chore(release): cut 0.1.71

Reload was a no-op for imported agents whose on-disk session was mutated
externally. The server preserved the in-memory timeline + historyPrimed
flag, so hydrate short-circuited and the epoch never changed. The client
also fetched without a cursor on reload, so the server returned
reset:false and the reducer dropped the new-epoch rows against a stale
cursor — the chat stayed stale until refocus.

Server: reloadAgentSession gains rehydrateFromDisk (used by the refresh
RPC). When set, durable + in-memory timeline are wiped and historyPrimed
is reset, so register mints a new epoch and hydrateTimelineFromProvider
re-streams the freshly read provider history. Voice-mode reloads keep
the existing preserve-timeline semantics.

Client: handleReloadAgent shows a sticky "Reloading agent…" toast,
sends the stale cursor with the fetch so the server returns reset:true,
and only flips to "Reloaded agent" after the response listener has
replaced the visible tail.

.github/FUNDING.yml

@@ -0,0 +1,15 @@
+# These are supported funding model platforms
+
+github: [boudra]
+patreon: # Replace with a single Patreon username
+open_collective: # Replace with a single Open Collective username
+ko_fi: # Replace with a single Ko-fi username
+tidelift: # Replace with a single Tidelift platform-name/package-name e.g., npm/babel
+community_bridge: # Replace with a single Community Bridge project-name e.g., cloud-foundry
+liberapay: # Replace with a single Liberapay username
+issuehunt: # Replace with a single IssueHunt username
+lfx_crowdfunding: # Replace with a single LFX Crowdfunding project-name e.g., cloud-foundry
+polar: # Replace with a single Polar username
+buy_me_a_coffee: # Replace with a single Buy Me a Coffee username
+thanks_dev: # Replace with a single thanks.dev username
+custom: # Replace with up to 4 custom sponsorship URLs e.g., ['link1', 'link2']

.github/workflows/android-apk-release.yml

@@ -34,15 +34,33 @@ jobs:
 
       - name: Resolve release tag
         shell: bash
+        run: node scripts/emit-release-env.mjs --source-tag "$SOURCE_TAG" >> "$GITHUB_ENV"
+
+      - name: Ensure GitHub release exists
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        shell: bash
         run: |
           set -euo pipefail
-          source_tag="${SOURCE_TAG}"
-          if [[ "$source_tag" =~ ^(android-)?v([0-9]+\.[0-9]+\.[0-9]+) ]]; then
-            release_tag="v${BASH_REMATCH[2]}"
-          else
-            release_tag="$source_tag"
+
+          if gh release view "$RELEASE_TAG" --repo "${{ github.repository }}" >/dev/null 2>&1; then
+            exit 0
+          fi
+
+          release_args=(
+            release create "$RELEASE_TAG"
+            --repo "${{ github.repository }}"
+            --title "Paseo $RELEASE_TAG"
+            --notes ""
+          )
+
+          if [[ "$IS_PRERELEASE" == "true" ]]; then
+            release_args+=(--prerelease)
+          fi
+
+          if ! gh "${release_args[@]}"; then
+            echo "Release creation raced with another workflow; continuing."
           fi
-          echo "RELEASE_TAG=$release_tag" >> "$GITHUB_ENV"
 
       - name: Setup Node
         uses: actions/setup-node@v4
@@ -107,24 +125,6 @@ jobs:
           echo "asset_name=$asset_name" >> "$GITHUB_OUTPUT"
           echo "asset_path=$asset_path" >> "$GITHUB_OUTPUT"
 
-      - name: Wait for GitHub release tag
-        env:
-          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        shell: bash
-        run: |
-          set -euo pipefail
-          for attempt in $(seq 1 90); do
-            if gh release view "$RELEASE_TAG" --repo "${{ github.repository }}" >/dev/null 2>&1; then
-              echo "Found release for tag $RELEASE_TAG"
-              exit 0
-            fi
-            echo "Release for $RELEASE_TAG is not available yet (attempt $attempt/90)."
-            sleep 20
-          done
-
-          echo "Timed out waiting for GitHub release tag $RELEASE_TAG."
-          exit 1
-
       - name: Upload APK to GitHub Release
         env:
           GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.github/workflows/ci.yml

@@ -0,0 +1,250 @@
+name: CI
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+  workflow_dispatch:
+
+jobs:
+  format:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
+
+      - name: Install dependencies
+        run: npm install
+
+      - name: Check formatting
+        run: npx oxfmt --check .
+
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
+
+      - name: Install dependencies
+        run: npm install
+
+      - name: Lint
+        run: npm run lint
+
+  typecheck:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
+
+      - name: Install dependencies
+        run: npm install
+
+      - name: Build highlight dependency
+        run: npm run build --workspace=@getpaseo/highlight
+
+      - name: Build relay dependency
+        run: npm run build --workspace=@getpaseo/relay
+
+      - name: Build server dependency
+        run: npm run build --workspace=@getpaseo/server
+
+      - name: Typecheck all packages
+        run: npm run typecheck
+
+  server-tests:
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest, windows-latest]
+    runs-on: ${{ matrix.os }}
+    name: server-tests (${{ matrix.os }})
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
+
+      - name: Fetch origin/main (worktree tests)
+        run: git fetch --no-tags origin main:refs/remotes/origin/main
+
+      - name: Install dependencies
+        run: npm install
+
+      - name: Install Claude Code CLI for provider tests
+        run: npm install -g @anthropic-ai/claude-code
+
+      - name: Build highlight dependency
+        run: npm run build --workspace=@getpaseo/highlight
+
+      - name: Build relay dependency
+        run: npm run build --workspace=@getpaseo/relay
+
+      - name: Run server tests
+        run: npm run test --workspace=@getpaseo/server
+        env:
+          CLAUDE_CODE_OAUTH_TOKEN: ${{ secrets.CLAUDE_CODE_OAUTH_TOKEN }}
+          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+
+  desktop-tests:
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest, windows-latest]
+    runs-on: ${{ matrix.os }}
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
+
+      - name: Install dependencies
+        run: npm install
+
+      - name: Build highlight dependency
+        run: npm run build --workspace=@getpaseo/highlight
+
+      - name: Build relay dependency
+        run: npm run build --workspace=@getpaseo/relay
+
+      - name: Build server dependency
+        run: npm run build --workspace=@getpaseo/server
+
+      - name: Run desktop tests
+        run: npm run test --workspace=@getpaseo/desktop
+
+  app-tests:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
+
+      - name: Install dependencies
+        run: npm install
+
+      - name: Install Playwright browsers
+        run: npx playwright install --with-deps chromium
+
+      - name: Build highlight dependency
+        run: npm run build --workspace=@getpaseo/highlight
+
+      - name: Run app unit tests
+        run: npm run test --workspace=@getpaseo/app
+
+  playwright:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
+
+      - name: Install dependencies
+        run: npm install
+
+      - name: Install Playwright browsers
+        run: npx playwright install --with-deps chromium
+
+      - name: Build highlight dependency
+        run: npm run build --workspace=@getpaseo/highlight
+
+      - name: Build relay dependency
+        run: npm run build --workspace=@getpaseo/relay
+
+      - name: Build server dependency
+        run: npm run build --workspace=@getpaseo/server
+
+      - name: Install agent CLIs for provider tests
+        run: npm install -g @anthropic-ai/claude-code @openai/codex@0.105.0 opencode-ai
+
+      - name: Run Playwright E2E tests
+        run: npm run test:e2e --workspace=@getpaseo/app
+        env:
+          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+
+      - name: Upload test artifacts
+        uses: actions/upload-artifact@v4
+        if: failure()
+        with:
+          name: playwright-results
+          path: |
+            packages/app/test-results/
+            packages/app/playwright-report/
+          retention-days: 7
+
+  relay-tests:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
+
+      - name: Install dependencies
+        run: npm install
+
+      - name: Build relay
+        run: npm run build --workspace=@getpaseo/relay
+
+      - name: Run relay tests
+        run: npm run test --workspace=@getpaseo/relay
+
+  cli-tests:
+    strategy:
+      fail-fast: false
+      matrix:
+        shard: [1, 2, 3]
+    runs-on: ubuntu-latest
+    name: cli-tests (shard ${{ matrix.shard }}/3)
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
+
+      - name: Install dependencies
+        run: npm install
+
+      - name: Install agent CLIs for provider tests
+        run: npm install -g @anthropic-ai/claude-code @openai/codex@0.105.0 opencode-ai
+
+      - name: Build highlight dependency
+        run: npm run build --workspace=@getpaseo/highlight
+
+      - name: Run CLI tests
+        run: npm run test --workspace=@getpaseo/cli
+        env:
+          PASEO_LOCAL_SPEECH_AUTO_DOWNLOAD: "0"
+          PASEO_DICTATION_ENABLED: "0"
+          PASEO_VOICE_MODE_ENABLED: "0"
+          PASEO_CLI_TEST_SHARD: ${{ matrix.shard }}
+          PASEO_CLI_TEST_SHARD_TOTAL: "3"

.github/workflows/deploy-app.yml

@@ -3,8 +3,10 @@ name: Deploy App
 on:
   push:
     tags:
-      - 'v*'
-      - 'app-v*'
+      - "v*"
+      - "!v*-beta.*"
+      - "app-v*"
+      - "!app-v*-beta.*"
   workflow_dispatch:
 
 jobs:
@@ -16,13 +18,13 @@ jobs:
 
       - uses: actions/setup-node@v4
         with:
-          node-version: '22'
-          cache: 'npm'
-          registry-url: 'https://npm.pkg.github.com'
-          scope: '@boudra'
+          node-version: "22"
+          cache: "npm"
+          registry-url: "https://npm.pkg.github.com"
+          scope: "@boudra"
 
       - name: Install dependencies
-        run: npm install --workspace=@getpaseo/app --include-workspace-root
+        run: npm ci
         env:
           NODE_AUTH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
 

.github/workflows/deploy-relay.yml

@@ -4,8 +4,8 @@ on:
   push:
     branches: [main]
     paths:
-      - 'packages/relay/**'
-      - '.github/workflows/deploy-relay.yml'
+      - "packages/relay/**"
+      - ".github/workflows/deploy-relay.yml"
   workflow_dispatch:
 
 jobs:
@@ -17,8 +17,8 @@ jobs:
 
       - uses: actions/setup-node@v4
         with:
-          node-version: '22'
-          cache: 'npm'
+          node-version: "22"
+          cache: "npm"
 
       - name: Install dependencies
         run: npm install --workspace=@getpaseo/relay --include-workspace-root
@@ -30,4 +30,3 @@ jobs:
         run: cd packages/relay && npx wrangler deploy
         env:
           CLOUDFLARE_API_TOKEN: ${{ secrets.CLOUDFLARE_API_TOKEN }}
-

.github/workflows/deploy-website.yml

@@ -4,17 +4,19 @@ on:
   push:
     branches: [main]
     paths:
-      - 'packages/website/**'
-      - 'package.json'
-      - 'package-lock.json'
-      - 'patches/**'
-      - '.github/workflows/deploy-website.yml'
+      - "CHANGELOG.md"
+      - "packages/website/**"
+      - "package.json"
+      - "package-lock.json"
+      - "patches/**"
+      - ".github/workflows/deploy-website.yml"
   release:
-    types: [published, edited]
+    types: [published]
   workflow_dispatch:
 
 jobs:
   deploy:
+    if: ${{ github.event_name != 'release' || (!github.event.release.prerelease && !github.event.release.draft) }}
     runs-on: ubuntu-latest
 
     steps:
@@ -22,8 +24,8 @@ jobs:
 
       - uses: actions/setup-node@v4
         with:
-          node-version: '22'
-          cache: 'npm'
+          node-version: "22"
+          cache: "npm"
 
       - name: Install dependencies
         run: npm install --workspace=@getpaseo/website --include-workspace-root

.github/workflows/desktop-release.yml

@@ -3,27 +3,45 @@ name: Desktop Release
 on:
   push:
     tags:
-      - 'v*'
-      - 'desktop-v*'
-      - 'desktop-macos-v*'
-      - 'desktop-linux-v*'
-      - 'desktop-windows-v*'
+      - "v*"
+      - "desktop-v*"
+      - "desktop-macos-v*"
+      - "desktop-linux-v*"
+      - "desktop-windows-v*"
   workflow_dispatch:
     inputs:
       tag:
-        description: 'Existing tag to build (e.g. v0.1.0)'
+        description: "Existing tag to build (e.g. v0.1.0)"
         required: true
         type: string
       platform:
-        description: 'Optional desktop platform to build.'
+        description: "Optional desktop platform to build."
         required: false
-        default: 'all'
+        default: "all"
         type: choice
         options:
           - all
           - macos
           - linux
           - windows
+      checkout_ref:
+        description: "Optional branch/ref to checkout while using tag for release metadata."
+        required: false
+        default: ""
+        type: string
+      publish:
+        description: "Publish built artifacts to GitHub Releases."
+        required: false
+        default: "true"
+        type: choice
+        options:
+          - "true"
+          - "false"
+      rollout_hours:
+        description: "Linear rollout duration in hours. Use 0 for instant rollout."
+        required: false
+        default: "24"
+        type: string
 
 concurrency:
   group: desktop-release-${{ github.ref }}
@@ -31,12 +49,51 @@ concurrency:
 
 env:
   SOURCE_TAG: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.tag || github.ref_name }}
-  DESKTOP_WORKSPACE: '@getpaseo/desktop'
-  DESKTOP_PACKAGE_PATH: 'packages/desktop'
+  CHECKOUT_REF: ${{ github.event_name == 'workflow_dispatch' && (github.event.inputs.checkout_ref || github.ref_name) || github.ref_name }}
+  SHOULD_PUBLISH: ${{ github.event_name != 'workflow_dispatch' || github.event.inputs.publish != 'false' }}
+  ROLLOUT_HOURS: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.rollout_hours || '24' }}
+  DESKTOP_PACKAGE_PATH: "packages/desktop"
 
 jobs:
+  create-release:
+    if: ${{ (github.event_name == 'push' && !startsWith(github.ref_name, 'desktop-macos-v') && !startsWith(github.ref_name, 'desktop-linux-v') && !startsWith(github.ref_name, 'desktop-windows-v')) || (github.event_name == 'workflow_dispatch' && github.event.inputs.platform == 'all' && github.event.inputs.publish != 'false') }}
+    permissions:
+      contents: write
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          sparse-checkout: scripts
+          ref: ${{ env.CHECKOUT_REF }}
+
+      - name: Resolve release metadata
+        shell: bash
+        run: node scripts/emit-release-env.mjs --source-tag "$SOURCE_TAG" >> "$GITHUB_ENV"
+
+      - name: Create GitHub release
+        if: env.IS_SMOKE_TAG != 'true'
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          if gh release view "$RELEASE_TAG" --repo "${{ github.repository }}" > /dev/null 2>&1; then
+            echo "Release $RELEASE_TAG already exists, skipping creation"
+          else
+            prerelease_flag=""
+            if [[ "$IS_PRERELEASE" == "true" ]]; then
+              prerelease_flag="--prerelease"
+            fi
+            gh release create "$RELEASE_TAG" \
+              --repo "${{ github.repository }}" \
+              --title "Paseo $RELEASE_TAG" \
+              --notes "" \
+              $prerelease_flag || {
+                echo "Release creation raced with another workflow; continuing."
+              }
+          fi
+
   publish-macos:
-    if: ${{ (github.event_name == 'workflow_dispatch' && (github.event.inputs.platform == 'all' || github.event.inputs.platform == 'macos')) || (github.event_name == 'push' && (startsWith(github.ref_name, 'v') || startsWith(github.ref_name, 'desktop-v') || startsWith(github.ref_name, 'desktop-macos-v'))) }}
+    needs: [create-release]
+    if: ${{ always() && (needs.create-release.result == 'success' || needs.create-release.result == 'skipped') && ((github.event_name == 'workflow_dispatch' && (github.event.inputs.platform == 'all' || github.event.inputs.platform == 'macos')) || (github.event_name == 'push' && (startsWith(github.ref_name, 'v') || startsWith(github.ref_name, 'desktop-v') || startsWith(github.ref_name, 'desktop-macos-v')))) }}
     strategy:
       fail-fast: false
       matrix:
@@ -54,37 +111,20 @@ jobs:
       - uses: actions/checkout@v4
         with:
           fetch-depth: 0
-          ref: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.tag || github.ref }}
+          ref: ${{ env.CHECKOUT_REF }}
 
       - name: Resolve release metadata
         shell: bash
-        run: |
-          set -euo pipefail
-          source_tag="${SOURCE_TAG}"
-          if [[ "$source_tag" =~ ^(desktop-(windows|linux|macos)-|desktop-)?v([0-9]+\.[0-9]+\.[0-9]+) ]]; then
-            release_tag="v${BASH_REMATCH[3]}"
-          else
-            release_tag="$source_tag"
-          fi
-          echo "RELEASE_TAG=$release_tag" >> "$GITHUB_ENV"
-
-          version="${release_tag#v}"
-          echo "DESKTOP_VERSION=$version" >> "$GITHUB_ENV"
-
-          if [[ "$source_tag" == *gha-smoke* ]]; then
-            echo "IS_SMOKE_TAG=true" >> "$GITHUB_ENV"
-          else
-            echo "IS_SMOKE_TAG=false" >> "$GITHUB_ENV"
-          fi
+        run: node scripts/emit-release-env.mjs --source-tag "$SOURCE_TAG" >> "$GITHUB_ENV"
 
       - name: Setup Node
         uses: actions/setup-node@v4
         with:
-          node-version: '22'
-          cache: 'npm'
+          node-version: "22"
+          cache: "npm"
           cache-dependency-path: package-lock.json
-          registry-url: 'https://npm.pkg.github.com'
-          scope: '@boudra'
+          registry-url: "https://npm.pkg.github.com"
+          scope: "@boudra"
 
       - name: Install JS dependencies
         run: npm ci
@@ -107,27 +147,6 @@ jobs:
           fs.writeFileSync(packageJsonPath, `${JSON.stringify(packageJson, null, 2)}\n`);
           NODE
 
-      - name: Build web app for desktop
-        run: npm run build:web --workspace=@getpaseo/app
-
-      - name: Detect existing GitHub release state
-        if: env.IS_SMOKE_TAG != 'true'
-        env:
-          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        shell: bash
-        run: |
-          set -euo pipefail
-          if release_draft="$(gh release view "$RELEASE_TAG" --repo "${{ github.repository }}" --json isDraft --jq '.isDraft' 2>/dev/null)"; then
-            if [[ "$release_draft" == "true" ]]; then
-              release_type="draft"
-            else
-              release_type="release"
-            fi
-          else
-            release_type="draft"
-          fi
-          echo "RELEASE_TYPE=$release_type" >> "$GITHUB_ENV"
-
       - name: Build desktop release
         shell: bash
         env:
@@ -138,19 +157,43 @@ jobs:
           APPLE_ID: ${{ secrets.APPLE_ID }}
           APPLE_APP_SPECIFIC_PASSWORD: ${{ secrets.APPLE_PASSWORD }}
           APPLE_TEAM_ID: ${{ secrets.APPLE_TEAM_ID }}
+          PASEO_DESKTOP_SMOKE: "1"
         run: |
           set -euo pipefail
-          publish_mode="never"
-          publish_args=()
-          if [[ "$IS_SMOKE_TAG" != "true" ]]; then
-            publish_mode="always"
-            publish_args+=("-c.publish.releaseType=$RELEASE_TYPE")
-          fi
+          build_args=(-- --publish never --mac --${{ matrix.electron_arch }})
+          build_args+=("-c.publish.releaseType=$RELEASE_TYPE")
+          build_args+=("-c.publish.channel=$RELEASE_CHANNEL")
+          npm run build:desktop "${build_args[@]}"
 
-          npm run build --workspace="$DESKTOP_WORKSPACE" -- --publish "$publish_mode" --mac --${{ matrix.electron_arch }} "${publish_args[@]}"
+      - name: Upload desktop artifacts to release
+        if: env.SHOULD_PUBLISH == 'true' && env.IS_SMOKE_TAG != 'true'
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        shell: bash
+        run: |
+          set -euo pipefail
+          release_dir="${DESKTOP_PACKAGE_PATH}/release"
+          files=()
+          while IFS= read -r -d '' f; do
+            files+=("$f")
+          done < <(find "$release_dir" -maxdepth 1 -type f ! -name '*.yml' -print0 | sort -z)
+          if (( ${#files[@]} == 0 )); then
+            echo "::error::No release artifacts found in $release_dir"
+            exit 1
+          fi
+          gh release upload "$RELEASE_TAG" "${files[@]}" --clobber --repo "${{ github.repository }}"
+
+      - name: Upload manifest artifact
+        if: env.SHOULD_PUBLISH == 'true' && env.IS_SMOKE_TAG != 'true'
+        uses: actions/upload-artifact@v4
+        with:
+          name: mac-manifest-${{ matrix.electron_arch }}
+          path: ${{ env.DESKTOP_PACKAGE_PATH }}/release/${{ env.RELEASE_CHANNEL }}-mac.yml
+          retention-days: 1
 
   publish-linux:
-    if: ${{ (github.event_name == 'workflow_dispatch' && (github.event.inputs.platform == 'all' || github.event.inputs.platform == 'linux')) || (github.event_name == 'push' && (startsWith(github.ref_name, 'v') || startsWith(github.ref_name, 'desktop-v') || startsWith(github.ref_name, 'desktop-linux-v'))) }}
+    needs: [create-release]
+    if: ${{ always() && (needs.create-release.result == 'success' || needs.create-release.result == 'skipped') && ((github.event_name == 'workflow_dispatch' && (github.event.inputs.platform == 'all' || github.event.inputs.platform == 'linux')) || (github.event_name == 'push' && (startsWith(github.ref_name, 'v') || startsWith(github.ref_name, 'desktop-v') || startsWith(github.ref_name, 'desktop-linux-v')))) }}
     permissions:
       contents: write
       packages: read
@@ -160,37 +203,20 @@ jobs:
       - uses: actions/checkout@v4
         with:
           fetch-depth: 0
-          ref: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.tag || github.ref }}
+          ref: ${{ env.CHECKOUT_REF }}
 
       - name: Resolve release metadata
         shell: bash
-        run: |
-          set -euo pipefail
-          source_tag="${SOURCE_TAG}"
-          if [[ "$source_tag" =~ ^(desktop-(windows|linux|macos)-|desktop-)?v([0-9]+\.[0-9]+\.[0-9]+) ]]; then
-            release_tag="v${BASH_REMATCH[3]}"
-          else
-            release_tag="$source_tag"
-          fi
-          echo "RELEASE_TAG=$release_tag" >> "$GITHUB_ENV"
-
-          version="${release_tag#v}"
-          echo "DESKTOP_VERSION=$version" >> "$GITHUB_ENV"
-
-          if [[ "$source_tag" == *gha-smoke* ]]; then
-            echo "IS_SMOKE_TAG=true" >> "$GITHUB_ENV"
-          else
-            echo "IS_SMOKE_TAG=false" >> "$GITHUB_ENV"
-          fi
+        run: node scripts/emit-release-env.mjs --source-tag "$SOURCE_TAG" >> "$GITHUB_ENV"
 
       - name: Setup Node
         uses: actions/setup-node@v4
         with:
-          node-version: '22'
-          cache: 'npm'
+          node-version: "22"
+          cache: "npm"
           cache-dependency-path: package-lock.json
-          registry-url: 'https://npm.pkg.github.com'
-          scope: '@boudra'
+          registry-url: "https://npm.pkg.github.com"
+          scope: "@boudra"
 
       - name: Install JS dependencies
         run: npm ci
@@ -213,45 +239,51 @@ jobs:
           fs.writeFileSync(packageJsonPath, `${JSON.stringify(packageJson, null, 2)}\n`);
           NODE
 
-      - name: Build web app for desktop
-        run: npm run build:web --workspace=@getpaseo/app
-
-      - name: Detect existing GitHub release state
-        if: env.IS_SMOKE_TAG != 'true'
-        env:
-          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        shell: bash
-        run: |
-          set -euo pipefail
-          if release_draft="$(gh release view "$RELEASE_TAG" --repo "${{ github.repository }}" --json isDraft --jq '.isDraft' 2>/dev/null)"; then
-            if [[ "$release_draft" == "true" ]]; then
-              release_type="draft"
-            else
-              release_type="release"
-            fi
-          else
-            release_type="draft"
-          fi
-          echo "RELEASE_TYPE=$release_type" >> "$GITHUB_ENV"
+      - name: Install Linux smoke display
+        run: sudo apt-get update && sudo apt-get install -y xvfb
 
       - name: Build desktop release
         shell: bash
         env:
           GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
           EP_GH_IGNORE_TIME: true
+          PASEO_DESKTOP_SMOKE: "1"
         run: |
           set -euo pipefail
-          publish_mode="never"
-          publish_args=()
-          if [[ "$IS_SMOKE_TAG" != "true" ]]; then
-            publish_mode="always"
-            publish_args+=("-c.publish.releaseType=$RELEASE_TYPE")
-          fi
+          build_args=(-- --publish never --linux --x64)
+          build_args+=("-c.publish.releaseType=$RELEASE_TYPE")
+          build_args+=("-c.publish.channel=$RELEASE_CHANNEL")
+          npm run build:desktop "${build_args[@]}"
 
-          npm run build --workspace="$DESKTOP_WORKSPACE" -- --publish "$publish_mode" --linux --x64 "${publish_args[@]}"
+      - name: Upload desktop artifacts to release
+        if: env.SHOULD_PUBLISH == 'true' && env.IS_SMOKE_TAG != 'true'
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        shell: bash
+        run: |
+          set -euo pipefail
+          release_dir="${DESKTOP_PACKAGE_PATH}/release"
+          files=()
+          while IFS= read -r -d '' f; do
+            files+=("$f")
+          done < <(find "$release_dir" -maxdepth 1 -type f ! -name '*.yml' -print0 | sort -z)
+          if (( ${#files[@]} == 0 )); then
+            echo "::error::No release artifacts found in $release_dir"
+            exit 1
+          fi
+          gh release upload "$RELEASE_TAG" "${files[@]}" --clobber --repo "${{ github.repository }}"
+
+      - name: Upload manifest artifact
+        if: env.SHOULD_PUBLISH == 'true' && env.IS_SMOKE_TAG != 'true'
+        uses: actions/upload-artifact@v4
+        with:
+          name: linux-manifest
+          path: ${{ env.DESKTOP_PACKAGE_PATH }}/release/${{ env.RELEASE_CHANNEL }}-linux.yml
+          retention-days: 1
 
   publish-windows:
-    if: ${{ (github.event_name == 'workflow_dispatch' && (github.event.inputs.platform == 'all' || github.event.inputs.platform == 'windows')) || (github.event_name == 'push' && (startsWith(github.ref_name, 'v') || startsWith(github.ref_name, 'desktop-v') || startsWith(github.ref_name, 'desktop-windows-v'))) }}
+    needs: [create-release]
+    if: ${{ always() && (needs.create-release.result == 'success' || needs.create-release.result == 'skipped') && ((github.event_name == 'workflow_dispatch' && (github.event.inputs.platform == 'all' || github.event.inputs.platform == 'windows')) || (github.event_name == 'push' && (startsWith(github.ref_name, 'v') || startsWith(github.ref_name, 'desktop-v') || startsWith(github.ref_name, 'desktop-windows-v')))) }}
     permissions:
       contents: write
       packages: read
@@ -261,37 +293,20 @@ jobs:
       - uses: actions/checkout@v4
         with:
           fetch-depth: 0
-          ref: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.tag || github.ref }}
+          ref: ${{ env.CHECKOUT_REF }}
 
       - name: Resolve release metadata
         shell: bash
-        run: |
-          set -euo pipefail
-          source_tag="${SOURCE_TAG}"
-          if [[ "$source_tag" =~ ^(desktop-(windows|linux|macos)-|desktop-)?v([0-9]+\.[0-9]+\.[0-9]+) ]]; then
-            release_tag="v${BASH_REMATCH[3]}"
-          else
-            release_tag="$source_tag"
-          fi
-          echo "RELEASE_TAG=$release_tag" >> "$GITHUB_ENV"
-
-          version="${release_tag#v}"
-          echo "DESKTOP_VERSION=$version" >> "$GITHUB_ENV"
-
-          if [[ "$source_tag" == *gha-smoke* ]]; then
-            echo "IS_SMOKE_TAG=true" >> "$GITHUB_ENV"
-          else
-            echo "IS_SMOKE_TAG=false" >> "$GITHUB_ENV"
-          fi
+        run: node scripts/emit-release-env.mjs --source-tag "$SOURCE_TAG" >> "$GITHUB_ENV"
 
       - name: Setup Node
         uses: actions/setup-node@v4
         with:
-          node-version: '22'
-          cache: 'npm'
+          node-version: "22"
+          cache: "npm"
           cache-dependency-path: package-lock.json
-          registry-url: 'https://npm.pkg.github.com'
-          scope: '@boudra'
+          registry-url: "https://npm.pkg.github.com"
+          scope: "@boudra"
 
       - name: Install JS dependencies
         run: npm ci
@@ -314,47 +329,157 @@ jobs:
           fs.writeFileSync(packageJsonPath, `${JSON.stringify(packageJson, null, 2)}\n`);
           NODE
 
-      - name: Build workspace dependencies
-        run: npm run build:workspace-deps --workspace=@getpaseo/app
-
-      - name: Build web app for desktop
-        shell: pwsh
+      - name: Build desktop release
+        shell: bash
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          EP_GH_IGNORE_TIME: true
+          PASEO_DESKTOP_SMOKE: "1"
         run: |
-          $patchPath = (Get-Item "$env:GITHUB_WORKSPACE/scripts/metro-config-windows-loader-patch.cjs").FullName
-          $env:NODE_OPTIONS = "--require=$patchPath"
-          npx expo export --platform web
-        working-directory: packages/app
+          set -euo pipefail
+          build_args=(-- --publish never --win --x64 --arm64)
+          build_args+=("-c.publish.releaseType=$RELEASE_TYPE")
+          build_args+=("-c.publish.channel=$RELEASE_CHANNEL")
+          npm run build:desktop "${build_args[@]}"
 
-      - name: Detect existing GitHub release state
+      - name: Upload desktop artifacts to release
+        if: env.SHOULD_PUBLISH == 'true' && env.IS_SMOKE_TAG != 'true'
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        shell: bash
+        run: |
+          set -euo pipefail
+          release_dir="${DESKTOP_PACKAGE_PATH}/release"
+          files=()
+          while IFS= read -r -d '' f; do
+            files+=("$f")
+          done < <(find "$release_dir" -maxdepth 1 -type f ! -name '*.yml' -print0 | sort -z)
+          if (( ${#files[@]} == 0 )); then
+            echo "::error::No release artifacts found in $release_dir"
+            exit 1
+          fi
+          gh release upload "$RELEASE_TAG" "${files[@]}" --clobber --repo "${{ github.repository }}"
+
+      - name: Upload manifest artifact
+        if: env.SHOULD_PUBLISH == 'true' && env.IS_SMOKE_TAG != 'true'
+        uses: actions/upload-artifact@v4
+        with:
+          name: windows-manifest
+          path: ${{ env.DESKTOP_PACKAGE_PATH }}/release/${{ env.RELEASE_CHANNEL }}.yml
+          retention-days: 1
+
+  finalize-rollout:
+    needs: [publish-macos, publish-linux, publish-windows]
+    if: ${{ always() && (needs.publish-macos.result == 'success' || needs.publish-macos.result == 'skipped') && (needs.publish-linux.result == 'success' || needs.publish-linux.result == 'skipped') && (needs.publish-windows.result == 'success' || needs.publish-windows.result == 'skipped') && (github.event_name != 'workflow_dispatch' || github.event.inputs.publish != 'false') }}
+    permissions:
+      contents: write
+    runs-on: ubuntu-latest
+    concurrency:
+      group: desktop-rollout-${{ github.event.inputs.tag || github.ref_name }}
+      cancel-in-progress: false
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          sparse-checkout: |
+            package.json
+            package-lock.json
+            scripts
+          ref: ${{ env.CHECKOUT_REF }}
+
+      - name: Resolve release tag
+        shell: bash
+        run: node scripts/emit-release-env.mjs --source-tag "$SOURCE_TAG" >> "$GITHUB_ENV"
+
+      - name: Setup Node
+        uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
+          cache-dependency-path: package-lock.json
+          registry-url: "https://npm.pkg.github.com"
+          scope: "@boudra"
+
+      - name: Install JS dependencies
+        run: npm ci
+        env:
+          NODE_AUTH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Download mac manifest artifacts
+        if: env.IS_SMOKE_TAG != 'true' && needs.publish-macos.result == 'success'
+        uses: actions/download-artifact@v4
+        with:
+          pattern: mac-manifest-*
+          path: release-manifests
+
+      - name: Download Linux manifest artifact
+        if: env.IS_SMOKE_TAG != 'true' && needs.publish-linux.result == 'success'
+        uses: actions/download-artifact@v4
+        with:
+          name: linux-manifest
+          path: release-manifests/linux-manifest
+
+      - name: Download Windows manifest artifact
+        if: env.IS_SMOKE_TAG != 'true' && needs.publish-windows.result == 'success'
+        uses: actions/download-artifact@v4
+        with:
+          name: windows-manifest
+          path: release-manifests/windows-manifest
+
+      - name: Assemble and upload stamped manifests
         if: env.IS_SMOKE_TAG != 'true'
         env:
           GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         shell: bash
         run: |
           set -euo pipefail
-          if release_draft="$(gh release view "$RELEASE_TAG" --repo "${{ github.repository }}" --json isDraft --jq '.isDraft' 2>/dev/null)"; then
-            if [[ "$release_draft" == "true" ]]; then
-              release_type="draft"
-            else
-              release_type="release"
-            fi
-          else
-            release_type="draft"
-          fi
-          echo "RELEASE_TYPE=$release_type" >> "$GITHUB_ENV"
+          cd release-manifests
+          manifests_dir="$PWD/final"
+          mkdir -p "$manifests_dir"
 
-      - name: Build desktop release
-        shell: bash
-        env:
-          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          EP_GH_IGNORE_TIME: true
-        run: |
-          set -euo pipefail
-          publish_mode="never"
-          publish_args=()
-          if [[ "$IS_SMOKE_TAG" != "true" ]]; then
-            publish_mode="always"
-            publish_args+=("-c.publish.releaseType=$RELEASE_TYPE")
+          if [[ "${{ needs.publish-macos.result }}" == "success" ]]; then
+            manifest_name="${RELEASE_CHANNEL}-mac.yml"
+            node ../scripts/merge-mac-manifest.mjs \
+              "mac-manifest-arm64/${manifest_name}" \
+              "mac-manifest-x64/${manifest_name}" \
+              "$manifests_dir/${manifest_name}"
           fi
 
-          npm run build --workspace="$DESKTOP_WORKSPACE" -- --publish "$publish_mode" --win --x64 "${publish_args[@]}"
+          if [[ "${{ needs.publish-linux.result }}" == "success" ]]; then
+            cp "linux-manifest/${RELEASE_CHANNEL}-linux.yml" "$manifests_dir/"
+          fi
+
+          if [[ "${{ needs.publish-windows.result }}" == "success" ]]; then
+            cp "windows-manifest/${RELEASE_CHANNEL}.yml" "$manifests_dir/"
+          fi
+
+          shopt -s nullglob
+          files=( "$manifests_dir"/*.yml )
+          if (( ${#files[@]} == 0 )); then
+            echo "::error::No manifest artifacts were available to publish"
+            exit 1
+          fi
+          timestamp="$(date -u +"%Y-%m-%dT%H:%M:%S.000Z")"
+          node ../scripts/stamp-rollout.mjs --release-date "$timestamp" --rollout-hours "$ROLLOUT_HOURS" "${files[@]}"
+          ROLLOUT_HOURS_EXPECTED="$ROLLOUT_HOURS" RELEASE_DATE_EXPECTED="$timestamp" node -e '
+            const yaml = require("js-yaml");
+            const fs = require("fs");
+            const expectedHours = Number(process.env.ROLLOUT_HOURS_EXPECTED);
+            const expectedDate = process.env.RELEASE_DATE_EXPECTED;
+            if (!Number.isFinite(expectedHours) || expectedHours < 0) {
+              throw new Error(`expected non-negative rolloutHours, got ${process.env.ROLLOUT_HOURS_EXPECTED}`);
+            }
+            for (const f of process.argv.slice(1)) {
+              const m = yaml.load(fs.readFileSync(f, "utf8")) ?? {};
+              if (m.rolloutHours !== expectedHours) {
+                throw new Error(`${f}: rolloutHours=${m.rolloutHours}, expected ${expectedHours}`);
+              }
+              if (m.releaseDate !== expectedDate) {
+                throw new Error(`${f}: releaseDate=${m.releaseDate}, expected ${expectedDate}`);
+              }
+              if (typeof m.version !== "string" || m.version.length === 0) {
+                throw new Error(`${f}: missing or invalid version`);
+              }
+            }
+          ' "${files[@]}"
+          gh release upload "$RELEASE_TAG" "${files[@]}" --clobber --repo "${{ github.repository }}"

.github/workflows/desktop-rollout.yml

@@ -0,0 +1,152 @@
+name: Desktop Rollout
+
+on:
+  workflow_dispatch:
+    inputs:
+      tag:
+        description: "Existing release tag to re-stamp (e.g. v0.1.42)."
+        required: true
+        type: string
+      rollout_hours:
+        description: "Total rollout duration since the original release date, in hours. Use 0 to admit everyone immediately."
+        required: true
+        type: string
+
+concurrency:
+  group: desktop-rollout-${{ inputs.tag }}
+  cancel-in-progress: false
+
+env:
+  SOURCE_TAG: ${{ inputs.tag }}
+
+jobs:
+  stamp:
+    permissions:
+      contents: write
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          sparse-checkout: |
+            package.json
+            package-lock.json
+            scripts
+
+      - name: Resolve release metadata
+        shell: bash
+        run: node scripts/emit-release-env.mjs --source-tag "$SOURCE_TAG" >> "$GITHUB_ENV"
+
+      - name: Setup Node
+        uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
+          cache-dependency-path: package-lock.json
+          registry-url: "https://npm.pkg.github.com"
+          scope: "@boudra"
+
+      - name: Install JS dependencies
+        run: npm ci
+        env:
+          NODE_AUTH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Download release manifests
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        shell: bash
+        run: |
+          set -euo pipefail
+          mkdir release-manifests
+          cd release-manifests
+          gh release download "$RELEASE_TAG" --repo "${{ github.repository }}" --pattern "${RELEASE_CHANNEL}*.yml"
+          shopt -s nullglob
+          files=( ./*.yml )
+          if (( ${#files[@]} == 0 )); then
+            echo "::error::No manifests matched ${RELEASE_CHANNEL}*.yml on $RELEASE_TAG"
+            exit 1
+          fi
+          echo "Downloaded ${#files[@]} manifest(s):"
+          printf '  %s\n' "${files[@]}"
+
+      - name: Capture before state
+        id: before
+        shell: bash
+        run: |
+          set -euo pipefail
+          cd release-manifests
+          summary=$(node -e '
+            const yaml = require("js-yaml");
+            const fs = require("fs");
+            for (const f of process.argv.slice(1)) {
+              const m = yaml.load(fs.readFileSync(f, "utf8")) ?? {};
+              console.log(`  ${f}: rolloutHours=${m.rolloutHours ?? "<unset>"} releaseDate=${m.releaseDate ?? "<unset>"}`);
+            }
+          ' ./*.yml)
+          echo "$summary"
+          {
+            echo "summary<<EOF"
+            echo "$summary"
+            echo "EOF"
+          } >> "$GITHUB_OUTPUT"
+
+      - name: Re-stamp rolloutHours
+        env:
+          NEW_HOURS: ${{ inputs.rollout_hours }}
+        shell: bash
+        run: |
+          set -euo pipefail
+          cd release-manifests
+          node ../scripts/stamp-rollout.mjs --rollout-hours "$NEW_HOURS" ./*.yml
+
+      - name: Validate rewritten manifests
+        env:
+          EXPECTED: ${{ inputs.rollout_hours }}
+        shell: bash
+        run: |
+          set -euo pipefail
+          cd release-manifests
+          node -e '
+            const yaml = require("js-yaml");
+            const fs = require("fs");
+            const expected = Number(process.env.EXPECTED);
+            if (!Number.isFinite(expected) || expected < 0) {
+              throw new Error(`EXPECTED must be a non-negative number, got ${process.env.EXPECTED}`);
+            }
+            for (const f of process.argv.slice(1)) {
+              const m = yaml.load(fs.readFileSync(f, "utf8")) ?? {};
+              if (m.rolloutHours !== expected) {
+                throw new Error(`${f}: rolloutHours=${m.rolloutHours}, expected ${expected}`);
+              }
+              if (typeof m.version !== "string" || m.version.length === 0) {
+                throw new Error(`${f}: missing or invalid version`);
+              }
+            }
+          ' ./*.yml
+
+      - name: Upload to release
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        shell: bash
+        run: |
+          set -euo pipefail
+          cd release-manifests
+          gh release upload "$RELEASE_TAG" ./*.yml --clobber --repo "${{ github.repository }}"
+
+      - name: Write summary
+        env:
+          BEFORE: ${{ steps.before.outputs.summary }}
+        shell: bash
+        run: |
+          {
+            echo "## Rollout updated"
+            echo ""
+            echo "**Tag:** \`$RELEASE_TAG\`"
+            echo "**Channel:** \`$RELEASE_CHANNEL\`"
+            echo "**New rolloutHours:** \`${{ inputs.rollout_hours }}\`"
+            echo ""
+            echo "### Before"
+            echo '```'
+            echo "$BEFORE"
+            echo '```'
+          } >> "$GITHUB_STEP_SUMMARY"

.github/workflows/fix-nix-hash.yml

@@ -1,45 +0,0 @@
-name: Fix Nix hash
-
-on:
-  push:
-    branches: [main]
-    paths:
-      - 'package.json'
-      - 'package-lock.json'
-  pull_request:
-    paths:
-      - 'package.json'
-      - 'package-lock.json'
-
-permissions:
-  contents: write
-
-jobs:
-  fix-nix-hash:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          ref: ${{ github.head_ref || github.ref }}
-          token: ${{ secrets.GITHUB_TOKEN }}
-
-      - uses: actions/setup-node@v4
-        with:
-          node-version: '22'
-          cache: 'npm'
-
-      - uses: cachix/install-nix-action@v31
-        with:
-          nix_path: nixpkgs=channel:nixos-unstable
-
-      - name: Fix lockfile and update hash
-        run: ./scripts/update-nix.sh
-
-      - name: Commit changes
-        run: |
-          git diff --quiet package-lock.json nix/package.nix && exit 0
-          git config user.name "github-actions[bot]"
-          git config user.email "github-actions[bot]@users.noreply.github.com"
-          git add package-lock.json nix/package.nix
-          git commit -m "fix: update lockfile signatures and Nix hash"
-          git push

.github/workflows/nix-build.yml

@@ -4,27 +4,33 @@ on:
   push:
     branches: [main]
     paths:
-      - 'nix/**'
-      - 'flake.nix'
-      - 'flake.lock'
-      - 'package.json'
-      - 'package-lock.json'
-      - 'packages/highlight/**'
-      - 'packages/server/**'
-      - 'packages/relay/**'
-      - 'packages/cli/**'
+      - "nix/**"
+      - "flake.nix"
+      - "flake.lock"
+      - "package.json"
+      - "package-lock.json"
+      - "packages/highlight/**"
+      - "packages/server/**"
+      - "packages/relay/**"
+      - "packages/cli/**"
+      - "scripts/update-nix.sh"
+      - "scripts/fix-lockfile.mjs"
+      - ".github/workflows/nix-build.yml"
   pull_request:
     branches: [main]
     paths:
-      - 'nix/**'
-      - 'flake.nix'
-      - 'flake.lock'
-      - 'package.json'
-      - 'package-lock.json'
-      - 'packages/highlight/**'
-      - 'packages/server/**'
-      - 'packages/relay/**'
-      - 'packages/cli/**'
+      - "nix/**"
+      - "flake.nix"
+      - "flake.lock"
+      - "package.json"
+      - "package-lock.json"
+      - "packages/highlight/**"
+      - "packages/server/**"
+      - "packages/relay/**"
+      - "packages/cli/**"
+      - "scripts/update-nix.sh"
+      - "scripts/fix-lockfile.mjs"
+      - ".github/workflows/nix-build.yml"
 
 permissions:
   contents: read
@@ -32,28 +38,35 @@ permissions:
 jobs:
   build:
     runs-on: ubuntu-latest
-    continue-on-error: false
+    permissions:
+      contents: write
     steps:
       - uses: actions/checkout@v4
+        with:
+          ref: ${{ github.head_ref || github.ref }}
+          token: ${{ secrets.GITHUB_TOKEN }}
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: "22"
+          cache: "npm"
 
       - uses: cachix/install-nix-action@v31
         with:
           nix_path: nixpkgs=channel:nixos-unstable
 
+      - name: Update lockfile + Nix hash if stale
+        run: ./scripts/update-nix.sh
+
       - name: Build Nix package
         run: nix build .#default -o result
 
-      - name: Verify lockfile is complete
-        # npm silently omits resolved/integrity fields in workspace monorepos.
-        # Nix needs them for offline builds. See https://github.com/npm/cli/issues/4460
+      - name: Commit hash/lockfile updates (main push only)
+        if: github.event_name == 'push' && github.ref == 'refs/heads/main'
         run: |
-          node scripts/fix-lockfile.mjs package-lock.json
-          git diff --exit-code package-lock.json || {
-            echo "ERROR: package-lock.json has missing resolved/integrity fields."
-            echo "This is a known npm bug: https://github.com/npm/cli/issues/4460"
-            echo "Run 'node scripts/fix-lockfile.mjs' and commit the result."
-            exit 1
-          }
-
-      - name: Check npmDepsHash is up to date
-        run: ./scripts/update-nix.sh --check
+          git diff --quiet package-lock.json nix/package.nix && exit 0
+          git config user.name "github-actions[bot]"
+          git config user.email "github-actions[bot]@users.noreply.github.com"
+          git add package-lock.json nix/package.nix
+          git commit -m "fix: update lockfile signatures and Nix hash"
+          git push

.github/workflows/release-notes-sync.yml

@@ -19,11 +19,6 @@ on:
         required: false
         default: false
         type: boolean
-      draft:
-        description: "Create missing release as draft."
-        required: false
-        default: false
-        type: boolean
 
 concurrency:
   group: release-notes-sync-${{ github.event_name == 'workflow_dispatch' && github.event.inputs.tag || github.ref }}
@@ -48,7 +43,6 @@ jobs:
           REF: ${{ github.ref }}
           INPUT_TAG: ${{ github.event_name == 'push' && startsWith(github.ref, 'refs/tags/') && github.ref_name || github.event.inputs.tag }}
           INPUT_CREATE_IF_MISSING: ${{ github.event.inputs.create_if_missing }}
-          INPUT_DRAFT: ${{ github.event.inputs.draft }}
         shell: bash
         run: |
           set -euo pipefail
@@ -60,7 +54,7 @@ jobs:
           fi
 
           create_if_missing="false"
-          if [ "$EVENT_NAME" = "push" ] && [[ "$REF" == refs/tags/v* ]]; then
+          if [[ "$EVENT_NAME" = "push" && "$REF" == refs/tags/v* ]]; then
             create_if_missing="true"
           elif [ "$EVENT_NAME" = "workflow_dispatch" ] && [ "${INPUT_CREATE_IF_MISSING:-false}" = "true" ]; then
             create_if_missing="true"
@@ -70,8 +64,4 @@ jobs:
             args+=(--create-if-missing)
           fi
 
-          if [ "${INPUT_DRAFT:-false}" = "true" ]; then
-            args+=(--draft)
-          fi
-
           node scripts/sync-release-notes-from-changelog.mjs "${args[@]}"

.github/workflows/server-ci.yml

@@ -4,17 +4,17 @@ on:
   push:
     branches: [main]
     paths:
-      - 'packages/server/**'
-      - 'package.json'
-      - 'package-lock.json'
-      - '.github/workflows/server-ci.yml'
+      - "packages/server/**"
+      - "package.json"
+      - "package-lock.json"
+      - ".github/workflows/server-ci.yml"
   pull_request:
     branches: [main]
     paths:
-      - 'packages/server/**'
-      - 'package.json'
-      - 'package-lock.json'
-      - '.github/workflows/server-ci.yml'
+      - "packages/server/**"
+      - "package.json"
+      - "package-lock.json"
+      - ".github/workflows/server-ci.yml"
 
 jobs:
   test:
@@ -27,8 +27,8 @@ jobs:
 
       - uses: actions/setup-node@v4
         with:
-          node-version: '20'
-          cache: 'npm'
+          node-version: "22"
+          cache: "npm"
 
       - name: Fetch origin/main (worktree tests)
         run: git fetch --no-tags origin main:refs/remotes/origin/main
@@ -36,6 +36,9 @@ jobs:
       - name: Install server dependencies
         run: npm install --workspace=@getpaseo/server --include-workspace-root
 
+      - name: Install Claude Code CLI for provider tests
+        run: npm install -g @anthropic-ai/claude-code
+
       - name: Build highlight dependency
         run: npm run build --workspace=@getpaseo/highlight
 

.gitignore

@@ -13,6 +13,8 @@ out/
 .env.local
 .env.test.local
 .env*.local
+.dev.vars
+**/.dev.vars
 
 # Logs
 *.log
@@ -64,6 +66,7 @@ CLAUDE.local.md
 .paseo/
 .wrangler/
 **/.wrangler/
+**/.tanstack/
 
 # Local agent/tooling artifacts (do not commit)
 PLAN.md
@@ -83,3 +86,4 @@ packages/server/src/server/fixtures/dictation/dictation-debug-largest.transcript
 /artifacts
 packages/desktop/.cache/
 packages/desktop/src-tauri/resources/managed-runtime/
+app.json

.oxfmtrc.json

@@ -0,0 +1,15 @@
+{
+  "$schema": "./node_modules/oxfmt/configuration_schema.json",
+  "useTabs": false,
+  "tabWidth": 2,
+  "printWidth": 100,
+  "singleQuote": false,
+  "jsxSingleQuote": false,
+  "quoteProps": "as-needed",
+  "trailingComma": "all",
+  "semi": true,
+  "arrowParens": "always",
+  "bracketSameLine": false,
+  "bracketSpacing": true,
+  "ignorePatterns": ["*.lock", "**/*.gen.ts", "**/*.gen.tsx"]
+}

.oxlintrc.json

@@ -0,0 +1,93 @@
+{
+  "$schema": "./node_modules/oxlint/configuration_schema.json",
+  "options": {
+    "typeAware": false
+  },
+  "plugins": ["react", "react-perf", "unicorn", "typescript", "oxc", "import", "promise"],
+  "categories": {
+    "correctness": "error",
+    "suspicious": "error",
+    "perf": "error"
+  },
+  "rules": {
+    "react/react-in-jsx-scope": "off",
+
+    "no-await-in-loop": "off",
+
+    "no-unused-expressions": "error",
+    "no-useless-catch": "error",
+    "preserve-caught-error": "error",
+    "require-await": "off",
+    "no-async-promise-executor": "error",
+    "no-useless-escape": "error",
+    "no-empty-pattern": "error",
+    "no-self-assign": "error",
+    "no-shadow": "error",
+    "unicorn/consistent-function-scoping": "off",
+    "unicorn/no-array-sort": "off",
+
+    "unicorn/no-useless-spread": "error",
+    "unicorn/no-useless-fallback-in-spread": "error",
+    "unicorn/no-new-array": "error",
+    "unicorn/no-empty-file": "error",
+
+    "promise/always-return": "error",
+    "promise/no-multiple-resolved": "error",
+
+    "react/no-array-index-key": "error",
+    "react/jsx-no-useless-fragment": "error",
+    "react/jsx-no-constructed-context-values": "error",
+    "react/no-unescaped-entities": "error",
+    "react/button-has-type": "error",
+    "react/jsx-max-depth": ["error", { "max": 6 }],
+
+    "react-hooks/rules-of-hooks": "error",
+    "react-hooks/exhaustive-deps": "error",
+
+    "react-perf/jsx-no-new-array-as-prop": "error",
+    "react-perf/jsx-no-new-function-as-prop": "error",
+    "react-perf/jsx-no-new-object-as-prop": "error",
+    "react-perf/jsx-no-jsx-as-prop": "error",
+
+    "oxc/no-map-spread": "error",
+    "oxc/no-async-endpoint-handlers": "error",
+    "oxc/only-used-in-recursion": "error",
+
+    "typescript/no-explicit-any": "error",
+    "typescript/prefer-as-const": "error",
+    "typescript/no-this-alias": "error",
+    "typescript/no-unnecessary-type-assertion": "error",
+    "typescript/consistent-type-definitions": ["error", "interface"],
+
+    "import/no-unassigned-import": [
+      "error",
+      {
+        "allow": [
+          "**/*.css",
+          "**/expo-router/entry",
+          "**/event-target-polyfill",
+          "**/dotenv/config",
+          "**/react-native",
+          "**/@/styles/unistyles",
+          "**/src/styles/unistyles",
+          "**/@/test/window-local-storage"
+        ]
+      }
+    ],
+
+    "no-nested-ternary": "error",
+    "no-unneeded-ternary": "error",
+
+    "complexity": ["error", { "max": 20 }],
+    "max-depth": ["error", { "max": 4 }],
+    "max-nested-callbacks": ["error", { "max": 3 }]
+  },
+  "overrides": [
+    {
+      "files": ["**/e2e/fixtures.ts"],
+      "rules": {
+        "no-empty-pattern": "off"
+      }
+    }
+  ]
+}

CLAUDE.md

@@ -15,18 +15,32 @@ This is an npm workspace monorepo:
 - `packages/desktop` — Electron desktop wrapper
 - `packages/website` — Marketing site (paseo.sh)
 
-## Documentation
+## Docs
 
-| Doc | What's in it |
-|---|---|
-| [docs/ARCHITECTURE.md](docs/ARCHITECTURE.md) | System design, package layering, WebSocket protocol, agent lifecycle, data flow |
-| [docs/CODING_STANDARDS.md](docs/CODING_STANDARDS.md) | Type hygiene, error handling, state design, React patterns, file organization |
-| [docs/TESTING.md](docs/TESTING.md) | TDD workflow, determinism, real dependencies over mocks, test organization |
-| [docs/DEVELOPMENT.md](docs/DEVELOPMENT.md) | Dev server, build sync gotchas, CLI reference, agent state, Playwright MCP |
-| [docs/RELEASE.md](docs/RELEASE.md) | Release playbook, draft releases, completion checklist |
-| [docs/ANDROID.md](docs/ANDROID.md) | App variants, local/cloud builds, EAS workflows |
-| [docs/DESIGN.md](docs/DESIGN.md) | How to design features before implementation |
-| [SECURITY.md](SECURITY.md) | Relay threat model, E2E encryption, DNS rebinding, agent auth |
+`docs/` is the source of truth for system-level and process-level knowledge. **"The docs", "check the docs", or "check the X docs" always mean this directory — not the web.** Look here before fetching anything online; the docs capture gotchas and conventions you cannot derive from the code or external sources.
+
+At the start of non-trivial work, list `docs/` and skim anything relevant to the task. When you learn something meta worth preserving — a gotcha, a convention, a workflow, a piece of system context that will outlive the current task — update an existing doc or propose a new one. Code-level facts belong in inline comments next to the code; system, process, and gotcha-level facts belong in `docs/`.
+
+| Doc                                                            | What's in it                                                                                  |
+| -------------------------------------------------------------- | --------------------------------------------------------------------------------------------- |
+| [docs/product.md](docs/product.md)                             | What Paseo is, who it's for, where it's going                                                 |
+| [docs/architecture.md](docs/architecture.md)                   | System design, package layering, WebSocket protocol, agent lifecycle, data flow               |
+| [docs/agent-lifecycle.md](docs/agent-lifecycle.md)             | Agent states, parent/child relationships, archive semantics, tabs vs archive, subagents track |
+| [docs/data-model.md](docs/data-model.md)                       | File-based JSON persistence, Zod schemas, atomic writes, no migrations                        |
+| [docs/glossary.md](docs/glossary.md)                           | Authoritative terminology — UI label wins, no synonyms                                        |
+| [docs/coding-standards.md](docs/coding-standards.md)           | Type hygiene, error handling, state design, React patterns, file organization                 |
+| [docs/design.md](docs/design.md)                               | Theme tokens — colors, fonts, spacing, radii, icons                                           |
+| [docs/unistyles.md](docs/unistyles.md)                         | Unistyles gotchas — `useUnistyles()` is forbidden, alternatives in order                      |
+| [docs/file-icons.md](docs/file-icons.md)                       | Material icon theme integration for the file explorer                                         |
+| [docs/providers.md](docs/providers.md)                         | Adding a new agent provider end-to-end                                                        |
+| [docs/custom-providers.md](docs/custom-providers.md)           | Custom provider config: Z.AI, Alibaba/Qwen, ACP agents, profiles, custom binaries             |
+| [docs/development.md](docs/development.md)                     | Dev server, build sync gotchas, CLI reference, agent state, Playwright MCP                    |
+| [docs/testing.md](docs/testing.md)                             | TDD workflow, determinism, real dependencies over mocks, test organization                    |
+| [docs/mobile-testing.md](docs/mobile-testing.md)               | Maestro and mobile test workflows                                                             |
+| [docs/ad-hoc-daemon-testing.md](docs/ad-hoc-daemon-testing.md) | Isolated in-process daemon test harness                                                       |
+| [docs/android.md](docs/android.md)                             | App variants, local/cloud builds, EAS workflows                                               |
+| [docs/release.md](docs/release.md)                             | Release playbook, draft releases, completion checklist                                        |
+| [SECURITY.md](SECURITY.md)                                     | Relay threat model, E2E encryption, DNS rebinding, agent auth                                 |
 
 ## Quick start
 
@@ -35,18 +49,94 @@ npm run dev                          # Start daemon + Expo in Tmux
 npm run cli -- ls -a -g              # List all agents
 npm run cli -- daemon status         # Check daemon status
 npm run typecheck                    # Always run after changes
+npm run lint                         # Always run after changes
+npm run format                       # Auto-format with Biome
+npm run format:check                 # Check formatting without writing
 ```
 
-See [docs/DEVELOPMENT.md](docs/DEVELOPMENT.md) for full setup, build sync requirements, and debugging.
+See [docs/development.md](docs/development.md) for full setup, build sync requirements, and debugging.
 
 ## Critical rules
 
 - **NEVER restart the main Paseo daemon on port 6767 without permission** — it manages all running agents. If you're an agent, restarting it kills your own process.
 - **NEVER assume a timeout means the service needs restarting** — timeouts can be transient.
 - **NEVER add auth checks to tests** — agent providers handle their own auth.
-- **Always run typecheck after every change.**
+- **NEVER run the full test suite locally.** The test suites are heavy and will freeze the machine, especially if multiple agents run them in parallel. Rules:
+  - Run only the specific test file you changed: `npx vitest run <file> --bail=1`
+  - Never run `npm run test` for an entire workspace unless explicitly asked.
+  - If you must run a broad suite, pipe output to a file and read it afterward: `npx vitest run <file> --bail=1 > /tmp/test-output.txt 2>&1` then read the file.
+  - Never re-run a test suite that another agent already ran and reported green — trust the result.
+  - For full suite verification, push to CI and check GitHub Actions instead.
+- **Always run typecheck and lint after every change.**
+- **Build workspace packages before diagnosing cross-package type errors.** This repo consumes generated declarations across workspaces. If typecheck fails in a package that depends on another workspace (especially CLI depending on server/daemon types), rebuild the owning package first so `dist` declarations are current:
+  - `npm run build:daemon` — rebuild highlight, relay, server, and CLI when daemon/server/CLI types may be stale.
+  - Do not patch inferred callback parameters or add local duplicate types just to silence stale declaration errors.
+- **Run `npm run format` before committing.** This repo uses Biome for formatting. Do not manually fix formatting — let the formatter handle it.
+- **Always use npm scripts for linting and formatting.** Do not run tools directly with `npx eslint`, `npx oxfmt`, `npx oxlint`, or package-local binaries. For targeted checks, pass file paths through the npm script:
+  - `npm run lint -- packages/app/src/components/message.tsx`
+  - `npm run format:files -- CLAUDE.md packages/app/src/components/message.tsx`
+- **The protocol stays backward-compatible. Features don't have to.** Two separate contracts:
+  - **Protocol contract (always):** schema changes must not break parsing in either direction. An old client must still parse messages from a new daemon; a new daemon must still parse messages from an old client.
+    - New fields: `.optional()` with a sensible default or `.transform()` fallback.
+    - Never flip optional → required, remove fields, or narrow types (`string` → `enum`, `nullable` → non-null).
+    - Removed fields stay accepted (we stop sending them, not stop reading them).
+    - Test with: "does a 6-month-old client still parse this?" and "does a 6-month-old daemon still send something this client accepts?"
+  - **Feature contract (per-feature):** a new feature may require a new daemon capability. The client detects whether the capability is present and either runs the feature or shows "Update the host to use this." That's it.
+    - **No fallback paths.** Don't write a degraded version of a new feature that runs on old daemons. Don't fan out across legacy RPCs to simulate a missing capability. The user upgrades or doesn't get the feature.
+    - **No defensive branches scattered through the feature.** Capability detection happens in one place; downstream code reads a clean shape.
+    - **Capability flags live in `server_info.features.*`** with a single `// COMPAT(featureName): added in v0.1.X, drop the gate when floor >= v0.1.X` comment marking the cleanup site.
+    - Existing functionality keeps working across versions — that's the protocol contract doing its job. New-feature degradation is not the goal.
+
+- **All back-compat shims are tagged and dated for cleanup.** Every shim that exists for old-client/old-daemon support carries a `COMPAT(name)` comment with the version it was added in and a target removal date (typically 6 months out). One grep — `rg "COMPAT\("` — should produce the full list of cleanup work. Don't bury back-compat in untagged `??`-fallbacks or optional-chain tunnels — that's how it stops being deletable.
+
+## Platform gating
+
+The app runs on iOS, Android, web (browser), and web (Electron desktop). Code is cross-platform by default. Gate only when you must. Import gates from `@/constants/platform`.
+
+### The four gates
+
+| Gate                       | Type      | When to use                                                                                                                 |
+| -------------------------- | --------- | --------------------------------------------------------------------------------------------------------------------------- |
+| `isWeb`                    | constant  | DOM APIs — `document`, `window`, `<div>`, `addEventListener`, `ResizeObserver`. This is the **exception**, not the default. |
+| `isNative`                 | constant  | Native-only APIs — Haptics, `StatusBar.currentHeight`, push tokens, camera/scanner, `expo-av`.                              |
+| `getIsElectron()`          | cached fn | Desktop wrapper features — file dialogs, titlebar drag region, daemon management, app updates, dock badges.                 |
+| `useIsCompactFormFactor()` | hook      | Layout decisions — sidebar overlay vs pinned, modal vs full screen, single-panel vs split. From `@/constants/layout`.       |
+
+### Decision matrix
+
+| I need to...                                                   | Use                                                                       |
+| -------------------------------------------------------------- | ------------------------------------------------------------------------- |
+| Access DOM (`document`, `window`, `<div>`, `addEventListener`) | `if (isWeb)`                                                              |
+| Use a native-only API (Haptics, push tokens, camera)           | `if (isNative)`                                                           |
+| Use an Electron bridge (file dialog, titlebar, updates)        | `if (getIsElectron())`                                                    |
+| Switch layout between phone and tablet/desktop                 | `useIsCompactFormFactor()`                                                |
+| Show something on hover, always-visible on native              | `isHovered \|\| isNative \|\| isCompact` (hover only works on web)        |
+| Gate to iOS or Android specifically                            | `Platform.OS === "ios"` / `Platform.OS === "android"` (rare, keep inline) |
+
+### Rules
+
+- **Default is cross-platform.** Don't gate unless you have a specific reason.
+- **Prefer Metro file extensions over `if` statements.** When a module has fundamentally different implementations per platform, use `.web.ts` / `.native.ts` file extensions instead of runtime `if (isWeb)` branches. Metro resolves the correct file at build time — the unused platform code is never bundled. Reserve `if (isWeb)` for small, inline checks (a single line or a few props). If you find yourself writing a large `if (isWeb) { ... } else { ... }` block, split into separate files instead.
+  ```
+  hooks/
+    use-audio-recorder.web.ts    ← uses Web Audio API
+    use-audio-recorder.native.ts ← uses expo-audio
+  ```
+  Import as `@/hooks/use-audio-recorder` — Metro picks the right file automatically.
+- **Use `.electron.ts` / `.electron.tsx` for Electron-only web modules.** Electron is still the Metro `web` platform, but desktop dev/build sets `PASEO_WEB_PLATFORM=electron`, so Metro first looks for `.electron.*` files and falls back to normal `.web.*` files. Use this when the implementation depends on Electron-only behavior such as `webviewTag`, desktop preload APIs, or the Electron bridge. Keep plain browser web in `.web.*`, and keep native fallbacks in the base file or `.native.*`.
+  ```
+  components/
+    browser-pane.electron.tsx ← Electron <webview> implementation
+    browser-pane.web.tsx      ← plain web fallback
+    browser-pane.tsx          ← native fallback
+  ```
+  Import as `@/components/browser-pane` — Electron desktop gets the `.electron.tsx` file, browser web gets `.web.tsx`, and native gets the native/base implementation.
+- **NEVER use raw DOM APIs without `isWeb` guard.** DOM APIs crash native. Casting a RN ref to `HTMLElement` is a red flag — ensure the block is web-only.
+- **NEVER use `onPointerEnter`/`onPointerLeave`.** They don't fire on native iOS.
+- **Hover only works on web.** React Native's `onHoverIn`/`onHoverOut` on `Pressable` does NOT fire on native iOS/iPad — the underlying W3C pointer events are behind disabled experimental flags. For hover-to-show UI (kebab menus, action buttons), use `isHovered || isNative || isCompact` so the controls are always visible on native and hover-to-show on web.
+- **Don't use Platform.OS as a proxy for layout capabilities.** Use breakpoints for layout decisions, not platform checks.
+- **Import `isWeb`/`isNative` from `@/constants/platform`.** Never write `const isWeb = Platform.OS === "web"` locally.
 
 ## Debugging
 
-
 Find the complete daemon logs and traces in the $PASEO_HOME/daemon.log

CONTRIBUTING.md

@@ -0,0 +1,168 @@
+# Contributing to Paseo
+
+Thanks for taking the time to contribute.
+
+## How this project works
+
+Paseo is a BDFL project. Product direction, scope, and what ships are the maintainer's call.
+
+This means:
+
+- PRs submitted without prior discussion will likely be rejected, heavily modified, or scoped down.
+- The maintainer may rewrite, split, cherry-pick from, or close any PR at their discretion.
+- There is no obligation to merge a PR as-submitted, regardless of code quality.
+
+This is not meant to discourage contributions. It is meant to set clear expectations so nobody wastes their time.
+
+## How to contribute
+
+1. **Open an issue first.** Describe the problem or improvement. Get a thumbs up before writing code.
+2. **Keep it small.** One bug, one flow, one focused change.
+3. **Open a PR** once there is alignment on scope.
+
+If you want to propose a direction change, start a conversation.
+
+## Before you start
+
+Please read these first:
+
+- [README.md](README.md)
+- [docs/architecture.md](docs/architecture.md)
+- [docs/development.md](docs/development.md)
+- [docs/coding-standards.md](docs/coding-standards.md)
+- [docs/testing.md](docs/testing.md)
+- [CLAUDE.md](CLAUDE.md)
+
+## What is most helpful
+
+The most useful contributions right now are:
+
+- bug fixes
+- windows and linux specific fixes
+- regression fixes
+- doc improvements
+- packaging / platform fixes
+- focused UX improvements that fit the existing product direction
+- tests that lock down important behavior
+
+## Scope expectations
+
+Please keep PRs narrow.
+
+Good:
+
+- fix one bug
+- improve one flow
+- add one focused panel or command
+- tighten one piece of UI
+
+Bad:
+
+- combine multiple product ideas in one PR
+- bundle unrelated refactors with a feature
+- sneak in roadmap decisions
+
+If a contribution contains multiple ideas, split it up.
+
+## Product fit matters
+
+Paseo is an opinionated product.
+
+When reviewing contributions, the bar is not just:
+
+- is this useful?
+- is this well implemented?
+
+It is also:
+
+- does this fit Paseo?
+- does this add product surface that will be hard to maintain?
+- does the value justify the maintenance surface it adds?
+- does this solve a common need or over-serve an edge case?
+- does this preserve the product's current direction?
+
+## Development setup
+
+### Prerequisites
+
+- Node.js matching `.tool-versions`
+- npm workspaces
+
+### Start local development
+
+```bash
+# runs both daemon and expo app
+npm run dev
+```
+
+Useful commands:
+
+```bash
+npm run dev:server
+npm run dev:app
+npm run dev:desktop
+npm run dev:website
+npm run cli -- ls -a -g
+```
+
+Read [docs/development.md](docs/development.md) for build-sync gotchas, local state, ports, and daemon details.
+
+## Multi-platform testing
+
+Paseo ships to mobile (iOS/Android), web, and desktop (Electron). Every UI change must be tested on mobile and web at minimum, and desktop if relevant. Things that look fine on one surface regularly break on another.
+
+Common checks:
+
+```bash
+npm run typecheck
+npm run test --workspaces --if-present
+```
+
+Important rules:
+
+- always run `npm run typecheck` after changes
+- tests should be deterministic
+- prefer real dependencies over mocks when possible
+- do not make breaking WebSocket / protocol changes
+- app and daemon versions in the wild lag each other, so compatibility matters
+
+If you touch protocol or shared client/server behavior, read the compatibility notes in [CLAUDE.md](CLAUDE.md).
+
+## Coding standards
+
+Paseo has explicit standards. Follow them.
+
+The full guide lives in [docs/coding-standards.md](docs/coding-standards.md).
+
+## PR checklist
+
+Before opening a PR, make sure:
+
+- there was prior discussion and alignment on scope (issue or conversation)
+- the change is focused, one idea per PR
+- the PR description explains what changed and why
+- **UI changes include screenshots or videos** for every affected platform (mobile, web, desktop)
+- UI changes have been tested on mobile and web at minimum
+- typecheck passes
+- tests pass, or you clearly explain what could not be run
+- relevant docs were updated if needed
+
+## Communication
+
+If you are unsure whether something fits, ask first.
+
+That is especially true for:
+
+- new core UX
+- naming / terminology changes
+- new extension points
+- new orchestration models
+- anything that would be hard to remove later
+
+Early alignment saves everyone time.
+
+## Forks are fine
+
+If you want to explore a different product direction, a fork is completely fine.
+
+Paseo is open source on purpose. Not every idea needs to land in the main repo to be valuable.

README.md

@@ -4,6 +4,21 @@
 
 <h1 align="center">Paseo</h1>
 
+<p align="center">
+  <a href="https://github.com/getpaseo/paseo/stargazers">
+    <img src="https://img.shields.io/github/stars/getpaseo/paseo?style=flat&logo=github" alt="GitHub stars">
+  </a>
+  <a href="https://github.com/getpaseo/paseo/releases">
+    <img src="https://img.shields.io/github/v/release/getpaseo/paseo?style=flat&logo=github" alt="GitHub release">
+  </a>
+  <a href="https://x.com/moboudra">
+    <img src="https://img.shields.io/badge/%40moboudra-555?logo=x" alt="X">
+  </a>
+  <a href="https://discord.gg/jz8T2uahpH">
+    <img src="https://img.shields.io/badge/Discord-555?logo=discord" alt="Discord">
+  </a>
+</p>
+
 <p align="center">One interface for all your Claude Code, Codex and OpenCode agents.</p>
 
 <p align="center">
@@ -54,6 +69,7 @@ paseo
 This shows a QR code in the terminal. Connect from any client. This path is useful for servers and remote machines.
 
 For full setup and configuration, see:
+
 - [Docs](https://paseo.sh/docs)
 - [Configuration reference](https://paseo.sh/docs/configuration)
 
@@ -75,9 +91,9 @@ paseo --host workstation.local:6767 run "run the full test suite"
 
 See the [full CLI reference](https://paseo.sh/docs/cli) for more.
 
-## Orchestration skills (Unstable)
+## Skills
 
-Experimental skills that teach agents how to use the Paseo CLI to orchestrate other agents. I am updating these very frequently as I learn new things, expect changes without notice, might be coupled to my own setup, use at your own risk.
+Skills teach your agent to use Paseo to orchestrate other agents.
 
 ```bash
 npx skills add getpaseo/paseo
@@ -85,22 +101,15 @@ npx skills add getpaseo/paseo
 
 Then use them in any agent conversation:
 
-```bash
-# Use handoff when you discuss something with an agent but want another one to implement.
-# I use this to plan with Claude and then handoff to Codex to implement.
-/paseo-handoff hand off the authentication fix to codex 5.4 in a worktree
-
-# Use loops when you have clear acceptance criteria (aka Ralph loops).
-/paseo-loop loop a codex agent to fix the backend tests, use sonnet to verify, max 10 iterations
-
-# Orchestrator teaches the agent how to create teams and manage them via a chat room.
-# Very opinionated and expects both Codex and Claude to work.
-/paseo-orchestrator spin up a team to implement the database refactor, use chat to coordinate. use claude to plan and codex to implement and review
-```
+- `/paseo-handoff` — hand off work between agents. I use this to plan with Claude and then handoff to Codex to implement.
+- `/paseo-loop` — loop an agent against clear acceptance criteria (aka Ralph loops), optionally with a verifier.
+- `/paseo-advisor` — spin up a single agent as an advisor for a second opinion, without delegating the work itself.
+- `/paseo-committee` — form a committee of two contrasting agents to step back, do root cause analysis, and produce a plan.
 
 ## Development
 
 Quick monorepo package map:
+
 - `packages/server`: Paseo daemon (agent process orchestration, WebSocket API, MCP server)
 - `packages/app`: Expo client (iOS, Android, web)
 - `packages/cli`: `paseo` CLI for daemon and agent workflows
@@ -127,6 +136,68 @@ npm run build:daemon
 npm run typecheck
 ```
 
+## Community
+
+- [paseo-relay](https://github.com/zenghongtu/paseo-relay) — self-hosted relay in Go
+
+### Self-hosted relay TLS
+
+Self-hosted relays use `ws://` unless TLS is opted in. For a relay behind nginx on 443, start the daemon with:
+
+```bash
+PASEO_RELAY_ENDPOINT=127.0.0.1:8080 \
+PASEO_RELAY_PUBLIC_ENDPOINT=relay.example.com:443 \
+PASEO_RELAY_USE_TLS=true \
+paseo daemon start
+```
+
+Equivalent config:
+
+```json
+{
+  "daemon": {
+    "relay": {
+      "enabled": true,
+      "endpoint": "127.0.0.1:8080",
+      "publicEndpoint": "relay.example.com:443",
+      "useTls": true
+    }
+  }
+}
+```
+
+Minimal nginx WebSocket proxy:
+
+```nginx
+server {
+  listen 443 ssl;
+  server_name relay.example.com;
+
+  ssl_certificate /etc/letsencrypt/live/relay.example.com/fullchain.pem;
+  ssl_certificate_key /etc/letsencrypt/live/relay.example.com/privkey.pem;
+
+  location /ws {
+    proxy_pass http://127.0.0.1:8080;
+    proxy_http_version 1.1;
+    proxy_set_header Upgrade $http_upgrade;
+    proxy_set_header Connection "upgrade";
+    proxy_set_header Host $host;
+  }
+}
+```
+
+---
+
+<p align="center">
+  <a href="https://star-history.com/#getpaseo/paseo&Date">
+    <picture>
+      <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=getpaseo/paseo&type=Date&theme=dark">
+      <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=getpaseo/paseo&type=Date">
+      <img src="https://api.star-history.com/svg?repos=getpaseo/paseo&type=Date" alt="Star history chart for getpaseo/paseo" width="600" style="max-width: 100%;">
+    </picture>
+  </a>
+</p>
+
 ## License
 
 AGPL-3.0

SECURITY.md

@@ -19,31 +19,44 @@ The relay is designed to be untrusted. All traffic between your phone and daemon
 
 ### How it works
 
-1. The daemon generates a persistent ECDH keypair and stores it locally
-2. When you scan the QR code or click the pairing link, your phone receives the daemon's public key
-3. Your phone sends a handshake message with its own public key. The daemon will not accept any commands until this handshake completes.
-4. Both sides perform an ECDH key exchange to derive a shared secret. All subsequent messages are encrypted with AES-256-GCM.
+1. The daemon generates a persistent Curve25519 keypair on first run and stores it at `$PASEO_HOME/daemon-keypair.json` with mode `0600`
+2. The pairing URL (rendered as a QR code or opened directly) carries the daemon's public key in its URL fragment (`https://app.paseo.sh/#offer=...`). Fragments are not sent to the web server, so `app.paseo.sh` never sees the key.
+3. When the phone connects via the relay, it generates a fresh ephemeral Curve25519 keypair and sends an `e2ee_hello` message containing its public key. The daemon will not process any application messages until this handshake completes.
+4. Both sides perform a Curve25519 ECDH key exchange to derive a shared key. All subsequent messages are encrypted with XSalsa20-Poly1305 (NaCl `box`). The wire format is `[24-byte nonce][ciphertext]`, base64-encoded as a WebSocket text frame.
 
-The relay sees only: IP addresses, timing, message sizes, and session IDs. It cannot read message contents, forge messages, or derive encryption keys from observing the handshake.
+The relay sees only: IP addresses, timing, message sizes, session IDs, and the plaintext `e2ee_hello` / `e2ee_ready` handshake frames (which contain only public keys). It cannot read message contents, forge messages, or derive encryption keys from observing the handshake.
 
 ### Why the relay can't attack you
 
 The daemon requires a valid cryptographic handshake before processing any commands. A compromised relay cannot:
 
-- **Send commands** — Without your phone's private key, it cannot complete the handshake
-- **Read your traffic** — All messages are encrypted with AES-256-GCM after the handshake
-- **Forge messages** — GCM provides authenticated encryption; tampered messages are rejected
-- **Replay old messages** — Each session derives fresh encryption keys
+- **Impersonate the daemon to your phone** — Without the daemon's secret key, it cannot derive the shared key, so any traffic it injects fails authenticated decryption on the phone
+- **Send commands as you** — The daemon only accepts traffic that decrypts and authenticates under a shared key derived with its own secret key. The phone's keypair is ephemeral per connection, so there is no persistent phone-side secret to steal; protection comes from the daemon's secret key never leaving the daemon.
+- **Read your traffic** — All messages are encrypted with XSalsa20-Poly1305 (NaCl box) after the handshake
+- **Forge messages** — NaCl box provides authenticated encryption; tampered messages are rejected
+- **Replay old messages across sessions** — Each session derives fresh encryption keys, so ciphertext from one session cannot be replayed into another session. Within a live session, replay protection is not yet implemented; the protocol uses random nonces and does not track nonce reuse or message counters.
 
 ### Trust model
 
 The QR code or pairing link is the trust anchor. It contains the daemon's public key, which is required to establish the encrypted connection. Treat it like a password — don't share it publicly.
 
+## Local daemon trust boundary
+
+By default, the daemon binds to `127.0.0.1`. With no password configured, the local control plane is trusted by network reachability — anything that can reach the daemon socket can control the daemon. This is the same security model Docker documents for its daemon: the security boundary is access to the socket or listening address.
+
+The daemon also supports an optional shared-secret password (set via `auth.password` in `config.json` or the `PASEO_PASSWORD` env var; stored bcrypt-hashed). When configured, every HTTP request must carry `Authorization: Bearer <password>` and every WebSocket upgrade must include a `Sec-WebSocket-Protocol: paseo.bearer.<password>` subprotocol. Browser WebSocket cannot set custom headers, which is why the token rides in the subprotocol. Health (`GET /api/health`) and CORS preflight (`OPTIONS`) are exempt. The password is intended for direct-TCP exposure (e.g. `tcp://host:port?ssl=true&password=...`); it is **not** a substitute for the relay's E2E encryption when traversing untrusted networks.
+
+If you expose the daemon beyond loopback, such as by binding to `0.0.0.0`, forwarding it through a tunnel or reverse proxy, or publishing it from a Docker container, you are responsible for restricting and securing that access. Setting a password is strongly recommended in that case.
+
+For remote access, use the relay connection. It is the supported path for reaching the daemon off-machine, and it adds end-to-end encryption plus a pairing handshake before commands are accepted.
+
+Host header validation and CORS origin checks are defense-in-depth controls for localhost exposure. They help block DNS rebinding and browser-based attacks, but they do not replace network isolation.
+
 ## DNS rebinding protection
 
 CORS is not a complete security boundary. It controls which browser origins can make requests, but does not prevent a malicious website from resolving its domain to your local machine (DNS rebinding).
 
-Paseo uses a host allowlist to validate the `Host` header on incoming requests. Requests with unrecognized hosts are rejected.
+Paseo validates the `Host` header on every HTTP request and every WebSocket upgrade against an allowlist (Vite-style semantics). By default, only `localhost`, `*.localhost`, and any literal IP address (IPv4 or IPv6) are accepted. Additional hostnames can be configured via `hostnames` in `config.json` or the `PASEO_HOSTNAMES` env var (comma-separated; entries beginning with `.` match a domain and its subdomains; the value `true` disables the allowlist entirely). Requests with unrecognized hosts are rejected with `403 Host not allowed`.
 
 ## Agent authentication
 

app.json

@@ -1,5 +0,0 @@
-{
-  "android": {
-    "package": "com.moboudra.paseo"
-  }
-}

biome.json

@@ -1,32 +0,0 @@
-{
-  "$schema": "https://biomejs.dev/schemas/2.0.0/schema.json",
-  "vcs": {
-    "enabled": true,
-    "clientKind": "git",
-    "useIgnoreFile": true
-  },
-  "files": {
-    "includes": ["**", "!*.lock"]
-  },
-  "formatter": {
-    "indentStyle": "space",
-    "indentWidth": 2,
-    "lineWidth": 100
-  },
-  "javascript": {
-    "formatter": {
-      "quoteStyle": "double",
-      "trailingCommas": "all",
-      "semicolons": "always"
-    }
-  },
-  "css": {
-    "parser": {
-      "cssModules": true,
-      "tailwindDirectives": true
-    }
-  },
-  "linter": {
-    "enabled": false
-  }
-}

docs/ARCHITECTURE.md

@@ -1,192 +0,0 @@
-# Architecture
-
-Paseo is a client-server system for monitoring and controlling local AI coding agents. The daemon runs on your machine, manages agent processes, and streams their output in real time over WebSocket. Clients (mobile app, CLI, desktop app) connect to the daemon to observe and interact with agents.
-
-Your code never leaves your machine. Paseo is local-first.
-
-## System overview
-
-```
-┌─────────────┐    ┌─────────────┐    ┌─────────────┐
-│  Mobile App  │    │     CLI     │    │ Desktop App │
-│   (Expo)     │    │ (Commander) │    │ (Electron)  │
-└──────┬───────┘    └──────┬──────┘    └──────┬──────┘
-       │                   │                  │
-       │    WebSocket      │    WebSocket     │    Managed subprocess
-       │    (direct or     │    (direct)      │    + WebSocket
-       │     via relay)    │                  │
-       └───────────┬───────┴──────────────────┘
-                   │
-            ┌──────▼──────┐
-            │   Daemon    │
-            │  (Node.js)  │
-            └──────┬──────┘
-                   │
-      ┌────────────┼────────────┐
-      │            │            │
-┌─────▼─────┐ ┌───▼────┐ ┌────▼─────┐
-│  Claude   │ │ Codex  │ │ OpenCode │
-│  Agent    │ │ Agent  │ │  Agent   │
-│  SDK      │ │ Server │ │          │
-└───────────┘ └────────┘ └──────────┘
-```
-
-## Components at a glance
-
-- **Daemon:** Local server that spawns and manages agent processes and exposes the WebSocket API.
-- **App:** Cross-platform Expo client for iOS, Android, web, and the shared UI used by desktop.
-- **CLI:** Terminal interface for agent workflows that can also start and manage the daemon.
-- **Desktop app:** Electron wrapper around the web app that bundles and auto-manages its own daemon.
-- **Relay:** Optional encrypted bridge for remote access without opening ports directly.
-
-## Packages
-
-### `packages/server` — The daemon
-
-The heart of Paseo. A Node.js process that:
-
-- Listens for WebSocket connections from clients
-- Manages agent lifecycle (create, run, stop, resume, archive)
-- Streams agent output in real time via a timeline model
-- Exposes an MCP server for agent-to-agent control
-- Optionally connects outbound to a relay for remote access
-
-**Key modules:**
-
-| Module | Responsibility |
-|---|---|
-| `bootstrap.ts` | Daemon initialization: HTTP server, WS server, agent manager, storage, relay |
-| `websocket-server.ts` | WebSocket connection management, hello/welcome handshake, binary multiplexing |
-| `session.ts` | Per-client session state, timeline subscriptions, terminal operations |
-| `agent/agent-manager.ts` | Agent lifecycle state machine, timeline tracking, subscriber management |
-| `agent/agent-storage.ts` | File-backed JSON persistence at `$PASEO_HOME/agents/` |
-| `agent/mcp-server.ts` | MCP server for sub-agent creation, permissions, timeouts |
-| `providers/` | Provider adapters: Claude (Agent SDK), Codex (AppServer), OpenCode |
-| `relay-transport.ts` | Outbound relay connection with E2E encryption |
-| `client/daemon-client.ts` | Client library for connecting to the daemon (used by CLI and app) |
-
-### `packages/app` — Mobile + web client (Expo)
-
-Cross-platform React Native app that connects to one or more daemons.
-
-- Expo Router navigation (`/h/[serverId]/agents`, etc.)
-- `DaemonRegistryContext` manages saved daemon connections
-- `SessionContext` wraps the daemon client for the active session
-- `Stream` model handles timeline with compaction, gap detection, sequence-based deduplication
-- Voice features: dictation (STT) and voice agent (realtime)
-
-### `packages/cli` — Command-line client
-
-Commander.js CLI with Docker-style commands:
-
-- `paseo agent ls/run/stop/logs/inspect/wait/send/attach`
-- `paseo daemon start/stop/restart/status/pair`
-- `paseo permit allow/deny/ls`
-- `paseo provider ls/models`
-- `paseo worktree ls/archive`
-
-Communicates with the daemon via the same WebSocket protocol as the app.
-
-### `packages/relay` — E2E encrypted relay
-
-Enables remote access when the daemon is behind a firewall.
-
-- ECDH key exchange + AES-256-GCM encryption
-- Relay server is zero-knowledge — it routes encrypted bytes, cannot read content
-- Client and daemon channels with identical API (`createClientChannel`, `createDaemonChannel`)
-- Pairing via QR code transfers the daemon's public key to the client
-
-See [SECURITY.md](../SECURITY.md) for the full threat model.
-
-### `packages/desktop` — Desktop app (Electron)
-
-Electron wrapper for macOS, Linux, and Windows.
-
-- Can spawn the daemon as a managed subprocess
-- Native file access for workspace integration
-- Same WebSocket client as mobile app
-
-### `packages/website` — Marketing site
-
-TanStack Router + Cloudflare Workers. Serves paseo.sh.
-
-## WebSocket protocol
-
-All clients speak the same binary-multiplexed WebSocket protocol.
-
-**Handshake:**
-
-```
-Client → Server:  WSHelloMessage { id, clientId, version, timestamp }
-Server → Client:  WSWelcomeMessage { clientId, daemonVersion, sessionId, capabilities }
-```
-
-**Message types:**
-
-- `agent_update` — Agent state changed (status, title, labels)
-- `agent_stream` — New timeline event from a running agent
-- `workspace_update` — Workspace state changed
-- `agent_permission_request` — Agent needs user approval for a tool call
-- Command-response pairs for fetch, list, create, etc.
-
-**Binary multiplexing:**
-
-Terminal I/O and agent streaming share the same connection via `BinaryMuxFrame`:
-- Channel 0: control messages
-- Channel 1: terminal data
-- 1-byte channel ID + 1-byte flags + variable payload
-
-## Agent lifecycle
-
-```
-initializing → idle → running → idle (or error → closed)
-                 ↑        │
-                 └────────┘  (agent completes a turn, awaits next prompt)
-```
-
-- **AgentManager** tracks up to 200 timeline items per agent
-- Timeline is append-only with epochs (each run starts a new epoch)
-- Events stream to all subscribed clients in real time
-- Agent state persists to `$PASEO_HOME/agents/{cwd-with-dashes}/{agent-id}.json`
-
-## Agent providers
-
-Each provider implements a common `AgentClient` interface:
-
-| Provider | Wraps | Session format |
-|---|---|---|
-| Claude | Anthropic Agent SDK | `~/.claude/projects/{cwd}/{session-id}.jsonl` |
-| Codex | CodexAppServer | `~/.codex/sessions/{date}/rollout-{ts}-{id}.jsonl` |
-| OpenCode | OpenCode CLI | Provider-managed |
-
-All providers:
-- Handle their own authentication (Paseo does not manage API keys)
-- Support session resume via persistence handles
-- Map tool calls to a normalized `ToolCallDetail` type
-- Expose provider-specific modes (plan, default, full-access)
-
-## Data flow: running an agent
-
-1. Client sends `CreateAgentRequestMessage` with config (prompt, cwd, provider, model, mode)
-2. Session routes to `AgentManager.create()`
-3. AgentManager creates a `ManagedAgent`, initializes provider session
-4. Provider runs the agent → emits `AgentStreamEvent` items
-5. Events append to the agent timeline, broadcast to all subscribed clients
-6. Tool calls are normalized to `ToolCallDetail` (shell, read, edit, write, search, etc.)
-7. Permission requests flow: agent → server → client → user decision → server → agent
-
-## Storage
-
-```
-$PASEO_HOME/
-├── agents/{cwd-with-dashes}/{agent-id}.json   # Agent state + config
-├── projects/projects.json                      # Project registry
-├── projects/workspaces.json                    # Workspace registry
-└── daemon.log                                  # Daemon trace logs
-```
-
-## Deployment models
-
-1. **Local daemon** (default): `paseo daemon start` on `127.0.0.1:6767`
-2. **Managed desktop**: Electron app spawns daemon as subprocess
-3. **Remote + relay**: Daemon behind firewall, relay bridges with E2E encryption

docs/CODING_STANDARDS.md

@@ -1,174 +0,0 @@
-# Coding Standards
-
-These standards apply to all code changes: features, bug fixes, refactors, and performance work.
-
-## Core principles
-
-- **Zero complexity budget** — justify every abstraction with specific benefits
-- **Fully typed TypeScript** — no `any`, no untyped boundaries
-- **YAGNI** — build features and abstractions only when needed
-- **Functional and declarative** over object-oriented
-- **`interface`** over `type` when possible
-- **`function` declarations** over arrow function assignments
-- **Single-purpose functions** — one function, one job
-- **Design for edge cases through types** rather than explicit handling
-- **Don't catch errors** unless there's a strong reason to
-- **No index.ts barrel files** that only re-export — they create unnecessary indirection
-- **No "while I'm at it" improvements** — stay focused on the task
-
-## Type hygiene
-
-### Infer from schemas
-
-Never hand-write a TypeScript type that can be inferred from a Zod schema.
-
-```typescript
-// Bad: duplicate type that can drift
-const schema = z.object({ procedure: z.string(), args: z.record(z.unknown()) });
-type RPCArgs = { procedure: string; args: Record<string, unknown> };
-
-// Good: infer from schema
-type RPCArgs = z.infer<typeof schema>;
-```
-
-### Named types over inline
-
-No complex inline types in public function signatures.
-
-```typescript
-// Bad
-function enqueueJob(input: { userId: string; priority: "low" | "normal" | "high" }) {}
-
-// Good
-interface EnqueueJobInput { userId: string; priority: "low" | "normal" | "high" }
-function enqueueJob(input: EnqueueJobInput) {}
-```
-
-### Object parameters
-
-If a function needs more than one argument, use a single object parameter.
-
-```typescript
-// Bad: positional args
-function createToolCall(provider: string, toolName: string, payload: unknown) {}
-
-// Good: object param
-interface CreateToolCallInput { provider: string; toolName: string; payload: unknown }
-function createToolCall(input: CreateToolCallInput) {}
-```
-
-### One canonical type per concept
-
-Don't redefine the same concept in different layer-specific shapes (`RpcX`, `DbX`, `UiX`). Keep one canonical type and add explicit layer wrappers that reference it.
-
-```typescript
-// Bad: duplicated fields across layers
-type RpcToolCall = { toolName: string; args: Record<string, unknown>; requestId: string };
-type DbToolCall = { toolName: string; args: Record<string, unknown>; id: string; createdAt: Date };
-
-// Good: canonical type + wrappers
-type ToolCall = { toolName: string; args: Record<string, unknown> };
-type ToolCallRequest = { requestId: string; toolCall: ToolCall };
-type ToolCallRecord = { id: string; createdAt: Date; toolCall: ToolCall };
-```
-
-## Make impossible states impossible
-
-Use discriminated unions instead of bags of booleans and optionals.
-
-```typescript
-// Bad
-interface FetchState { isLoading: boolean; error?: Error; data?: Data }
-
-// Good
-type FetchState =
-  | { status: "idle" }
-  | { status: "loading" }
-  | { status: "error"; error: Error }
-  | { status: "success"; data: Data };
-```
-
-## Optionality is a design decision
-
-Don't mark fields optional to avoid migrations. Decide deliberately:
-
-1. Is optionality actually needed?
-2. If there are distinct valid states → discriminated union
-3. If value can be intentionally empty → explicit `null`
-4. Keep optionality at real boundaries (external input), then resolve it
-
-## Validate at boundaries, trust internally
-
-Parse external data once at the boundary with schema validation. Then use typed values everywhere else.
-
-```typescript
-// Bad: optional chaining because shape is unclear
-const value = response?.data?.items?.[0]?.name;
-
-// Good: validate at boundary, trust the types
-const parsed = responseSchema.parse(rawResponse);
-const value = parsed.data.items[0].name;
-```
-
-## Error handling
-
-- **Fail explicitly** — if caller requests X and X is unavailable, throw rather than silently returning Y
-- **Use typed domain errors** — not plain `Error`. Carry structured metadata for handling, logging, and user messaging
-- **Preserve error semantics** — don't collapse meaningful typed errors into generic `Error`
-
-```typescript
-class TimeoutError extends Error {
-  constructor(
-    public readonly operation: string,
-    public readonly waitedMs: number,
-  ) {
-    super(`${operation} timed out after ${waitedMs}ms`);
-    this.name = "TimeoutError";
-  }
-}
-```
-
-## Keep logic density low
-
-Avoid packing branching, lookup, and transformation into single dense expressions.
-
-```typescript
-// Bad: nested ternaries + inline lookups
-const billing = shouldUseLegacy(account)
-  ? getLegacy(account)
-  : buildBilling(account, rates.find((r) => r.region === account.region));
-
-// Good: named steps, then assemble
-const rate = rates.find((r) => r.region === account.region);
-if (!rate) throw new MissingRateError(account.region);
-const billing = shouldUseLegacy(account) ? getLegacy(account) : buildBilling(account, rate);
-```
-
-## Centralize policy
-
-When the same discriminator (`plan`, `provider`, `kind`, `status`) is checked across multiple files, centralize it into a policy model. A new case should require editing one place, not many.
-
-## React: keep components dumb
-
-- Components render state and dispatch events — they don't compute transitions
-- If a component has more than two interacting `useState` calls, extract a state machine or reducer
-- `useRef` for mutable coordination state (flags, timers) is a smell — model states explicitly
-- Never mirror a source of truth into local state; derive from it
-- Test state logic as pure functions without rendering
-
-## File organization
-
-- Organize by domain first (`providers/claude/`), not by technical type (`tool-parsers/`)
-- Name files after the main export (`create-toolcall.ts`)
-- Use `index.ts` as an entrypoint, not a dumping ground
-- Collocate tests with implementation (`thing.ts` + `thing.test.ts`)
-
-## Refactoring contract
-
-Refactoring is structure work, not feature work.
-
-- Preserve behavior by default, especially user-facing behavior
-- Do not remove features to simplify code without explicit approval
-- Have a verification strategy before you start
-- Fully migrate callers and remove old paths in the same refactor
-- No fallback behavior by default — prefer explicit error over silent degradation

docs/DESIGN.md

@@ -1,73 +0,0 @@
-# Designing Features
-
-How to think through a feature before writing code.
-
-## Start from the user
-
-Even for backend work, start from the user's perspective:
-
-- What problem does this solve?
-- What triggers it? User action, schedule, event?
-- What does success look like from the user's perspective?
-- What data does it need? Where does that data come from?
-
-## Map existing code
-
-Before designing anything new, understand what exists:
-
-- Where does similar functionality live?
-- What patterns does the codebase already use?
-- What layers exist? (See [ARCHITECTURE.md](./ARCHITECTURE.md))
-- What types and data shapes are already defined?
-
-New features rarely mean only new code. Usually they require modifying existing interfaces, extending existing types, or refactoring to accommodate the new functionality. Identify what needs to change, not just what needs to be added.
-
-## Define verification before implementation
-
-Before designing the solution, define how you'll know it works:
-
-- What tests will prove this feature is correct?
-- At what layer? Unit, integration, E2E?
-- What's the simplest way to verify the core behavior?
-
-If you can't define verification, you don't understand the feature well enough yet.
-
-## Design the shape
-
-### Data
-
-- What types are needed?
-- Use discriminated unions — make impossible states impossible
-- One canonical type per concept (see [CODING_STANDARDS.md](./CODING_STANDARDS.md))
-
-### Layers
-
-- What belongs in each layer?
-- Where are the boundaries?
-- What does each layer expose to the layer above?
-
-### Interactions
-
-- How does data flow through the system?
-- What triggers what?
-- Where do side effects happen?
-
-### Refactoring
-
-- What existing code needs to change?
-- Is existing code testable enough? If not, that's part of the plan.
-
-## Create a concrete plan
-
-Once the design is clear:
-
-1. **Acceptance criteria** — specific, verifiable outcomes (not "should work well" but "returns X when given Y")
-2. **Ordered steps** — what to build first (usually: types, then lowest layer, then up)
-3. **What to refactor** before adding new code
-4. **How to verify** each step
-
-## Principles
-
-- **Fit, don't force** — new code should fit existing patterns, or refactor first
-- **Simple** — the best design is the simplest one that works
-- **Verify early** — define how to test before designing the implementation

docs/DEVELOPMENT.md

@@ -1,130 +0,0 @@
-# Development
-
-## Prerequisites
-
-- Node.js (see `.tool-versions` for exact version)
-- npm workspaces (comes with Node)
-
-## Running the dev server
-
-```bash
-npm run dev
-```
-
-The dev script automatically picks an available port. Both the server and Expo app run in a Tmux session — see `CLAUDE.local.md` for system-specific session details.
-
-### Running alongside the main checkout
-
-Set `PASEO_HOME` to isolate state when running a second instance (e.g., in a worktree):
-
-```bash
-PASEO_HOME=~/.paseo-blue npm run dev
-```
-
-- `PASEO_HOME` — path for runtime state (agents, sockets, etc.). Defaults to `~/.paseo`.
-
-### Default ports
-
-In the main checkout:
-- Daemon: `localhost:6767`
-- Expo app: `localhost:8081`
-
-In worktrees or with `npm run dev`, ports may differ. Never assume defaults.
-
-### Daemon logs
-
-Check `$PASEO_HOME/daemon.log` for trace-level logs.
-
-## Build sync gotchas
-
-### Relay → Daemon
-
-When changing `packages/relay/src/*`, rebuild before running the daemon:
-
-```bash
-npm run build --workspace=@getpaseo/relay
-```
-
-The Node daemon imports `@getpaseo/relay` from `packages/relay/dist/*`, not `src/*`.
-
-### Server → CLI
-
-When changing `packages/server/src/client/*` (especially `daemon-client.ts`) or shared WS protocol types, rebuild before running CLI commands:
-
-```bash
-npm run build --workspace=@getpaseo/server
-```
-
-The CLI imports `@getpaseo/server` via package exports resolving to `dist/*`. Stale `dist` means the CLI speaks an old protocol and fails with handshake warnings or timeouts.
-
-## CLI reference
-
-Use `npm run cli` to run the local CLI (instead of the globally installed `paseo` which points to the main checkout).
-
-```bash
-npm run cli -- ls -a -g              # List all agents globally
-npm run cli -- ls -a -g --json       # Same, as JSON
-npm run cli -- inspect <id>          # Show detailed agent info
-npm run cli -- logs <id>             # View agent timeline
-npm run cli -- daemon status         # Check daemon status
-```
-
-Use `--host <host:port>` to point the CLI at a different daemon:
-
-```bash
-npm run cli -- --host localhost:7777 ls -a
-```
-
-## Agent state
-
-Agent data lives at:
-
-```
-$PASEO_HOME/agents/{cwd-with-dashes}/{agent-id}.json
-```
-
-Find an agent by ID:
-```bash
-find $PASEO_HOME/agents -name "{agent-id}.json"
-```
-
-Find by content:
-```bash
-rg -l "some title text" $PASEO_HOME/agents/
-```
-
-## Provider session files
-
-Get the session ID from the agent JSON (`persistence.sessionId`), then:
-
-**Claude:**
-```
-~/.claude/projects/{cwd-with-dashes}/{session-id}.jsonl
-```
-
-**Codex:**
-```
-~/.codex/sessions/{YYYY}/{MM}/{DD}/rollout-{timestamp}-{session-id}.jsonl
-```
-
-## Testing with Playwright MCP
-
-Use Playwright MCP connecting to Metro at `http://localhost:8081` for UI testing.
-
-Do NOT use browser history (back/forward). Always navigate by clicking UI elements or using `browser_navigate` with the full URL — the app uses client-side routing and browser history breaks state.
-
-## Expo troubleshooting
-
-```bash
-npx expo-doctor
-```
-
-Diagnoses version mismatches and native module issues.
-
-## Typecheck
-
-Always run typecheck after changes:
-
-```bash
-npm run typecheck
-```

docs/RELEASE.md

@@ -1,86 +0,0 @@
-# Release
-
-All workspaces share one version and release together.
-
-## Standard release (patch)
-
-```bash
-npm run release:patch
-```
-
-This bumps the version across all workspaces, runs checks, publishes to npm, and pushes the branch + tag (triggering desktop, APK, and EAS mobile workflows).
-
-If asked to "release paseo" without specifying major/minor, treat it as a patch release.
-
-## Manual step-by-step
-
-```bash
-npm run release:check        # Typecheck, build, dry-run pack
-npm run version:all:patch    # Bump version, create commit + tag
-npm run release:publish      # Publish to npm
-npm run release:push         # Push HEAD + tag (triggers CI workflows)
-```
-
-## Draft release flow
-
-```bash
-npm run draft-release:patch    # Bump, push tag, create draft GitHub Release
-# ... test builds from the draft release assets ...
-npm run release:finalize       # Publish npm, promote draft to published
-```
-
-- `draft-release:patch` creates the GitHub Release as a draft so desktop assets, APK uploads, and synced notes attach to it
-- `release:finalize` publishes npm and promotes the same draft release
-- Use the same semver tag for both; don't cut a second tag
-- Desktop assets now come from the Electron package at `packages/desktop`
-- **Do NOT create a changelog entry for drafts.** The changelog entry is written only when finalizing. The website parses `CHANGELOG.md` to determine the latest published version for download links — adding an entry for a draft will point the homepage at untested assets.
-
-## Fixing a failed release build
-
-**NEVER bump the version to fix a build problem.** New versions are reserved for meaningful product changes (features, fixes, improvements). Build/CI failures are fixed on the current version.
-
-**NEVER use `workflow_dispatch` to retry release builds.** The `workflow_dispatch` trigger runs the workflow file from the default branch but checks out the code at the tag ref (`ref: ${{ inputs.tag }}`). This means build fixes committed to `main` won't be picked up — the old broken code at the tag gets built again.
-
-To retry a failed workflow, **always push a retry tag** on the commit you want to build:
-
-```bash
-# Desktop (all platforms)
-git tag -f desktop-v0.1.28 HEAD && git push origin desktop-v0.1.28 --force
-
-# Desktop (single platform)
-git tag -f desktop-macos-v0.1.28 HEAD && git push origin desktop-macos-v0.1.28 --force
-git tag -f desktop-linux-v0.1.28 HEAD && git push origin desktop-linux-v0.1.28 --force
-git tag -f desktop-windows-v0.1.28 HEAD && git push origin desktop-windows-v0.1.28 --force
-
-# Android APK
-git tag -f android-v0.1.28 HEAD && git push origin android-v0.1.28 --force
-```
-
-This ensures the checkout ref matches the actual code on `main` with the fix included.
-
-## Notes
-
-- `version:all:*` bumps root + syncs workspace versions and `@getpaseo/*` dependency versions
-- `release:prepare` refreshes workspace `node_modules` links to prevent stale types
-- `npm run dev:desktop` and `npm run build:desktop` target the Electron desktop package in `packages/desktop`
-- If `release:publish` partially fails, re-run it — npm skips already-published versions
-- The website parses the first `## X.Y.Z` heading in `CHANGELOG.md` to determine the download version. This is why changelog entries must only be added at finalization, not during drafts.
-
-## Changelog format
-
-The website depends on the changelog to determine the latest download version. The heading format **must** be strictly followed:
-
-```
-## X.Y.Z - YYYY-MM-DD
-```
-
-No prefix (`v`), no extra text. The parser matches the first `## X.Y.Z` line to extract the version. A malformed heading will break download links on the homepage.
-
-## Completion checklist
-
-- [ ] Update `CHANGELOG.md` with user-facing release notes (features, fixes — not refactors)
-- [ ] Verify the changelog heading follows strict `## X.Y.Z - YYYY-MM-DD` format
-- [ ] `npm run release:patch` (or `release:finalize` for drafts) completes successfully
-- [ ] GitHub `Desktop Release` workflow for the `v*` tag is green
-- [ ] GitHub `Android APK Release` workflow for the same tag is green
-- [ ] EAS `release-mobile.yml` workflow for the same tag is green

docs/ad-hoc-daemon-testing.md

@@ -0,0 +1,165 @@
+# Ad-hoc daemon testing
+
+Spin up an isolated in-process daemon test harness without touching the main daemon on port 6767.
+
+This is for test code only. Executable daemon processes must start through
+`scripts/supervisor-entrypoint.ts` or `dist/scripts/supervisor-entrypoint.js`;
+do not use `createPaseoDaemon` as a product launch path.
+
+## Quick start
+
+```typescript
+import os from "node:os";
+import path from "node:path";
+import { mkdir, mkdtemp, rm } from "node:fs/promises";
+import pino from "pino";
+import { createPaseoDaemon } from "./bootstrap.js";
+import { DaemonClient } from "./test-utils/daemon-client.js";
+
+const logger = pino({ level: "warn" });
+const paseoHomeRoot = await mkdtemp(path.join(os.tmpdir(), "paseo-test-"));
+const paseoHome = path.join(paseoHomeRoot, ".paseo");
+await mkdir(paseoHome, { recursive: true });
+const staticDir = await mkdtemp(path.join(os.tmpdir(), "paseo-static-"));
+
+const daemon = await createPaseoDaemon(
+  {
+    listen: "127.0.0.1:0", // OS picks a free port
+    paseoHome,
+    corsAllowedOrigins: [],
+    hostnames: true,
+    mcpEnabled: false,
+    staticDir,
+    mcpDebug: false,
+    agentClients: {},
+    agentStoragePath: path.join(paseoHome, "agents"),
+    relayEnabled: false,
+    relayEndpoint: "relay.paseo.sh:443",
+    appBaseUrl: "https://app.paseo.sh",
+    // Add custom config here, e.g.:
+    // providerOverrides: { ... },
+  },
+  logger,
+);
+
+await daemon.start();
+const target = daemon.getListenTarget();
+const port = target!.type === "tcp" ? target!.port : null;
+
+const client = new DaemonClient({
+  url: `ws://127.0.0.1:${port}/ws`,
+  appVersion: "0.1.70", // see gotcha #1
+});
+await client.connect();
+await client.fetchAgents({ subscribe: { subscriptionId: "test" } });
+
+// ... do your testing ...
+
+await client.close();
+await daemon.stop();
+await rm(paseoHomeRoot, { recursive: true, force: true });
+await rm(staticDir, { recursive: true, force: true });
+```
+
+Run with:
+
+```bash
+npx tsx packages/server/src/server/your-script.ts
+```
+
+## Using the test helper
+
+For simpler cases, `createTestPaseoDaemon` + `DaemonClient` handles temp dirs and port selection:
+
+```typescript
+import { createTestPaseoDaemon } from "./test-utils/paseo-daemon.js";
+import { DaemonClient } from "./test-utils/daemon-client.js";
+
+const daemon = await createTestPaseoDaemon();
+const client = new DaemonClient({
+  url: `ws://127.0.0.1:${daemon.port}/ws`,
+  appVersion: "0.1.70",
+});
+await client.connect();
+await client.fetchAgents({ subscribe: { subscriptionId: "test" } });
+
+// ... test ...
+
+await client.close();
+await daemon.close(); // stops daemon + cleans up temp dirs
+```
+
+The test helper does **not** expose `providerOverrides`. In test harnesses, use `createPaseoDaemon` directly when you need it (see quick start above).
+
+## Common client methods
+
+```typescript
+// Provider discovery
+const snapshot = await client.getProvidersSnapshot({ cwd: "/tmp" });
+const models = await client.listProviderModels("claude");
+const modes = await client.listProviderModes("claude");
+
+// Agent lifecycle
+const agent = await client.createAgent({ provider: "claude", cwd: "/tmp" });
+await client.sendMessage(agent.id, "Hello");
+const updated = await client.waitForAgentUpsert(agent.id, (s) => s.status === "idle");
+```
+
+## Gotchas
+
+### 1. appVersion gates provider visibility
+
+The daemon hides non-legacy providers (anything other than claude, codex, opencode) from clients that don't send an `appVersion >= 0.1.45`. The `DaemonClient` sends no version by default, so custom providers like ACP-based ones will be invisible in snapshot responses.
+
+Always pass `appVersion`:
+
+```typescript
+const client = new DaemonClient({
+  url: `ws://127.0.0.1:${port}/ws`,
+  appVersion: "0.1.70",
+});
+```
+
+### 2. Provider snapshots are async
+
+After the daemon starts, providers are probed in the background. The first `getProvidersSnapshot()` call will likely return `status: "loading"` for most providers. Poll until the provider you care about is no longer loading:
+
+```typescript
+let snapshot = await client.getProvidersSnapshot({ cwd: "/tmp" });
+for (let i = 0; i < 20; i++) {
+  const entry = snapshot.entries.find((e) => e.provider === "gemini");
+  if (entry && entry.status !== "loading") break;
+  await new Promise((r) => setTimeout(r, 2_000));
+  snapshot = await client.getProvidersSnapshot({ cwd: "/tmp" });
+}
+```
+
+### 3. fetchAgents is required before most operations
+
+Call `client.fetchAgents()` after connecting. The daemon session expects this handshake before it processes other requests — without it, messages like `get_providers_snapshot_request` will silently hang.
+
+### 4. listen: "127.0.0.1:0" for port allocation
+
+Always use port `0` so the OS picks a free port. Never hardcode a port — it will collide with the main daemon or other test runs.
+
+### 5. Script must live inside packages/server
+
+The test utilities use relative imports through the TypeScript project. Place your script somewhere under `packages/server/src/` and import from there. Scripts outside the repo will fail with module resolution errors.
+
+### 6. Cleanup on failure
+
+Wrap your test logic in try/finally to ensure the daemon stops and temp dirs are cleaned up, even if an assertion fails:
+
+```typescript
+try {
+  // ... test logic ...
+} finally {
+  await client.close();
+  await daemon.stop().catch(() => undefined);
+  await rm(paseoHomeRoot, { recursive: true, force: true });
+}
+```
+
+### 7. ACP providers spawn real processes
+
+When testing ACP providers (e.g., Gemini with `extends: "acp"`), the daemon will spawn real processes to probe for models and modes. The binary must be installed and on PATH. Probing can take 5-15 seconds depending on the provider.

docs/agent-lifecycle.md

@@ -0,0 +1,107 @@
+# Agent lifecycle
+
+How an agent is created, runs, becomes a subagent, gets archived, and disappears from the UI. The model spans the daemon (lifecycle, archive) and the client (tabs, the subagents track).
+
+## States
+
+```
+initializing → idle → running → idle (or error → closed)
+                 ↑        │
+                 └────────┘  (agent completes a turn, awaits next prompt)
+```
+
+Each agent in `AgentManager` carries a `lastStatus` of `initializing`, `idle`, `running`, `error`, or `closed`. State transitions persist to disk and stream to subscribed clients via WebSocket.
+
+## Relationships
+
+Agents can launch other agents via the `create_agent` MCP tool. When they do, the daemon stamps the new agent with a label `paseo.parent-agent-id` pointing back at the caller (`packages/server/src/server/agent/mcp-server.ts:804`). The client surfaces that as `agent.parentAgentId`.
+
+There is exactly one relationship type today: `parentAgentId`. The daemon does not distinguish between:
+
+- **Subagents** — children that exist as part of the parent's work (e.g. orchestration tasks the parent delegates and waits on)
+- **Detached agents** — children launched to take over from the parent (e.g. handoffs, fire-and-forget delegations)
+
+Both look the same in storage. This is an accepted limitation — see [Limitations](#limitations).
+
+## Archive
+
+Archive is a **soft delete**: the agent record stays on disk with `archivedAt` set, the runtime is closed, and the agent disappears from active lists. Archive is **global** — it lives on the server and propagates to every connected client.
+
+Archiving runs through `AgentManager.archiveAgent` (`packages/server/src/server/agent/agent-manager.ts`):
+
+1. Snapshot the current session into the registry
+2. Set `archivedAt` and normalize `lastStatus` away from `running`/`initializing`
+3. Notify subscribers
+4. Close the runtime (kills the process if still running)
+5. **Cascade-archive children** — any agent whose `paseo.parent-agent-id` label matches the archived agent gets archived too, recursively
+
+Cascade is what keeps subagent fleets from outliving their orchestrator.
+
+## Tabs vs archive
+
+These are two distinct concepts that used to be conflated:
+
+| Concept                    | Scope      | Triggers                   |
+| -------------------------- | ---------- | -------------------------- |
+| **Tab** (workspace layout) | Per-client | User opens/closes a view   |
+| **Archive** (lifecycle)    | Global     | Explicit lifecycle gesture |
+
+Closing a tab on a **root agent** still archives — the tab is the agent's home, so closing it means "I'm done with this agent." A confirm dialog protects against archiving a running agent by accident.
+
+Closing a tab on a **subagent** (any agent with `parentAgentId`) is **layout-only**. The agent stays unarchived and stays in its parent's track. The user can re-open the tab from the track at any time. This is implemented in `handleCloseAgentTab` (`packages/app/src/screens/workspace/workspace-screen.tsx`).
+
+The asymmetry is intentional: a subagent's home is the parent's track, not the tab. Tabs are ephemeral viewing slots; the track is the persistent record of the parent's children.
+
+## The subagents track
+
+The collapsible section above the composer in an agent's pane (`packages/app/src/subagents/subagents-section.tsx`). Membership rule (`packages/app/src/subagents/subagents.ts`):
+
+```
+parentAgentId === thisAgent.id  AND  !archivedAt
+```
+
+Archived subagents disappear from the track, by design. To remove a subagent from the track without closing its tab, use the **archive button (X)** on the row — it opens a confirm dialog and archives the subagent on confirm. That same archive shows the subagent leave the track on every connected client.
+
+## Why this shape
+
+The decision was to **decouple "close tab" from "archive" only for subagents**, rather than universally:
+
+- **Closing a tab on a root agent still archives** — preserves the existing UX users are trained on
+- **Closing a tab on a subagent is layout-only** — fixes the lossy "click to read, close to dismiss view, lose the row" flow
+- **Archive button on track rows** — gives subagents an explicit lifecycle gesture in their home surface
+- **Cascade archive on parent** — keeps subagents from leaking when the parent is archived
+
+We considered universal decoupling (no tab close ever archives, archive is always explicit) but rejected it: it changes a behavior root-agent users rely on.
+
+## Limitations
+
+### Detached agents are cascade-archived
+
+The daemon can't tell a "subagent" apart from a "detached agent" — both carry `paseo.parent-agent-id`. So when you archive an agent that previously launched a detached child (e.g. via `/paseo-handoff`), cascade will archive the detached child too, even though semantically it should outlive the originator.
+
+Until a richer relation model lands (e.g. a `relation: "subagent" | "detached"` field on creation, or a separate channel for handoff launches), this trade-off stands. Workaround: don't archive an agent whose work was handed off, or unarchive the detached child afterward.
+
+### Subagent accumulation under long-lived parents
+
+A parent that spawns many subagents will see the track grow. There's no automatic cleanup for completed subagents — the user prunes via the archive button on each row. A bulk gesture (e.g. "archive all idle children") could land later if this becomes a real problem.
+
+### Cross-client tab dismissal
+
+Closing a subagent's tab on one client doesn't affect other clients' layouts. This is the expected behavior of decoupled tabs and is consistent with how layouts have always worked. Archive remains the global gesture for cross-client cleanup.
+
+## Storage
+
+```
+$PASEO_HOME/agents/{cwd-with-dashes}/{agent-id}.json
+```
+
+Each agent is a single JSON file. Fields relevant to this doc:
+
+| Field                             | Type          | Meaning                                                       |
+| --------------------------------- | ------------- | ------------------------------------------------------------- |
+| `id`                              | `string`      | Stable identifier                                             |
+| `archivedAt`                      | `string?`     | Soft-delete timestamp (ISO 8601)                              |
+| `labels["paseo.parent-agent-id"]` | `string?`     | Parent agent ID, set automatically by `create_agent` MCP tool |
+| `lastStatus`                      | `AgentStatus` | `initializing` / `idle` / `running` / `error` / `closed`      |
+
+See [`docs/data-model.md`](./data-model.md) for the full agent record.

docs/android.md

@@ -4,9 +4,9 @@
 
 Controlled by `APP_VARIANT` in `packages/app/app.config.js` (vanilla Expo, no custom Gradle plugin):
 
-| Variant | App name | Package ID |
-|---|---|---|
-| `production` | Paseo | `sh.paseo` |
+| Variant       | App name    | Package ID       |
+| ------------- | ----------- | ---------------- |
+| `production`  | Paseo       | `sh.paseo`       |
 | `development` | Paseo Debug | `sh.paseo.debug` |
 
 EAS profiles: `development`, `production`, and `production-apk` in `packages/app/eas.json`.
@@ -46,11 +46,15 @@ adb exec-out screencap -p > screenshot.png
 
 ## Cloud build + submit (EAS)
 
-Tag pushes like `v0.1.0` trigger:
+Stable tag pushes like `v0.1.0` trigger:
 
 - `packages/app/.eas/workflows/release-mobile.yml` on Expo servers (iOS + Android build + submit)
 - `.github/workflows/android-apk-release.yml` on GitHub Actions (APK asset on GitHub Release)
 
+Beta tags like `v0.1.1-beta.1` only trigger the GitHub APK workflow. They publish a GitHub prerelease APK for testing and do not submit to the stores.
+
+`android-v*` tags also trigger only the GitHub APK workflow — useful when you want to ship an APK without going through stores. Both workflows also support `workflow_dispatch`; the GitHub APK one takes an existing `tag` input so you can rebuild without cutting a new tag.
+
 ### Useful commands
 
 ```bash

docs/architecture.md

@@ -0,0 +1,267 @@
+# Architecture
+
+Paseo is a client-server system for monitoring and controlling local AI coding agents. The daemon runs on your machine, manages agent processes, and streams their output in real time over WebSocket. Clients (mobile app, CLI, desktop app) connect to the daemon to observe and interact with agents.
+
+Your code never leaves your machine. Paseo is local-first.
+
+## System overview
+
+```
+┌─────────────┐    ┌─────────────┐    ┌─────────────┐
+│  Mobile App  │    │     CLI     │    │ Desktop App │
+│   (Expo)     │    │ (Commander) │    │ (Electron)  │
+└──────┬───────┘    └──────┬──────┘    └──────┬──────┘
+       │                   │                  │
+       │    WebSocket      │    WebSocket     │    Managed subprocess
+       │    (direct or     │    (direct)      │    + WebSocket
+       │     via relay)    │                  │
+       └───────────┬───────┴──────────────────┘
+                   │
+            ┌──────▼──────┐
+            │   Daemon    │
+            │  (Node.js)  │
+            └──────┬──────┘
+                   │
+      ┌────────────┼────────────┐
+      │            │            │
+┌─────▼─────┐ ┌───▼────┐ ┌────▼─────┐
+│  Claude   │ │ Codex  │ │ OpenCode │
+│  Agent    │ │ Agent  │ │  Agent   │
+│  SDK      │ │ Server │ │          │
+└───────────┘ └────────┘ └──────────┘
+```
+
+## Components at a glance
+
+- **Daemon:** Local server that spawns and manages agent processes and exposes the WebSocket API.
+- **App:** Cross-platform Expo client for iOS, Android, web, and the shared UI used by desktop.
+- **CLI:** Terminal interface for agent workflows that can also start and manage the daemon.
+- **Desktop app:** Electron wrapper around the web app that bundles and auto-manages its own daemon.
+- **Relay:** Optional encrypted bridge for remote access without opening ports directly.
+
+## Packages
+
+### `packages/server` — The daemon
+
+The heart of Paseo. A Node.js process that:
+
+- Listens for WebSocket connections from clients
+- Manages agent lifecycle (create, run, stop, resume, archive)
+- Streams agent output in real time via a timeline model
+- Exposes an MCP server for agent-to-agent control
+- Optionally connects outbound to a relay for remote access
+
+All paths are under `packages/server/src/`.
+
+**Key modules:**
+
+| Module                          | Responsibility                                                               |
+| ------------------------------- | ---------------------------------------------------------------------------- |
+| `server/bootstrap.ts`           | Daemon initialization: HTTP server, WS server, agent manager, storage, relay |
+| `server/websocket-server.ts`    | WebSocket connection management, hello handshake, binary frame routing       |
+| `server/session.ts`             | Per-client session state, timeline subscriptions, terminal operations        |
+| `server/agent/agent-manager.ts` | Agent lifecycle state machine, timeline tracking, subscriber management      |
+| `server/agent/agent-storage.ts` | File-backed JSON persistence at `$PASEO_HOME/agents/`                        |
+| `server/agent/mcp-server.ts`    | MCP server for sub-agent creation, permissions, timeouts                     |
+| `server/agent/providers/`       | Provider adapters (see "Agent providers" below)                              |
+| `server/relay-transport.ts`     | Outbound relay connection with E2E encryption                                |
+| `server/schedule/`              | Cron-based scheduled agents                                                  |
+| `server/loop-service.ts`        | Looping agent runs that retry until an exit condition                        |
+| `server/chat/`                  | Chat rooms for agent-to-agent and human-to-agent messaging                   |
+| `client/daemon-client.ts`       | Client library for connecting to the daemon (used by CLI and app)            |
+| `shared/messages.ts`            | Zod schemas for the entire wire protocol                                     |
+| `shared/binary-frames/`         | Terminal stream and file transfer binary frame codecs                        |
+
+### `packages/app` — Mobile + web client (Expo)
+
+Cross-platform React Native app that connects to one or more daemons.
+
+- Expo Router navigation (`/h/[serverId]/workspace/[workspaceId]`, `/h/[serverId]/agent/[agentId]`, etc.)
+- `HostRuntimeController` manages saved host connections, reconnection, and per-host runtime state
+- `SessionContext` wraps the daemon client for the active session
+- Timeline reducers in `timeline/session-stream-reducers.ts` handle compaction, gap detection, sequence-based deduplication
+- Voice features: dictation (STT) and voice agent (realtime)
+
+### `packages/cli` — Command-line client
+
+Commander.js CLI with Docker-style commands. Common agent operations are also exposed at the top level (e.g. `paseo ls`, `paseo run`).
+
+- `paseo agent ls/run/import/attach/logs/stop/delete/send/inspect/wait/archive/reload/update/mode`
+- `paseo daemon start/stop/restart/status/pair/set-password`
+- `paseo chat ls/create/inspect/post/read/wait/delete`
+- `paseo terminal ls/create/capture/send-keys/kill`
+- `paseo loop run/ls/inspect/logs/stop`
+- `paseo schedule create/ls/inspect/update/pause/resume/run-once/logs/delete`
+- `paseo permit allow/deny/ls`
+- `paseo provider ls/models`
+- `paseo worktree create/ls/archive`
+- `paseo speech …`
+
+Communicates with the daemon via the same WebSocket protocol as the app.
+
+### `packages/relay` — E2E encrypted relay
+
+Enables remote access when the daemon is behind a firewall.
+
+- ECDH key exchange + AES-256-GCM encryption
+- Relay server is zero-knowledge — it routes encrypted bytes, cannot read content
+- Client and daemon channels with identical API (`createClientChannel`, `createDaemonChannel`)
+- Pairing via QR code transfers the daemon's public key to the client
+- Self-hosted relays opt into TLS with `daemon.relay.useTls` or `PASEO_RELAY_USE_TLS=true`
+
+See [SECURITY.md](../SECURITY.md) for the full threat model.
+
+### `packages/desktop` — Desktop app (Electron)
+
+Electron wrapper for macOS, Linux, and Windows.
+
+- Can spawn the daemon as a managed subprocess
+- Native file access for workspace integration
+- Same WebSocket client as mobile app
+
+### `packages/website` — Marketing site
+
+TanStack Router + Cloudflare Workers. Serves paseo.sh.
+
+## WebSocket protocol
+
+All clients speak the same WebSocket protocol over a single connection that mixes JSON text frames and a small binary framing for terminal streams. Schemas live in `packages/server/src/shared/messages.ts`.
+
+**Handshake:**
+
+```
+Client → Server:  WSHelloMessage {
+                    type: "hello",
+                    clientId,
+                    clientType: "mobile" | "browser" | "cli" | "mcp",
+                    protocolVersion,
+                    appVersion?,
+                    capabilities?: { voice?, pushNotifications?, ... },
+                  }
+Server → Client:  status message with payload { status: "server_info",
+                    serverId, hostname, version, capabilities?, features }
+```
+
+There is no dedicated welcome message; the server emits a `status` session message after accepting the hello, then begins streaming. The session stores client capabilities from the hello and rehydrates them on reconnect, so the wire boundary can ask one question: `session.supports(...)`.
+
+**Top-level WS envelopes** are `hello`, `recording_state`, `ping`/`pong`, and `session` (which wraps the rich union of session messages).
+
+**Notable session message types:**
+
+- `agent_update` — Agent state changed (status, title, labels)
+- `agent_stream` — New timeline event from a running agent
+- `workspace_update`, `script_status_update`, `workspace_setup_progress` — Workspace state
+- `agent_permission_request` / `agent_permission_resolved` — Tool-call permission flow
+- `agent_deleted`, `agent_archived`, `agent_status`, `agent_list`
+- `checkout_status_update`, `checkout_diff_update`, and the full `checkout_*` request/response set for git operations
+- Terminal subscribe/input/capture commands
+- Voice/dictation streaming events (`dictation_stream_*`, `assistant_chunk`, `audio_output`, `transcription_result`)
+- Request/response pairs for fetch, list, create, etc., correlated by `requestId`; failures use `rpc_error`
+
+**Binary frames (terminal stream protocol):**
+
+Terminal I/O is sent as binary WebSocket frames decoded by `decodeTerminalStreamFrame` in `shared/binary-frames/terminal.ts`. The layout is:
+
+- 1-byte opcode: `Output (0x01)`, `Input (0x02)`, `Resize (0x03)`, `Snapshot (0x04)`
+- 1-byte slot: terminal slot id
+- variable payload: bytes for output/input, JSON-encoded `{ rows, cols }` for resize, terminal snapshot for snapshot
+
+There is also a separate file-transfer binary frame format in the same directory, used for download/upload streams.
+
+### Compatibility rules
+
+- WebSocket schemas are append-only. Add fields, do not remove fields, and never make optional fields required.
+- New wire enum values must be gated at serialization with `session.supports(CLIENT_CAPS.someCapability)`.
+- `Session` stores client capabilities from the `hello` handshake and rehydrates them on reconnect, so the wire boundary can ask one question: `session.supports(...)`.
+
+Example: adding a new enum value
+
+```ts
+// 1. Add CLIENT_CAPS.newThing = "new_thing"
+// 2. Let new clients advertise it in WS hello
+// 3. Keep the shared producer schema strict
+// 4. Gate the new emitted value: session.supports(CLIENT_CAPS.newThing) ? "new_value" : "old_value"
+```
+
+## Agent lifecycle
+
+The lifecycle states are defined in `shared/agent-lifecycle.ts`:
+
+```
+initializing → idle ⇄ running
+        ↓       ↓        ↓
+              error
+                ↓
+              closed
+```
+
+- `initializing` — provider session is being created
+- `idle` — has a live session, awaiting the next prompt
+- `running` — provider is currently producing a turn
+- `error` — last attempt failed; session is still attached
+- `closed` — terminal state, no live session
+
+`ManagedAgent` is a discriminated union over those lifecycle tags. Notes:
+
+- **AgentManager** is the source of truth for agent state and broadcasts updates to all subscribers
+- Timeline is append-only with epochs (each run starts a new epoch). Storage uses sequence numbers for client-side dedup; the default fetch page is 200 items
+- Events stream to all subscribed clients in real time
+- Agent state persists to `$PASEO_HOME/agents/{cwd-with-dashes}/{agent-id}.json` (timeline rows live alongside the record)
+
+## Agent providers
+
+Each provider implements the `AgentClient` interface in `agent/agent-sdk-types.ts`. Provider implementations live in `agent/providers/`.
+
+The three first-class, user-facing providers are Claude Code, Codex, and OpenCode. Additional adapters exist in the same directory for ACP-compatible agents and internal use:
+
+| Provider           | Wraps                                | Session format                                     |
+| ------------------ | ------------------------------------ | -------------------------------------------------- |
+| Claude (`claude/`) | Anthropic Agent SDK                  | `~/.claude/projects/{cwd}/{session-id}.jsonl`      |
+| Codex              | Codex AppServer (`codex-app-server`) | `~/.codex/sessions/{date}/rollout-{ts}-{id}.jsonl` |
+| OpenCode           | OpenCode server / CLI                | Provider-managed                                   |
+| Cursor / Copilot   | ACP wrapper (`acp-agent`)            | Provider-managed                                   |
+| Generic ACP        | ACP wrapper                          | Provider-managed                                   |
+| Pi-direct          | Direct Anthropic API call            | Stateless                                          |
+| Mock load test     | In-process fake                      | In-memory                                          |
+
+All providers:
+
+- Handle their own authentication (Paseo does not manage API keys)
+- Support session resume via persistence handles
+- Map tool calls to a normalized `ToolCallDetail` type
+- Expose provider-specific modes (plan, default, full-access)
+
+## Data flow: running an agent
+
+1. Client sends `CreateAgentRequestMessage` with config (prompt, cwd, provider, model, mode)
+2. Session routes to `AgentManager.create()`
+3. AgentManager creates a `ManagedAgent`, initializes provider session
+4. Provider runs the agent → emits `AgentStreamEvent` items
+5. Events append to the agent timeline, broadcast to all subscribed clients
+6. Tool calls are normalized to `ToolCallDetail` (shell, read, edit, write, search, etc.)
+7. Permission requests flow: agent → server → client → user decision → server → agent
+
+## Storage
+
+`$PASEO_HOME` defaults to `~/.paseo`. The most important files:
+
+```
+$PASEO_HOME/
+├── agents/{cwd-with-dashes}/{agent-id}.json   # Agent record + persisted timeline rows
+├── projects/projects.json                      # Project registry
+├── projects/workspaces.json                    # Workspace registry
+├── chat/                                       # Chat rooms
+├── schedules/                                  # Scheduled-agent definitions and runs
+├── loops/                                      # Loop runs and logs
+├── config.json                                 # Daemon config (mutable)
+├── daemon-keypair.json                         # Daemon identity for relay/E2EE
+├── push-tokens.json                            # Mobile push tokens
+├── paseo.sock / paseo.pid                      # Local IPC socket and pidfile
+└── daemon.log                                  # Daemon trace logs (rotated)
+```
+
+## Deployment models
+
+1. **Local daemon** (default): `paseo daemon start` on `127.0.0.1:6767`
+2. **Managed desktop**: Electron app spawns daemon as subprocess
+3. **Remote + relay**: Daemon behind firewall, relay bridges with E2E encryption

docs/coding-standards.md

@@ -0,0 +1,98 @@
+# Coding Standards
+
+The core instinct: AI-generated code hedges — it covers every case, layers over instead of cutting in, scatters uncertainty everywhere, wraps in case. A senior engineer commits — to a shape, a boundary, a name, a happy path, a type — and lets everything else fall into place. Every rule below catches a different form of indecision.
+
+For testing rules, see [testing.md](testing.md).
+
+## Core principles
+
+- **Zero complexity budget** — every abstraction must justify itself with a specific, current benefit.
+- **YAGNI** — build features and abstractions only when needed. A function called once is indirection, not abstraction.
+- **No "while I'm at it" cleanups** — make the change you came for. Drive-by edits hide in the diff.
+- **Functional and declarative** over object-oriented.
+- **`function` declarations** over arrow function assignments.
+- **`interface`** over `type` when both work.
+- **No `index.ts` barrel files** that only re-export — they create indirection and circular-dep risk. Import from the source.
+
+## Comments and noise
+
+- Delete any comment where removing it loses zero information. Comments explain _why_, not _what_.
+- No tutorial comments explaining language features (`// Use destructuring to...`).
+- No decorative section dividers (`// ===== Helpers =====`). Use files and modules to organize, not ASCII art.
+- No hedging comments (`// might need to revisit`, `// should work for most cases`). If you're unsure, investigate.
+- No commented-out code. Git remembers.
+- No `console.log` / `debugger` left behind. No `TODO: implement` stubs — if it needs to exist, write it.
+
+## Confidence: commit to a shape
+
+- Validate at boundaries (network, IPC, user input, file I/O), trust types internally. After the parse, the value is what its type says.
+- Every `?.` and `??` past the validation boundary is unconfident code — either the boundary should resolve it, or the type should reflect reality.
+- No defensive checks for conditions the type system already rules out (`if (!agent) return` on a non-nullable parameter).
+- No `try/catch` "just in case." If you can't say what you're catching and why, don't catch.
+- Optionality is a design decision, not a migration shortcut. Distinct valid states → discriminated union. Intentionally empty → explicit `null`. Keep optionality at real boundaries.
+
+## Types
+
+- No `any`. No `as` casts to bypass errors. No `@ts-ignore` / `@ts-expect-error`. Narrow with `if` / schema validation; let the compiler check harder, not less.
+- If a Zod schema exists, the TypeScript type is `z.infer<typeof schema>`. Never hand-write a parallel type.
+- One canonical type per concept. Layer-specific views are `Pick` / `Omit`, not duplicated fields.
+- Name multi-property object shapes — no inline `Array<{ ... }>` or `Promise<{ ... }>` in signatures, returns, or generic args.
+- Use string literal unions, not raw `string`, when the value is one of a known set. Catches typos at compile time.
+- Object parameters past the obvious-name threshold: 3+ args, any boolean arg, any optional arg → object. `(thing, true, false, true)` is unreadable at the call site.
+- Make impossible states impossible — discriminated unions over `{ isLoading; error?; data? }` bags.
+
+## Errors
+
+- Throw typed error classes that carry the fields a caller would want to read. Plain `Error("Provider X not found")` collapses structured info into a string.
+- Catch blocks branch on `instanceof` for what they can handle; rethrow the rest. No `catch (e) { return null }`.
+- Separate user-facing copy from log/debug strings — don't make one string serve telemetry, logs, and the UI.
+- Fail explicitly. If the caller asked for X and X isn't available, throw — don't silently substitute Y.
+
+## Density
+
+- Nested ternaries are forbidden. A single ternary is fine only when both branches are a single identifier or trivial access (`x ? a : b`).
+- Boolean expressions with 2+ clauses or mixed concerns → name the conditions.
+- Object literals assemble pre-computed values; don't pack branching and lookups into property positions.
+- Operations wrapping operations (`Object.fromEntries(arr.filter(...).map(...))`, `Math.max(...xs.map(...))`) → break into named intermediates.
+- Max 3 levels of nesting (callbacks, JSX, control flow). Above that, extract.
+
+## Structure and modules
+
+- A directory is a module, not a namespace. One intentional public surface; internal files stay internal.
+- Path is part of the name — prefer `provider/registry.ts` over `provider/provider-registry.ts`. If the filename has to do double duty, deepen the path.
+- Filenames ending in `-utils`, `-helpers`, `-manager`, `-handler`, `-controller`, `-formatter`, `-builder` are a smell — the path didn't carry enough domain.
+- Boundary returns answer the caller's question (`getActiveAgents()`), not "here's my storage" (`getAgents().filter(...)` repeated everywhere).
+- One adapter means a hypothetical seam; two adapters means a real one. Don't define a port until something actually varies across it.
+- Pass-through modules fail the deletion test — if removing the module makes callers go straight to what they wanted, delete it.
+- Centralize policy. The same discriminator (`plan`, `provider`, `kind`, `status`) branched in 3+ files → policy table, not another `else if` per case.
+- New features get a home before implementation. A feature smeared across 5 shared files is the same slop as a flat-peer namespace.
+- Don't drop new files at the nearest root just because placement is unclear — say so and ask.
+
+## Refactoring is a bolt-on test
+
+- A change should look like a thoughtful edit to existing code, not a new layer next to it. New coordinator wrapping a coordinator, new flag bypassing the normal path, new helper duplicating an existing selector — stop and reshape instead.
+- Refactors preserve behavior by default. No removing features to simplify code without explicit approval.
+- Have a verification plan _before_ refactoring — name the invariants, confirm a test holds them, write one if not. See [testing.md](testing.md).
+- Migrate all callers and remove old paths in the same refactor. No fallback behavior unless explicitly designed.
+
+## React
+
+- `useEffect` is for synchronizing with external systems (DOM, network, timers, subscriptions). Not for transforming React state. Derived state → compute in render or `useMemo`.
+- No effect cascades — chains of effects setting state that triggers more effects almost always want React Query or a reducer.
+- `useRef` is for DOM refs and non-rendering identities (timer IDs, AbortController, latest-callback caches). If the value affects what renders next, it's state — model it explicitly with `useReducer` and a discriminated union.
+- Server state goes through React Query. Manual `useState` + `useEffect` + `isLoading` + `error` for fetched data is always worse.
+- Components render and dispatch — they don't compute transitions. Two-plus interacting `useState`s → extract a reducer.
+- Never define components inside other components. Module-scope only.
+- Subscribe narrowly: select primitives from stores, pass `status` not `agent`, use `useShallow` / deep-equal when returning derived arrays/objects.
+- Stable references for props that cross `memo` boundaries or feed dependency arrays. Static literals at module scope `as const`; derived with `useMemo`; handlers with `useCallback` only when there's a memoized beneficiary.
+- Use stable ids for `key`, never array index for reorderable/filterable lists.
+- Context for stable values (theme, auth). Store with selectors for state that changes.
+
+## Naming
+
+- Names describe meaning, not mechanics. `submitForm` over `handleOnClickButtonSubmit`. `running` over `filteredArrayOfRunningAgents`.
+- The right length is the shortest unambiguous in context. Inside `AgentManager`, methods are `start`, `stop`, `list`.
+- Match the surrounding code's vocabulary. If the codebase uses `getX`, don't introduce `fetchX` / `retrieveX` for the same shape.
+- Don't leak implementation into names — `getAgent`, not `queryPostgresForAgent`. If swapping the impl would force a rename, the name is wrong.
+- Booleans read as yes/no questions: `isX`, `hasX`, `canX`. Avoid negative booleans (`isNotConnected`).
+- `data`, `result`, `info`, `manager`, `temp` are smells — say what the thing _is_.

docs/custom-providers.md

@@ -0,0 +1,591 @@
+# Custom Provider Configuration
+
+Paseo supports configuring custom agent providers through `config.json` (located at `$PASEO_HOME/config.json`, typically `~/.paseo/config.json`). You can extend built-in providers with different API backends, add ACP-compatible agents, set custom binaries, disable providers, and create multiple profiles for the same underlying provider.
+
+All provider configuration lives under `agents.providers` in config.json:
+
+```json
+{
+  "version": 1,
+  "agents": {
+    "providers": {
+      "provider-id": { ... }
+    }
+  }
+}
+```
+
+Provider IDs must be lowercase alphanumeric with hyphens (`/^[a-z][a-z0-9-]*$/`).
+
+---
+
+## Table of Contents
+
+- [Extending a built-in provider](#extending-a-built-in-provider)
+- [Z.AI (Zhipu) coding plan](#zai-zhipu-coding-plan)
+- [Alibaba Cloud (Qwen) coding plan](#alibaba-cloud-qwen-coding-plan)
+- [Multiple profiles for the same provider](#multiple-profiles-for-the-same-provider)
+- [Custom binary for a provider](#custom-binary-for-a-provider)
+- [Disabling a provider](#disabling-a-provider)
+- [ACP providers](#acp-providers)
+- [Provider override reference](#provider-override-reference)
+
+---
+
+## Extending a built-in provider
+
+Use `extends` to create a new provider entry that inherits from a built-in provider (claude, codex, copilot, opencode, pi). The new provider gets its own entry in the provider list, with its own label, environment, and model definitions.
+
+```json
+{
+  "agents": {
+    "providers": {
+      "my-claude": {
+        "extends": "claude",
+        "label": "My Claude",
+        "description": "Claude with custom API endpoint",
+        "env": {
+          "ANTHROPIC_API_KEY": "sk-ant-...",
+          "ANTHROPIC_BASE_URL": "https://my-proxy.example.com/v1"
+        }
+      }
+    }
+  }
+}
+```
+
+Required fields for custom providers:
+
+- `extends` — which built-in provider to inherit from (or `"acp"`)
+- `label` — display name in the UI
+
+---
+
+## Z.AI (Zhipu) coding plan
+
+[Z.AI](https://z.ai) is a Chinese AI company (Zhipu AI) that offers an Anthropic-compatible API endpoint. Their GLM Coding Plan provides flat-rate access to GLM models through Claude Code's Anthropic API protocol. These are **not** Anthropic Claude models — they are Zhipu's own GLM models exposed through an Anthropic-compatible API.
+
+### Setup
+
+1. Register at [z.ai](https://z.ai) and subscribe to a coding plan
+2. Create an API key from the Z.AI dashboard
+3. Add a provider entry in config.json:
+
+```json
+{
+  "agents": {
+    "providers": {
+      "zai": {
+        "extends": "claude",
+        "label": "ZAI",
+        "env": {
+          "ANTHROPIC_AUTH_TOKEN": "<your-zai-api-key>",
+          "ANTHROPIC_BASE_URL": "https://api.z.ai/api/anthropic",
+          "API_TIMEOUT_MS": "3000000"
+        },
+        "disallowedTools": ["WebSearch"],
+        "models": [
+          { "id": "glm-4.5-air", "label": "GLM 4.5 Air" },
+          { "id": "glm-5-turbo", "label": "GLM 5 Turbo", "isDefault": true },
+          { "id": "glm-5.1", "label": "GLM 5.1" }
+        ]
+      }
+    }
+  }
+}
+```
+
+### Available models
+
+| Model         | Tier                |
+| ------------- | ------------------- |
+| `glm-5.1`     | Advanced (flagship) |
+| `glm-5-turbo` | Advanced            |
+| `glm-4.7`     | Standard            |
+| `glm-4.5-air` | Lightweight         |
+
+### Notes
+
+- `ANTHROPIC_AUTH_TOKEN` is used instead of `ANTHROPIC_API_KEY` — this is the z.ai API key
+- The `API_TIMEOUT_MS` env var extends the request timeout (z.ai can be slower than direct Anthropic)
+- If you get auth errors, run `/logout` inside Claude Code before switching to the z.ai provider
+- Web search (`WebSearch` tool) is an Anthropic-only server-side feature — third-party endpoints don't support it. Add `"disallowedTools": ["WebSearch"]` to avoid errors.
+- Automated setup is also available: `npx @z_ai/coding-helper`
+- Official docs: [docs.z.ai/devpack/tool/claude](https://docs.z.ai/devpack/tool/claude)
+
+---
+
+## Alibaba Cloud (Qwen) coding plan
+
+[Alibaba Cloud Model Studio](https://www.alibabacloud.com/en/campaign/ai-scene-coding) offers a coding plan that routes Claude Code requests to Qwen models through an Anthropic-compatible API. Like z.ai, these are **not** Anthropic Claude models.
+
+### Setup
+
+1. Go to the [Coding Plan page](https://modelstudio.console.alibabacloud.com/ap-southeast-1/?tab=globalset#/efm/coding_plan) on Alibaba Cloud Model Studio (Singapore region)
+2. Subscribe to the Pro plan ($50/month)
+3. Obtain your plan-specific API key (format: `sk-sp-xxxxx`) — this is different from a standard Model Studio key
+4. Add a provider entry in config.json:
+
+```json
+{
+  "agents": {
+    "providers": {
+      "qwen": {
+        "extends": "claude",
+        "label": "Qwen (Alibaba)",
+        "env": {
+          "ANTHROPIC_AUTH_TOKEN": "sk-sp-<your-coding-plan-key>",
+          "ANTHROPIC_BASE_URL": "https://coding-intl.dashscope.aliyuncs.com/apps/anthropic"
+        },
+        "disallowedTools": ["WebSearch"],
+        "models": [
+          { "id": "qwen3.5-plus", "label": "Qwen 3.5 Plus", "isDefault": true },
+          { "id": "qwen3-coder-next", "label": "Qwen 3 Coder Next" },
+          { "id": "kimi-k2.5", "label": "Kimi K2.5" }
+        ]
+      }
+    }
+  }
+}
+```
+
+### API endpoints
+
+| Mode                            | Base URL                                                    |
+| ------------------------------- | ----------------------------------------------------------- |
+| Coding plan (subscription)      | `https://coding-intl.dashscope.aliyuncs.com/apps/anthropic` |
+| Pay-as-you-go (no subscription) | `https://dashscope-intl.aliyuncs.com/apps/anthropic`        |
+
+For pay-as-you-go, use `ANTHROPIC_API_KEY` with a standard Model Studio key (`sk-xxxxx`) instead of `ANTHROPIC_AUTH_TOKEN`.
+
+### Available models
+
+**Recommended for coding plan:**
+
+| Model              | Notes                       |
+| ------------------ | --------------------------- |
+| `qwen3.5-plus`     | Vision capable, recommended |
+| `qwen3-coder-next` | Optimized for coding        |
+| `kimi-k2.5`        | Vision capable              |
+| `glm-5`            | Zhipu GLM                   |
+| `MiniMax-M2.5`     | MiniMax                     |
+
+**Additional models (pay-as-you-go):**
+`qwen3-max`, `qwen3.5-flash`, `qwen3-coder-plus`, `qwen3-coder-flash`, `qwen3-vl-plus`, `qwen3-vl-flash`
+
+### Notes
+
+- API keys must be created in the **Singapore region**
+- The coding plan is for personal use only in interactive coding tools
+- Web search (`WebSearch` tool) is an Anthropic-only server-side feature — third-party endpoints don't support it. Add `"disallowedTools": ["WebSearch"]` to avoid errors.
+- Official docs: [alibabacloud.com/help/en/model-studio/claude-code-coding-plan](https://www.alibabacloud.com/help/en/model-studio/claude-code-coding-plan)
+
+---
+
+## Multiple profiles for the same provider
+
+You can create multiple entries that extend the same built-in provider. Each gets its own entry in the provider list with independent credentials, models, and environment.
+
+Example: two different Anthropic accounts as separate profiles:
+
+```json
+{
+  "agents": {
+    "providers": {
+      "claude-work": {
+        "extends": "claude",
+        "label": "Claude (Work)",
+        "description": "Work Anthropic account",
+        "env": {
+          "ANTHROPIC_API_KEY": "sk-ant-work-..."
+        }
+      },
+      "claude-personal": {
+        "extends": "claude",
+        "label": "Claude (Personal)",
+        "description": "Personal Anthropic account",
+        "env": {
+          "ANTHROPIC_API_KEY": "sk-ant-personal-..."
+        }
+      }
+    }
+  }
+}
+```
+
+Each profile appears as a separate provider in the Paseo app. You can select which one to use when launching an agent.
+
+You can also combine profiles with model overrides to pin specific models per profile:
+
+```json
+{
+  "agents": {
+    "providers": {
+      "claude-fast": {
+        "extends": "claude",
+        "label": "Claude (Fast)",
+        "models": [{ "id": "claude-sonnet-4-6", "label": "Sonnet 4.6", "isDefault": true }]
+      },
+      "claude-smart": {
+        "extends": "claude",
+        "label": "Claude (Smart)",
+        "models": [{ "id": "claude-opus-4-6", "label": "Opus 4.6", "isDefault": true }]
+      }
+    }
+  }
+}
+```
+
+---
+
+## Custom binary for a provider
+
+Override the command used to launch any provider with the `command` field. This is an array where the first element is the binary and the rest are arguments.
+
+### Override a built-in provider's binary
+
+```json
+{
+  "agents": {
+    "providers": {
+      "claude": {
+        "command": ["/opt/claude-nightly/claude"]
+      }
+    }
+  }
+}
+```
+
+### Use a custom wrapper script
+
+```json
+{
+  "agents": {
+    "providers": {
+      "claude": {
+        "command": ["/usr/local/bin/my-claude-wrapper", "--verbose"]
+      }
+    }
+  }
+}
+```
+
+### Custom binary on a derived provider
+
+```json
+{
+  "agents": {
+    "providers": {
+      "my-codex": {
+        "extends": "codex",
+        "label": "Codex (Custom Build)",
+        "command": ["/home/user/codex-dev/target/release/codex"]
+      }
+    }
+  }
+}
+```
+
+The `command` array completely replaces the default command for that provider. The binary must exist on the system — Paseo checks for its availability and will mark the provider as unavailable if not found.
+
+---
+
+## Disabling a provider
+
+Set `enabled: false` to hide a provider from the provider list. The provider will not appear in the app or CLI.
+
+```json
+{
+  "agents": {
+    "providers": {
+      "copilot": { "enabled": false },
+      "codex": { "enabled": false }
+    }
+  }
+}
+```
+
+This works for both built-in and custom providers. To re-enable, set `enabled: true` or remove the `enabled` field entirely (providers are enabled by default).
+
+---
+
+## ACP providers
+
+The [Agent Client Protocol (ACP)](https://agentclientprotocol.com) is an open standard for communication between editors and AI coding agents — think LSP but for AI agents. Any agent that supports ACP can be added to Paseo as a custom provider.
+
+ACP agents communicate over JSON-RPC 2.0 on stdio. Paseo spawns the agent process and talks to it through stdin/stdout.
+
+### Adding a generic ACP provider
+
+Set `extends: "acp"` and provide a `command`:
+
+```json
+{
+  "agents": {
+    "providers": {
+      "my-agent": {
+        "extends": "acp",
+        "label": "My Agent",
+        "command": ["my-agent-binary", "--acp"],
+        "env": {
+          "MY_API_KEY": "..."
+        }
+      }
+    }
+  }
+}
+```
+
+Required fields for ACP providers:
+
+- `extends: "acp"`
+- `label`
+- `command` — the command to spawn the agent process (must support ACP over stdio)
+
+### Example: Google Gemini CLI
+
+[Gemini CLI](https://github.com/google-gemini/gemini-cli) supports ACP via the `--acp` flag.
+
+1. Install: `npm install -g @google/gemini-cli` or see [Gemini CLI docs](https://github.com/google-gemini/gemini-cli)
+2. Authenticate with Google (Gemini CLI handles its own auth)
+3. Add to config.json:
+
+```json
+{
+  "agents": {
+    "providers": {
+      "gemini": {
+        "extends": "acp",
+        "label": "Google Gemini",
+        "command": ["gemini", "--acp"]
+      }
+    }
+  }
+}
+```
+
+Ref: [Gemini CLI ACP mode docs](https://github.com/google-gemini/gemini-cli/blob/main/docs/cli/acp-mode.md)
+
+### Example: Hermes (Nous Research)
+
+[Hermes](https://github.com/NousResearch/hermes-agent) is an open-source coding agent by Nous Research with persistent memory and multi-provider LLM support. It supports ACP via the `acp` subcommand.
+
+1. Install: `curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash`
+2. Install ACP support: `pip install -e '.[acp]'`
+3. Configure Hermes credentials in `~/.hermes/`
+4. Add to config.json:
+
+```json
+{
+  "agents": {
+    "providers": {
+      "hermes": {
+        "extends": "acp",
+        "label": "Hermes",
+        "description": "Nous Research self-improving AI agent",
+        "command": ["hermes", "acp"]
+      }
+    }
+  }
+}
+```
+
+Ref: [Hermes ACP docs](https://hermes-agent.nousresearch.com/docs/user-guide/features/acp)
+
+### How ACP providers work in Paseo
+
+When you launch an agent with an ACP provider:
+
+1. Paseo spawns the process using the configured `command`
+2. Sends an `initialize` JSON-RPC request over stdin
+3. The agent responds with its capabilities, available modes, and models
+4. Paseo creates a session and sends prompts through the ACP protocol
+5. The agent streams responses, tool calls, and permission requests back over stdout
+
+Models and modes are discovered dynamically at runtime from the agent process. If you want to override the model list (e.g., to curate which models appear in the UI), use the `models` field:
+
+```json
+{
+  "agents": {
+    "providers": {
+      "my-agent": {
+        "extends": "acp",
+        "label": "My Agent",
+        "command": ["my-agent", "--acp"],
+        "models": [
+          { "id": "fast-model", "label": "Fast", "isDefault": true },
+          { "id": "smart-model", "label": "Smart" }
+        ]
+      }
+    }
+  }
+}
+```
+
+Profile models (defined in config.json) completely replace runtime-discovered models when present.
+
+If you want to keep runtime-discovered models and add or relabel a few entries, use `additionalModels` instead.
+
+Example: add an experimental model while keeping every model the provider discovers at runtime:
+
+```json
+{
+  "agents": {
+    "providers": {
+      "my-agent": {
+        "extends": "acp",
+        "label": "My Agent",
+        "command": ["my-agent", "--acp"],
+        "additionalModels": [
+          { "id": "experimental-model", "label": "Experimental", "isDefault": true }
+        ]
+      }
+    }
+  }
+}
+```
+
+Example: relabel a discovered model without replacing the full list:
+
+```json
+{
+  "agents": {
+    "providers": {
+      "my-agent": {
+        "extends": "acp",
+        "label": "My Agent",
+        "command": ["my-agent", "--acp"],
+        "additionalModels": [{ "id": "provider/model-id", "label": "My Preferred Label" }]
+      }
+    }
+  }
+}
+```
+
+When an `additionalModels` entry has the same `id` as a discovered model, it updates that model in place.
+
+---
+
+## Provider override reference
+
+Every entry under `agents.providers` accepts these fields:
+
+| Field              | Type                     | Required          | Description                                                        |
+| ------------------ | ------------------------ | ----------------- | ------------------------------------------------------------------ |
+| `extends`          | `string`                 | Yes (custom only) | Built-in provider ID to inherit from, or `"acp"`                   |
+| `label`            | `string`                 | Yes (custom only) | Display name in the UI                                             |
+| `description`      | `string`                 | No                | Short description shown in the UI                                  |
+| `command`          | `string[]`               | Yes (ACP only)    | Command to spawn the agent process                                 |
+| `env`              | `Record<string, string>` | No                | Environment variables to set for the agent process                 |
+| `models`           | `ProviderProfileModel[]` | No                | Static model list (overrides runtime discovery)                    |
+| `additionalModels` | `ProviderProfileModel[]` | No                | Static model additions (merged with runtime discovery or `models`) |
+| `disallowedTools`  | `string[]`               | No                | Tool names to disable for this provider (e.g. `["WebSearch"]`)     |
+| `enabled`          | `boolean`                | No                | Set to `false` to hide the provider (default: `true`)              |
+| `order`            | `number`                 | No                | Sort order in the provider list                                    |
+
+### Model definition
+
+Each entry in the `models` array:
+
+| Field             | Type               | Required | Description                           |
+| ----------------- | ------------------ | -------- | ------------------------------------- |
+| `id`              | `string`           | Yes      | Model identifier sent to the provider |
+| `label`           | `string`           | Yes      | Display name in the UI                |
+| `description`     | `string`           | No       | Short description                     |
+| `isDefault`       | `boolean`          | No       | Mark as the default model selection   |
+| `thinkingOptions` | `ThinkingOption[]` | No       | Available thinking/reasoning levels   |
+
+### Thinking option
+
+| Field         | Type      | Required | Description                         |
+| ------------- | --------- | -------- | ----------------------------------- |
+| `id`          | `string`  | Yes      | Thinking option identifier          |
+| `label`       | `string`  | Yes      | Display name                        |
+| `description` | `string`  | No       | Short description                   |
+| `isDefault`   | `boolean` | No       | Mark as the default thinking option |
+
+### Gotcha: `extends: "claude"` with third-party endpoints
+
+When a custom provider extends `"claude"` but points `ANTHROPIC_BASE_URL` at a non-Anthropic API (Z.AI, Alibaba/Qwen, proxies), the Claude Agent SDK may try to use Anthropic-only server-side tools like `WebSearch`. Third-party APIs don't support these tools, causing errors.
+
+Use `disallowedTools` to disable unsupported tools:
+
+```json
+{
+  "agents": {
+    "providers": {
+      "my-proxy": {
+        "extends": "claude",
+        "label": "My Proxy",
+        "env": {
+          "ANTHROPIC_BASE_URL": "https://my-proxy.example.com/v1"
+        },
+        "disallowedTools": ["WebSearch"]
+      }
+    }
+  }
+}
+```
+
+### Valid `extends` values
+
+Built-in providers: `claude`, `codex`, `copilot`, `opencode`, `pi`
+
+Special value: `acp` — creates a generic ACP provider (requires `command`)
+
+### Full example
+
+A config.json with multiple custom providers:
+
+```json
+{
+  "version": 1,
+  "agents": {
+    "providers": {
+      "copilot": { "enabled": false },
+
+      "zai": {
+        "extends": "claude",
+        "label": "ZAI",
+        "env": {
+          "ANTHROPIC_AUTH_TOKEN": "<zai-api-key>",
+          "ANTHROPIC_BASE_URL": "https://api.z.ai/api/anthropic",
+          "API_TIMEOUT_MS": "3000000"
+        },
+        "disallowedTools": ["WebSearch"],
+        "models": [
+          { "id": "glm-4.5-air", "label": "GLM 4.5 Air" },
+          { "id": "glm-5-turbo", "label": "GLM 5 Turbo", "isDefault": true },
+          { "id": "glm-5.1", "label": "GLM 5.1" }
+        ]
+      },
+
+      "qwen": {
+        "extends": "claude",
+        "label": "Qwen (Alibaba)",
+        "env": {
+          "ANTHROPIC_AUTH_TOKEN": "sk-sp-<coding-plan-key>",
+          "ANTHROPIC_BASE_URL": "https://coding-intl.dashscope.aliyuncs.com/apps/anthropic"
+        },
+        "disallowedTools": ["WebSearch"],
+        "models": [
+          { "id": "qwen3.5-plus", "label": "Qwen 3.5 Plus", "isDefault": true },
+          { "id": "qwen3-coder-next", "label": "Qwen 3 Coder Next" }
+        ]
+      },
+
+      "gemini": {
+        "extends": "acp",
+        "label": "Google Gemini",
+        "command": ["gemini", "--acp"]
+      },
+
+      "hermes": {
+        "extends": "acp",
+        "label": "Hermes",
+        "command": ["hermes", "acp"]
+      }
+    }
+  }
+}
+```

docs/data-model.md

@@ -0,0 +1,449 @@
+# Data Model
+
+Paseo uses **file-based JSON persistence** instead of a traditional database. All data is validated at runtime with Zod schemas. Most stores write atomically (write to temp file, then rename); a few still use plain `writeFile` — see each section. There is no schema-versioning/migration framework — schemas rely on optional fields with defaults for forward compatibility, with a small amount of inline normalization in `persisted-config.ts` for legacy provider/speech entries.
+
+All server-side stores live under `$PASEO_HOME` (defaults to `~/.paseo`).
+
+---
+
+## Directory layout
+
+```
+$PASEO_HOME/
+├── config.json                          # Daemon configuration
+├── server-id                            # Stable daemon identifier (plain text, "srv_<base64url>")
+├── daemon-keypair.json                  # E2EE keypair for relay (mode 0600)
+├── paseo.pid                            # Daemon PID lock file
+├── daemon.log                           # Default log file (path configurable)
+├── agents/
+│   └── {sanitized-cwd}/
+│       └── {agentId}.json               # One file per agent
+├── schedules/
+│   └── {scheduleId}.json                # One file per schedule
+├── chat/
+│   └── rooms.json                       # All rooms + messages
+├── loops/
+│   └── loops.json                       # All loop records
+├── projects/
+│   ├── projects.json                    # Project registry
+│   └── workspaces.json                  # Workspace registry
+└── push-tokens.json                     # Expo push notification tokens
+```
+
+The `agents/{sanitized-cwd}/` directory name is derived from the agent's `cwd` by stripping the filesystem root and replacing path separators with `-` (Windows drive letters become a `C-` style prefix). Atomic writes (temp file + rename): agent records, chat, project/workspace registries, push tokens. Non-atomic (plain `writeFile`): `config.json`, `schedules/*.json`, `loops/loops.json`, `server-id`, `daemon-keypair.json`.
+
+---
+
+## 1. Agent Record
+
+**Path:** `$PASEO_HOME/agents/{project-dir}/{agentId}.json`
+
+Each agent is stored as a separate JSON file, grouped by project directory.
+
+| Field                | Type                                     | Description                                                                                                                                                               |
+| -------------------- | ---------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `id`                 | `string`                                 | UUID, primary key                                                                                                                                                         |
+| `provider`           | `string`                                 | Agent provider (`"claude"`, `"codex"`, `"opencode"`, etc.)                                                                                                                |
+| `cwd`                | `string`                                 | Working directory the agent operates in                                                                                                                                   |
+| `createdAt`          | `string` (ISO 8601)                      | Creation timestamp                                                                                                                                                        |
+| `updatedAt`          | `string` (ISO 8601)                      | Last update timestamp                                                                                                                                                     |
+| `lastActivityAt`     | `string?` (ISO 8601)                     | Last activity timestamp                                                                                                                                                   |
+| `lastUserMessageAt`  | `string?` (ISO 8601)                     | Last user message timestamp                                                                                                                                               |
+| `title`              | `string?`                                | User-visible title                                                                                                                                                        |
+| `labels`             | `Record<string, string>`                 | Key-value labels (default `{}`). `paseo.parent-agent-id` set automatically when launched via the `create_agent` MCP tool — see [agent-lifecycle.md](./agent-lifecycle.md) |
+| `lastStatus`         | `AgentStatus`                            | One of: `"initializing"`, `"idle"`, `"running"`, `"error"`, `"closed"`                                                                                                    |
+| `lastModeId`         | `string?`                                | Last active mode ID                                                                                                                                                       |
+| `config`             | `SerializableConfig?`                    | Agent session configuration (see below)                                                                                                                                   |
+| `runtimeInfo`        | `RuntimeInfo?`                           | Live runtime state (see below)                                                                                                                                            |
+| `features`           | `AgentFeature[]?`                        | Provider-reported features (toggles/selects)                                                                                                                              |
+| `persistence`        | `PersistenceHandle?`                     | Handle for resuming sessions                                                                                                                                              |
+| `lastError`          | `string?` (nullable)                     | Last error message, if any                                                                                                                                                |
+| `requiresAttention`  | `boolean?`                               | Whether the agent needs user attention                                                                                                                                    |
+| `attentionReason`    | `"finished" \| "error" \| "permission"?` | Why attention is needed                                                                                                                                                   |
+| `attentionTimestamp` | `string?` (ISO 8601)                     | When attention was flagged                                                                                                                                                |
+| `internal`           | `boolean?`                               | Whether this is a system-internal agent (loop workers, etc.)                                                                                                              |
+| `archivedAt`         | `string?` (ISO 8601)                     | Soft-delete timestamp                                                                                                                                                     |
+
+### Nested: SerializableConfig
+
+| Field              | Type                       | Description                  |
+| ------------------ | -------------------------- | ---------------------------- |
+| `title`            | `string?`                  | Configured title             |
+| `modeId`           | `string?`                  | Configured mode              |
+| `model`            | `string?`                  | Configured model             |
+| `thinkingOptionId` | `string?`                  | Thinking/reasoning level     |
+| `featureValues`    | `Record<string, unknown>?` | Feature preference overrides |
+| `extra`            | `Record<string, any>?`     | Provider-specific config     |
+| `systemPrompt`     | `string?`                  | Custom system prompt         |
+| `mcpServers`       | `Record<string, any>?`     | MCP server configurations    |
+
+### Nested: RuntimeInfo
+
+| Field              | Type                       | Description                    |
+| ------------------ | -------------------------- | ------------------------------ |
+| `provider`         | `string`                   | Active provider                |
+| `sessionId`        | `string?`                  | Active session ID              |
+| `model`            | `string?`                  | Active model                   |
+| `thinkingOptionId` | `string?`                  | Active thinking option         |
+| `modeId`           | `string?`                  | Active mode                    |
+| `extra`            | `Record<string, unknown>?` | Provider-specific runtime data |
+
+### Nested: PersistenceHandle
+
+| Field          | Type                   | Description                                                           |
+| -------------- | ---------------------- | --------------------------------------------------------------------- |
+| `provider`     | `string`               | Provider that owns the session                                        |
+| `sessionId`    | `string`               | Session ID for resumption                                             |
+| `nativeHandle` | `any?`                 | Provider-specific handle (Codex thread ID, Claude resume token, etc.) |
+| `metadata`     | `Record<string, any>?` | Extra metadata                                                        |
+
+### Nested: AgentFeature (discriminated union on `type`)
+
+**Toggle:**
+
+| Field         | Type       |
+| ------------- | ---------- |
+| `type`        | `"toggle"` |
+| `id`          | `string`   |
+| `label`       | `string`   |
+| `description` | `string?`  |
+| `tooltip`     | `string?`  |
+| `icon`        | `string?`  |
+| `value`       | `boolean`  |
+
+**Select:**
+
+| Field         | Type                  |
+| ------------- | --------------------- |
+| `type`        | `"select"`            |
+| `id`          | `string`              |
+| `label`       | `string`              |
+| `description` | `string?`             |
+| `tooltip`     | `string?`             |
+| `icon`        | `string?`             |
+| `value`       | `string \| null`      |
+| `options`     | `AgentSelectOption[]` |
+
+---
+
+## 2. Daemon Configuration
+
+**Path:** `$PASEO_HOME/config.json`
+
+Single file, validated with `PersistedConfigSchema`.
+
+```
+{
+  version: 1,
+  daemon: {
+    listen: "127.0.0.1:6767",
+    hostnames: true | string[],   // legacy alias `allowedHosts` is migrated on load
+    mcp: { enabled: boolean, injectIntoAgents: boolean },
+    cors: { allowedOrigins: string[] },
+    relay: { enabled: boolean, endpoint: string, publicEndpoint: string, useTls: boolean },
+    auth: { password: string }    // bcrypt hash, optional
+  },
+  app: {
+    baseUrl: string
+  },
+  providers: {
+    openai: { apiKey: string },
+    local: { modelsDir: string }
+  },
+  agents: {
+    // ProviderOverrideSchema; legacy entries with `command: { mode, ... }` are migrated to the
+    // current shape on load via `migrateProviderSettings`. Custom provider IDs must declare
+    // `extends` (one of the built-ins or `"acp"`) and `label`. See `provider-launch-config.ts`.
+    providers: Record<providerId, ProviderOverride>
+  },
+  features: {
+    dictation: { enabled, stt: { provider, model, confidenceThreshold } },
+    voiceMode: { enabled, llm, stt, turnDetection, tts: { provider, model, voice, speakerId, speed } }
+  },
+  log: {
+    level, format,
+    console: { level, format },
+    file: { level, path, rotate: { maxSize, maxFiles } }
+  }
+}
+```
+
+All fields are optional with sensible defaults.
+
+---
+
+## 3. Schedule
+
+**Path:** `$PASEO_HOME/schedules/{id}.json`
+
+One file per schedule. ID is 8 hex characters. Writes are direct (not atomic).
+
+| Field       | Type                                  | Description                      |
+| ----------- | ------------------------------------- | -------------------------------- |
+| `id`        | `string`                              | 8-char hex ID                    |
+| `name`      | `string?`                             | Human-readable name              |
+| `prompt`    | `string`                              | The prompt to send               |
+| `cadence`   | `ScheduleCadence`                     | Timing (see below)               |
+| `target`    | `ScheduleTarget`                      | What to run (see below)          |
+| `status`    | `"active" \| "paused" \| "completed"` | Current state                    |
+| `createdAt` | `string` (ISO 8601)                   |                                  |
+| `updatedAt` | `string` (ISO 8601)                   |                                  |
+| `nextRunAt` | `string?` (ISO 8601)                  | Next scheduled execution         |
+| `lastRunAt` | `string?` (ISO 8601)                  | Last execution time              |
+| `pausedAt`  | `string?` (ISO 8601)                  | When paused                      |
+| `expiresAt` | `string?` (ISO 8601)                  | Auto-expire time                 |
+| `maxRuns`   | `number?`                             | Max executions before completing |
+| `runs`      | `ScheduleRun[]`                       | Execution history                |
+
+### Nested: ScheduleCadence (discriminated union on `type`)
+
+- `{ type: "every", everyMs: number }` — interval in milliseconds
+- `{ type: "cron", expression: string }` — cron expression
+
+### Nested: ScheduleTarget (discriminated union on `type`)
+
+- `{ type: "agent", agentId: string }` — send to existing agent
+- `{ type: "new-agent", config: { provider, cwd, modeId?, model?, thinkingOptionId?, title?, approvalPolicy?, sandboxMode?, networkAccess?, webSearch?, extra?, systemPrompt?, mcpServers? } }` — create a new agent
+
+### Nested: ScheduleRun
+
+| Field          | Type                                   | Description             |
+| -------------- | -------------------------------------- | ----------------------- |
+| `id`           | `string`                               | Run ID                  |
+| `scheduledFor` | `string` (ISO 8601)                    | Intended execution time |
+| `startedAt`    | `string` (ISO 8601)                    |                         |
+| `endedAt`      | `string?` (ISO 8601)                   |                         |
+| `status`       | `"running" \| "succeeded" \| "failed"` |                         |
+| `agentId`      | `string?` (UUID)                       | Agent used for this run |
+| `output`       | `string?`                              | Agent output text       |
+| `error`        | `string?`                              | Error message if failed |
+
+---
+
+## 4. Chat
+
+**Path:** `$PASEO_HOME/chat/rooms.json`
+
+Single file containing all rooms and messages.
+
+```json
+{
+  "rooms": [ ... ],
+  "messages": [ ... ]
+}
+```
+
+### ChatRoom
+
+| Field       | Type                | Description                         |
+| ----------- | ------------------- | ----------------------------------- |
+| `id`        | `string` (UUID)     |                                     |
+| `name`      | `string`            | Unique room name (case-insensitive) |
+| `purpose`   | `string?`           | Room description                    |
+| `createdAt` | `string` (ISO 8601) |                                     |
+| `updatedAt` | `string` (ISO 8601) | Updated on each new message         |
+
+### ChatMessage
+
+| Field              | Type                | Description                         |
+| ------------------ | ------------------- | ----------------------------------- |
+| `id`               | `string` (UUID)     |                                     |
+| `roomId`           | `string`            | FK to ChatRoom.id                   |
+| `authorAgentId`    | `string`            | Agent ID of the author              |
+| `body`             | `string`            | Message text (supports `@mentions`) |
+| `replyToMessageId` | `string?`           | FK to another ChatMessage.id        |
+| `mentionAgentIds`  | `string[]`          | Extracted `@mention` agent IDs      |
+| `createdAt`        | `string` (ISO 8601) |                                     |
+
+---
+
+## 5. Loop
+
+**Path:** `$PASEO_HOME/loops/loops.json`
+
+Single file containing an array of all loop records. Writes are direct (not atomic) and serialized through an in-memory queue. On daemon startup any record with `status: "running"` is recovered as `"stopped"` with an interruption log entry.
+
+| Field                   | Type                                                | Description                                |
+| ----------------------- | --------------------------------------------------- | ------------------------------------------ |
+| `id`                    | `string`                                            | 8-char UUID prefix                         |
+| `name`                  | `string?`                                           | Human-readable name                        |
+| `prompt`                | `string`                                            | Worker prompt                              |
+| `cwd`                   | `string`                                            | Working directory                          |
+| `provider`              | `string`                                            | Default provider                           |
+| `model`                 | `string?`                                           | Default model                              |
+| `modeId`                | `string?`                                           | Default mode ID                            |
+| `workerProvider`        | `string?`                                           | Override provider for workers              |
+| `workerModel`           | `string?`                                           | Override model for workers                 |
+| `verifierProvider`      | `string?`                                           | Override provider for verifiers            |
+| `verifierModel`         | `string?`                                           | Override model for verifiers               |
+| `verifierModeId`        | `string?`                                           | Override mode ID for verifiers             |
+| `verifyPrompt`          | `string?`                                           | LLM verification prompt                    |
+| `verifyChecks`          | `string[]`                                          | Shell commands to run as checks            |
+| `archive`               | `boolean`                                           | Whether to archive worker agents after use |
+| `sleepMs`               | `number`                                            | Delay between iterations (ms)              |
+| `maxIterations`         | `number?`                                           | Cap on iterations                          |
+| `maxTimeMs`             | `number?`                                           | Total time budget (ms)                     |
+| `status`                | `"running" \| "succeeded" \| "failed" \| "stopped"` |                                            |
+| `createdAt`             | `string` (ISO 8601)                                 |                                            |
+| `updatedAt`             | `string` (ISO 8601)                                 |                                            |
+| `startedAt`             | `string` (ISO 8601)                                 |                                            |
+| `completedAt`           | `string?` (ISO 8601)                                |                                            |
+| `stopRequestedAt`       | `string?` (ISO 8601)                                |                                            |
+| `iterations`            | `LoopIteration[]`                                   |                                            |
+| `logs`                  | `LoopLogEntry[]`                                    |                                            |
+| `nextLogSeq`            | `number`                                            | Monotonic log sequence counter             |
+| `activeIteration`       | `number?`                                           | Currently executing iteration index        |
+| `activeWorkerAgentId`   | `string?`                                           | Currently running worker agent             |
+| `activeVerifierAgentId` | `string?`                                           | Currently running verifier agent           |
+
+### Nested: LoopIteration
+
+| Field               | Type                                                | Description              |
+| ------------------- | --------------------------------------------------- | ------------------------ |
+| `index`             | `number`                                            | 1-based iteration index  |
+| `workerAgentId`     | `string?`                                           | Agent ID of the worker   |
+| `workerStartedAt`   | `string` (ISO 8601)                                 |                          |
+| `workerCompletedAt` | `string?` (ISO 8601)                                |                          |
+| `verifierAgentId`   | `string?`                                           | Agent ID of the verifier |
+| `status`            | `"running" \| "succeeded" \| "failed" \| "stopped"` |                          |
+| `workerOutcome`     | `"completed" \| "failed" \| "canceled"?`            |                          |
+| `failureReason`     | `string?`                                           |                          |
+| `verifyChecks`      | `LoopVerifyCheckResult[]`                           | Shell check results      |
+| `verifyPrompt`      | `LoopVerifyPromptResult?`                           | LLM verification result  |
+
+### Nested: LoopLogEntry
+
+| Field       | Type                                                 |
+| ----------- | ---------------------------------------------------- |
+| `seq`       | `number` (monotonic)                                 |
+| `timestamp` | `string` (ISO 8601)                                  |
+| `iteration` | `number?`                                            |
+| `source`    | `"loop" \| "worker" \| "verifier" \| "verify-check"` |
+| `level`     | `"info" \| "error"`                                  |
+| `text`      | `string`                                             |
+
+### Nested: LoopVerifyCheckResult
+
+| Field         | Type                |
+| ------------- | ------------------- |
+| `command`     | `string`            |
+| `exitCode`    | `number`            |
+| `passed`      | `boolean`           |
+| `stdout`      | `string`            |
+| `stderr`      | `string`            |
+| `startedAt`   | `string` (ISO 8601) |
+| `completedAt` | `string` (ISO 8601) |
+
+### Nested: LoopVerifyPromptResult
+
+| Field             | Type                |
+| ----------------- | ------------------- |
+| `passed`          | `boolean`           |
+| `reason`          | `string`            |
+| `verifierAgentId` | `string?`           |
+| `startedAt`       | `string` (ISO 8601) |
+| `completedAt`     | `string` (ISO 8601) |
+
+---
+
+## 6. Project Registry
+
+**Path:** `$PASEO_HOME/projects/projects.json`
+
+Array of project records.
+
+| Field         | Type                        | Description                              |
+| ------------- | --------------------------- | ---------------------------------------- |
+| `projectId`   | `string`                    | Primary key                              |
+| `rootPath`    | `string`                    | Filesystem root of the project           |
+| `kind`        | `"git" \| "non_git"`        |                                          |
+| `displayName` | `string`                    |                                          |
+| `createdAt`   | `string` (ISO 8601)         |                                          |
+| `updatedAt`   | `string` (ISO 8601)         |                                          |
+| `archivedAt`  | `string \| null` (ISO 8601) | Soft-delete timestamp; required nullable |
+
+---
+
+## 7. Workspace Registry
+
+**Path:** `$PASEO_HOME/projects/workspaces.json`
+
+Array of workspace records. A workspace is a specific working directory within a project.
+
+| Field         | Type                                            | Description                    |
+| ------------- | ----------------------------------------------- | ------------------------------ |
+| `workspaceId` | `string`                                        | Primary key                    |
+| `projectId`   | `string`                                        | FK to Project.projectId        |
+| `cwd`         | `string`                                        | Filesystem path                |
+| `kind`        | `"local_checkout" \| "worktree" \| "directory"` |                                |
+| `displayName` | `string`                                        |                                |
+| `createdAt`   | `string` (ISO 8601)                             |                                |
+| `updatedAt`   | `string` (ISO 8601)                             |                                |
+| `archivedAt`  | `string \| null` (ISO 8601)                     | Soft-delete; required nullable |
+
+---
+
+## 8. Push Token Store
+
+**Path:** `$PASEO_HOME/push-tokens.json`
+
+```json
+{
+  "tokens": ["ExponentPushToken[...]", ...]
+}
+```
+
+Simple set of Expo push notification tokens. Loaded with permissive parsing (filters non-string entries). Persisted with atomic temp-file rename.
+
+---
+
+## 9. Daemon meta files
+
+These small files are not validated as full Zod schemas but are persisted under `$PASEO_HOME` for daemon identity and runtime coordination.
+
+| Path                  | Format                                                         | Notes                                                                             |
+| --------------------- | -------------------------------------------------------------- | --------------------------------------------------------------------------------- |
+| `server-id`           | Plain text, e.g. `srv_<base64url>`                             | Stable per-`$PASEO_HOME` daemon ID. Overridable via `PASEO_SERVER_ID` env.        |
+| `daemon-keypair.json` | `{ v: 2, publicKeyB64, secretKeyB64 }` (libsodium box keypair) | E2EE relay identity. Written with mode `0600`. Regenerated if file is unreadable. |
+| `paseo.pid`           | JSON `{ pid, startedAt, ... }`                                 | PID lock; prevents two daemons sharing one `$PASEO_HOME`.                         |
+| `daemon.log`          | Pino log output                                                | Default location; path/rotation configurable via `log.file` in `config.json`.     |
+
+---
+
+## Client-side stores (App)
+
+These live in React Native `AsyncStorage` or browser `IndexedDB`, not on the daemon filesystem.
+
+### Draft Store
+
+**AsyncStorage key:** `paseo-drafts` (version 2)
+
+```typescript
+{
+  drafts: Record<draftKey, {
+    input: { text: string, images: AttachmentMetadata[] },
+    lifecycle: "active" | "abandoned" | "sent",
+    updatedAt: number,     // epoch ms
+    version: number        // optimistic concurrency
+  }>,
+  createModalDraft: DraftRecord | null
+}
+```
+
+### Attachment Store (Web)
+
+**IndexedDB database:** `paseo-attachment-bytes`, object store: `attachments`
+
+Stores binary attachment blobs keyed by attachment ID.
+
+### AttachmentMetadata
+
+| Field         | Type      | Description                    |
+| ------------- | --------- | ------------------------------ |
+| `id`          | `string`  | Unique attachment ID           |
+| `mimeType`    | `string`  | MIME type                      |
+| `storageType` | `string`  | Storage backend identifier     |
+| `storageKey`  | `string`  | Key within the storage backend |
+| `createdAt`   | `number`  | Epoch ms                       |
+| `fileName`    | `string?` | Original filename              |
+| `byteSize`    | `number?` | Size in bytes                  |

docs/design.md

@@ -0,0 +1,247 @@
+# Design
+
+Tokens — every color, font size, weight, spacing step, radius, icon size — live in `packages/app/src/styles/theme.ts`.
+
+---
+
+## 1. Character
+
+Paseo is minimal, spacious, quiet, confident. Whitespace is deliberate. Nothing crowds, nothing decorates, nothing apologizes. A row, a label, a control. That is the bar.
+
+The app is calm so the user's work is not. Every visual decision serves either _act on this_ or _understand this_ — never _look at this_.
+
+Consistency comes from component reuse, not from hand-matching styles across surfaces. A row in the projects list, a row in settings, and a row in a modal are the same component, not three implementations that happen to look alike. When two surfaces do the same semantic thing in two different ways, one of them is wrong.
+
+---
+
+## 2. Component reuse
+
+A semantic element used in three or more places is a primitive. One of a kind is a screen.
+
+Primitives live in `packages/app/src/components/ui/` and `packages/app/src/components/headers/`. Card and row layout live in `packages/app/src/styles/settings.ts`. Section structure lives in `packages/app/src/screens/settings/settings-section.tsx`.
+
+A pressable styled to look like a button is wrong; the button is `<Button>` (`packages/app/src/components/ui/button.tsx`). A bare `<Text>` styled to look like a section header is wrong; the section header is `<SettingsSection>` (`packages/app/src/screens/settings/settings-section.tsx`). A custom `Modal` for a confirmation is wrong; the confirmation is `confirmDialog` (`packages/app/src/utils/confirm-dialog.ts`). A hand-rolled overflow menu is wrong; the menu is `<DropdownMenu>` (`packages/app/src/components/ui/dropdown-menu.tsx`). A hand-rolled status pill is wrong; the pill is `<StatusBadge>` (`packages/app/src/components/ui/status-badge.tsx`).
+
+Before adding a new component, read `components/ui/`. The primitive usually exists.
+
+---
+
+## 3. Hierarchy
+
+Hierarchy is conveyed through weight and color, not size. Most labels, titles, and hints across the app are `fontSize.base` or `fontSize.xs`. The distinction between a row's primary line and its secondary line is `foreground` versus `foregroundMuted`.
+
+Weight has three tiers, applied by role:
+
+- **Screen titles** — the title at the top of a screen — use `<ScreenTitle>` (`packages/app/src/components/headers/screen-title.tsx`), which renders `fontSize.base` at weight `400` on compact and `300` on desktop. Top-of-screen titles are lighter on desktop, not heavier. The workspace screen header follows the same rule (`packages/app/src/screens/workspace/workspace-screen.tsx`).
+- **Structural labels** use `fontWeight.medium`. This applies to section labels above a stack of rows (`packages/app/src/components/agent-list.tsx:519-523`, `packages/app/src/components/keyboard-shortcuts-dialog.tsx:63-67`), form field labels above an input inside a modal (`packages/app/src/components/add-host-modal.tsx:19-23`, `packages/app/src/components/pair-link-modal.tsx:24-28`), the title at the top of a modal/sheet/dialog (`packages/app/src/components/adaptive-modal-sheet.tsx:90-94`, `packages/app/src/components/ui/combobox.tsx:1607-1611`, `packages/app/src/components/welcome-screen.tsx:48-53`), action button labels in tight components such as the sidebar callout actions (`packages/app/src/components/sidebar-callout.tsx:218-221`), and inline data emphasis on dense metadata rows (`packages/app/src/components/git-diff-pane.tsx:2322-2327`, `packages/app/src/components/file-explorer-pane.tsx:1115-1122`).
+- **Content** uses `fontWeight.normal`. This applies to settings rows (`packages/app/src/styles/settings.ts`), sidebar primary list-item titles (`packages/app/src/components/sidebar-workspace-list.tsx:2680-2686`, `packages/app/src/components/agent-list.tsx:572-578`), `<Button>` text (`packages/app/src/components/ui/button.tsx:80-84`), `<StatusBadge>` text (`packages/app/src/components/ui/status-badge.tsx:56-60`), and `<SidebarCallout>` titles (`packages/app/src/components/sidebar-callout.tsx:175-180`).
+
+The rule, condensed: text that _names_ a surface or a group is `medium`. Text that lives _inside_ a surface or a group is `normal`. Top-of-screen titles are `<ScreenTitle>`, which is lighter still.
+
+Foreground is for the thing being acted on: row titles, section headings, the selected sidebar item. `foregroundMuted` is for context: hints, descriptions, secondary metadata, idle sidebar items, placeholders, status text.
+
+Accent is the one CTA per surface. A `<Button variant="default">` filled with `accent` appears at most once on a page. Most pages have zero — settings is mostly toggles and text, the workspace pane is mostly content, the chat composer is the input itself.
+
+Destructive is a color, not a click. Restart-daemon and remove-host are `<Button variant="outline">` in the row trailing slot; the destructive surface only appears inside the `confirmDialog` (`packages/app/src/screens/settings/host-page.tsx:541-547`). Workspace archive opens a confirm dialog before any red appears (`packages/app/src/components/sidebar-workspace-list.tsx`). Red appears after the user has indicated intent.
+
+---
+
+## 4. Buttons
+
+The button is `<Button>` (`packages/app/src/components/ui/button.tsx`). It has five variants. Each has one job.
+
+`default` is the one primary action on a surface — filled with `accent`. At most one per page. The primary slot inside an `<AdaptiveModalSheet>` and the highlighted action on the welcome screen are the canonical uses.
+
+`secondary` is the paired action when two actions carry equal weight — filled with `surface3`. The component default is `secondary`, which matches its frequency in the codebase.
+
+`outline` is the low-frequency action that lives on a row — transparent with `borderAccent`. Restart, Remove, Update on host detail (`packages/app/src/screens/settings/host-page.tsx:585-594`).
+
+`ghost` is structural and non-committal — no border, no fill. Back arrows, header toggles, "Load more" footers (`packages/app/src/screens/sessions-screen.tsx:54-63`), more-affordances. Ghost is used when the affordance is part of the chrome, not a decision.
+
+`destructive` is filled with `destructive`. It only appears inside a confirm. The button on the page is `outline`; the destructive button is the confirm button inside the dialog.
+
+Sizes: `xs` for ultra-tight inline triggers. `sm` for any button sitting in a row. `md` is the page default. `lg` is reserved for large standalone CTAs.
+
+A `<Pressable>` wrapping a `<Text>` is a sixth variant. It is wrong. `<Button>` accepts `style`, `textStyle`, `leftIcon`, `disabled`, `size`, and `variant`.
+
+---
+
+## 5. Borders
+
+Borders group, separate, or rarely emphasize.
+
+A logical block of related rows lives inside a card — one border around the whole group. The card primitive is `settingsStyles.card`; the keyboard-shortcuts dialog uses the same shape inline (`packages/app/src/components/keyboard-shortcuts-dialog.tsx:68-73`). The border defines what belongs together.
+
+Rows after the first inside a card carry `settingsStyles.rowBorder` — a single top border. The first row never has one. The same divider pattern appears in the keyboard-shortcuts dialog rows (`packages/app/src/components/keyboard-shortcuts-dialog.tsx:74-83`). Rows do not need their own background to feel separated.
+
+A list that is itself the page content — sidebar items in `sidebar-workspace-list.tsx`, the workspace list, the agent list (`packages/app/src/components/agent-list.tsx`) — uses spacing and surface, not borders, to separate items. Rows-in-a-card is an interior pattern; lists-as-pages are not.
+
+Pane chrome — the workspace pane header, the file-explorer header, the diff pane header — uses a single bottom border to separate the header from the content (`packages/app/src/components/git-diff-pane.tsx:2328-2331`). One border, no shadow.
+
+`borderAccent` is reserved for the outline button. Inputs use `border`. Single-thing borders are wrong; a single bordered element is either a card with one row (use the card) or it does not need a border.
+
+---
+
+## 6. Pickers
+
+Five primitives. The pick is determined by option count, the need to search, and how the picker is anchored.
+
+`<DropdownMenu>` is for a small fixed set anchored to a trigger. Theme picker, kebab menus on workspace and project rows (`packages/app/src/components/sidebar-workspace-list.tsx:684-770`), row "more" menus. Items can be async (`status: "pending"`) and can include destructive entries. Under ~10 options where the user knows what they're looking for.
+
+`<Combobox>` is for a large or searchable list. Host switcher in the sidebar footer, model selector in the composer, branch switcher in the workspace header (`packages/app/src/components/branch-switcher.tsx`). The user types to find the option, or the list is long enough to scroll.
+
+`<ContextMenu>` is for right-click and long-press on a target. The row is the trigger; there is no visible affordance. Used for incidental actions on workspace rows in the sidebar (`packages/app/src/components/sidebar-workspace-list.tsx`).
+
+`<AdaptiveModalSheet>` is for a focused task. Multi-field forms (`packages/app/src/components/add-host-modal.tsx`, `packages/app/src/components/pair-link-modal.tsx`, `packages/app/src/components/project-picker-modal.tsx`), confirmations with detail, anything that earns a backdrop. Bottom sheet on compact, centered card on desktop. Raw `Modal` is wrong for any of these.
+
+`confirmDialog` is for destructive yes/no and imperative confirmation. Promise-based: `await confirmDialog({ destructive: true, ... })`. Anything where a wrong click loses work.
+
+Three themes is `DropdownMenu`. Thirty hosts is `Combobox`. A label and a value is `AdaptiveModalSheet`. "Are you sure?" is `confirmDialog`.
+
+---
+
+## 7. Density and rhythm
+
+Settings detail pages, the projects detail page, and any list+detail content sit inside a centered, max-width 720 column (`packages/app/src/screens/settings-screen.tsx`, `packages/app/src/screens/projects-screen.tsx`). Lines stay readable, the eye does not have to track wide horizontal distances. Form modals carry their own narrower content frame (`packages/app/src/components/add-host-modal.tsx`).
+
+Workspace and chat surfaces use the full width — these are working surfaces, not reading surfaces. The composer carries `MAX_CONTENT_WIDTH` from `packages/app/src/constants/layout.ts` to keep lines readable while letting the workspace pane fill the rest.
+
+Sections sit apart. `<SettingsSection>` owns its own bottom margin; the next thing is wrapped in another `<SettingsSection>`. The agent-list `sectionHeading` carries the same `marginTop`/`marginBottom` rhythm (`packages/app/src/components/agent-list.tsx:511-517`). Adding `marginBottom` to a section is wrong.
+
+Cards inside a section sit closer than sections. Rows inside a card touch — only the divider separates them. The rhythm is page → spacious; section → spacious; card → tight.
+
+Rows have generous vertical padding: roughly 16px of content plus 16px of vertical padding for settings rows, 8–12px for sidebar list items where many rows must fit. Compressing rows below the established density to fit more on the screen is wrong. Too many rows means more cards or more sections, not smaller rows.
+
+The whitespace is the design.
+
+---
+
+## 8. Responsiveness
+
+Compact-first. The small case is designed; the large case adds chrome around it.
+
+The list+detail pattern is canonical and reused across surfaces. The settings shell (`packages/app/src/screens/settings-screen.tsx`) and the projects screen (`packages/app/src/screens/projects-screen.tsx`) implement it identically:
+
+- On compact: full-screen list with `<BackHeader>` at the top. Tapping a row pushes a full-screen detail with its own `<BackHeader>` that returns to the list.
+- On desktop: a 320px sidebar on the left holds the list with `surfaceSidebar` background. The content pane on the right holds the selected detail with `<ScreenHeader>`, `<HeaderIconBadge>`, and `<ScreenTitle>`.
+
+The branching is one `useIsCompactFormFactor()` check at the top of the screen component. The list and the detail are the same components in both layouts; only the framing changes.
+
+The workspace screen (`packages/app/src/screens/workspace/workspace-screen.tsx`) follows a different but parallel rule: tabs collapse on compact, panes split on desktop. The sidebar (`packages/app/src/components/left-sidebar.tsx`) is overlaid on compact and pinned on desktop.
+
+A new list+detail feature copies the settings shell. A new workspace-shaped feature copies the workspace shell. Inventing a third shape happens in design review, not in a PR.
+
+---
+
+## 9. Copy and voice
+
+Sentence case. "Pair a device", "Danger zone", "Restart daemon", "Inject Paseo tools", "No sessions yet", "Load more". Proper nouns retain casing — Paseo, Beta, Stable, Local. Title case is wrong.
+
+No trailing periods on row titles, labels, or buttons. No trailing period on a single-clause hint: "What happens when you press Enter while the agent is running" (`packages/app/src/screens/settings-screen.tsx:271-272`). Periods exist inside multi-sentence prose: "Restarts the daemon process. The app will reconnect automatically."
+
+Empty-state strings are short noun phrases or short sentences: "No projects yet", "Select a project", "No sessions yet" (`packages/app/src/screens/sessions-screen.tsx:74-76`), "Host not found".
+
+Buttons are imperative: Save, Cancel, Restart, Remove, Update, Install update, Add host, Load more. In-flight labels are present-participle with a literal three-dot ellipsis: "Saving...", "Restarting...", "Removing...", "Loading...".
+
+Error copy is direct. "Unable to remove host" (`packages/app/src/screens/settings/host-page.tsx:697`), not "Sorry, we couldn't remove the host." Recovery instructions are concrete: "Wait for it to come online before restarting." Errors describe state; they do not editorialize.
+
+Terminology:
+
+- Workspace, never "checkout".
+- Host, except where the user-facing concept is the daemon process itself ("Restart daemon").
+- Project, not "repo" or "repository".
+- Provider, not "model provider".
+- Session and agent are distinct: a session is a historical entry in `sessions-screen.tsx`; an agent is a live entity in the workspace.
+
+---
+
+## 10. States
+
+Loading is inline by default. `<LoadingSpinner size={14} color={foregroundMuted} />` sits next to the thing it relates to (`packages/app/src/screens/settings/providers-section.tsx:227-231`). Page-level loading is a centered `<LoadingSpinner size="large">` (`packages/app/src/screens/sessions-screen.tsx:69-72`). Card-level loading is a single short line, not a spinner. In-row dropdown items use `<DropdownMenuItem status="pending" pendingLabel="Removing...">`; the menu item handles its own pending state.
+
+Empty states are short noun phrases. Centered, muted, one or two lines. Sessions screen pairs the empty noun with a single ghost button to navigate back (`packages/app/src/screens/sessions-screen.tsx:74-81`); that pairing is the maximum elaboration. Illustrations and CTAs disguised as empty states are wrong.
+
+Inline errors are a single sentence in `palette.red[300]` `xs`, sitting under the field or inside the card it relates to (`packages/app/src/screens/settings/providers-section.tsx:115-119`).
+
+Page-level alerts — informational notices, success confirmations, warnings, or recoverable errors that need a small visible block on the page — use `<Alert>` (`packages/app/src/components/ui/alert.tsx`). Variants: `default`, `info`, `success`, `warning`, `error`. The chrome is quiet by design: a 1px tinted border, transparent background, a small variant-tinted icon, the title in the variant accent, the description in `foregroundMuted`. Actions go in the `children` slot as `<Button variant="outline" size="sm">` — recovery actions are low-frequency and outline keeps them quiet alongside the alert's accent (`packages/app/src/screens/project-settings-screen.tsx`). One `<Alert>` at a time per region.
+
+Sidebar callouts — cross-cutting alerts that apply across the whole app, like worktree setup, Rosetta install, and desktop update available — register through `useSidebarCallouts()` and render in the left sidebar via `<SidebarCallout>` (`packages/app/src/components/sidebar-callout.tsx`). The chrome (top-border-only, full-width action buttons) is tuned for that ~280px column. Canonical sources: `packages/app/src/components/worktree-setup-callout-source.tsx`, `packages/app/src/desktop/updates/rosetta-callout-source.tsx`, `packages/app/src/desktop/updates/update-callout-source.tsx`. Never import `<SidebarCallout>` into a page — that's what `<Alert>` is for.
+
+Imperative errors are `Alert.alert("Error", "Unable to ...")` (the React Native `Alert` API, not this component) for failures that interrupt the flow and have no place on the page.
+
+Disabled state is `opacity: theme.opacity[50]` on the outer pressable. Color changes for disabled state are wrong; a disabled button is the same button, dimmer.
+
+Partial failure (a list mostly fine but one source errored) is a bordered banner above the list, listing each failure in red-300 `xs` (`packages/app/src/screens/projects-screen.tsx:151-159`). The list still renders.
+
+State surfaces at the smallest scope it affects. Field error stays under the field; page error is a banner; flow-stopping error is an `Alert`.
+
+---
+
+## 11. List rows
+
+The row anatomy is a content column with an optional trailing slot. Inside a card the row is `settingsStyles.row`. Inside a sidebar list the row carries its own padding and `borderRadius.lg` per item (`packages/app/src/components/sidebar-workspace-list.tsx:2614-2625`).
+
+Rows that drill into a detail lead with a chevron in the trailing slot (`ChevronRight`, `iconSize.sm`, `foregroundMuted`). The whole row is the `<Pressable>`. Pair-device row (`packages/app/src/screens/settings/host-page.tsx:644-668`), provider row (`packages/app/src/screens/settings/providers-section.tsx:92-132`), project row in the projects list. Chevron means navigation.
+
+Kebab menus (`<DropdownMenu>` with `<MoreVertical size={14} />` trigger) are for actions on the row, not navigation. Trigger style: `padding: 2`, `borderRadius: 4`, hover background `surface2`. Menu position: `align="end"`. Items use `<DropdownMenuItem leading={<Icon size={14} color={foregroundMuted} />} ...>`. Visibility is `isHovered || isTouchPlatform` — hover-revealed on web, always visible on native (`packages/app/src/components/sidebar-workspace-list.tsx:684-770`).
+
+A row may carry both a chevron and a kebab when both navigation and row-level actions apply. Chevron sits at the end; kebab sits before it.
+
+Switches and segmented controls also sit in the trailing slot. A row that both navigates and toggles is a `<Pressable>` with a `<Switch>` in the trailing slot — the switch calls `event.stopPropagation()` so the row press does not fire (`packages/app/src/screens/settings/providers-section.tsx:92-132`). Sidebar items that hold a status dot, a count, and a kebab follow the same rule (`packages/app/src/components/sidebar-workspace-list.tsx`).
+
+Selected state on rows in a desktop list+detail uses `surfaceSidebarHover` as the background (`packages/app/src/screens/projects-screen.tsx`). Selected state on rows in the sidebar list uses `surface2` (`packages/app/src/components/agent-list.tsx:563-571`).
+
+---
+
+## 12. Status pills and badges
+
+Status pills are `palette.<color>[300]` foreground on a 10%-alpha background of the same color. Success uses green, warning uses amber, danger uses red, muted uses zinc. The `<StatusBadge>` primitive (`packages/app/src/components/ui/status-badge.tsx`) is canonical.
+
+Status dots — the small filled circles next to a host or agent name — are `borderRadius.full` filled with the status color (`statusSuccess`, `statusWarning`, `statusDanger`, or `foregroundMuted`). They sit in the trailing slot of a sidebar row or as a leading marker on a status pill.
+
+The bespoke pills in `packages/app/src/screens/settings/host-page.tsx:97-116`, `packages/app/src/components/agent-list.tsx:607-632`, and `packages/app/src/components/sidebar-workspace-list.tsx:2889-2894` are drift to be removed. New code uses `<StatusBadge>`.
+
+---
+
+## 13. Forbidden
+
+- `fontWeight.medium` on row titles, body text, button labels, badge text, or `<SidebarCallout>` titles. Medium is reserved for the structural-label tier described in §3 — section labels, modal/sheet titles, dense metadata emphasis, and tight action labels. Anything else is `normal`. `<ScreenTitle>` is responsive `400/300` and is never overridden.
+- `<Pressable>` wrapping `<Text>` to make a button. `<Button>` exists.
+- Bare `<Text>` for a section header inside settings. `<SettingsSection>` exists.
+- A "Settings" CTA on a detail page. Detail pages are settings; settings is reached from the sidebar, the host entry, or a row's kebab menu.
+- The word "checkout" in UI strings or identifiers. The term is "workspace".
+- New color tokens or hardcoded hex outside the palette. Status pill rgba backgrounds are the documented pattern (§12), not a license.
+- Placeholder text dimmed beyond `foregroundMuted`. No extra opacity, no italics, no ghost-text.
+- `onPointerEnter` and `onPointerLeave`. They do not fire on native iOS. Hover uses Pressable's `onHoverIn`/`onHoverOut` gated with `isHovered || isCompact || isNative`.
+- Raw DOM APIs without an `isWeb` guard.
+- Spacing values outside the scale. `padding: 20` and `gap: 10` are wrong.
+- Color changes for disabled state. Opacity only.
+- Destructive actions without `confirmDialog`. Restart, remove, archive, and any future destructive action are confirmed.
+- Bespoke status pills. `<StatusBadge>` is the pill primitive.
+- Raw `Modal` for a focused task. `<AdaptiveModalSheet>` is the modal primitive.
+- Importing `ActivityIndicator` directly. `<LoadingSpinner>` is the loading primitive.
+
+---
+
+## 14. Canonical surfaces by pattern
+
+| Pattern                                             | Reference                                                                                                                                                                                                                                                                                                |
+| --------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| List+detail (compact stack, desktop sidebar+pane)   | `packages/app/src/screens/settings-screen.tsx`, `packages/app/src/screens/projects-screen.tsx`                                                                                                                                                                                                           |
+| Detail card+row                                     | `packages/app/src/screens/settings/host-page.tsx`, `packages/app/src/screens/settings/providers-section.tsx`                                                                                                                                                                                             |
+| Section grouping inside a card list                 | `packages/app/src/screens/settings/settings-section.tsx`                                                                                                                                                                                                                                                 |
+| Form modal (label + input fields, primary + cancel) | `packages/app/src/components/add-host-modal.tsx`, `packages/app/src/components/pair-link-modal.tsx`, `packages/app/src/components/project-picker-modal.tsx`                                                                                                                                              |
+| Destructive confirmation                            | `confirmDialog` invoked from `packages/app/src/screens/settings/host-page.tsx:541-547`                                                                                                                                                                                                                   |
+| Centered hero / first-run                           | `packages/app/src/components/welcome-screen.tsx`                                                                                                                                                                                                                                                         |
+| Sidebar list (workspaces, hosts)                    | `packages/app/src/components/sidebar-workspace-list.tsx`, `packages/app/src/components/left-sidebar.tsx`                                                                                                                                                                                                 |
+| Live list of items with sections (agents)           | `packages/app/src/components/agent-list.tsx`                                                                                                                                                                                                                                                             |
+| Historical list (sessions)                          | `packages/app/src/screens/sessions-screen.tsx`                                                                                                                                                                                                                                                           |
+| Workspace pane (multi-tab, split)                   | `packages/app/src/screens/workspace/workspace-screen.tsx`                                                                                                                                                                                                                                                |
+| Composer / message input                            | `packages/app/src/components/composer.tsx`, `packages/app/src/components/message-input.tsx`                                                                                                                                                                                                              |
+| Pane chrome with single bottom border               | `packages/app/src/components/git-diff-pane.tsx`, `packages/app/src/components/file-explorer-pane.tsx`, `packages/app/src/components/terminal-pane.tsx`                                                                                                                                                   |
+| Page-level alert (info / success / warning / error) | `packages/app/src/components/ui/alert.tsx`, `packages/app/src/screens/project-settings-screen.tsx`                                                                                                                                                                                                       |
+| Sidebar callout (cross-cutting alert)               | `packages/app/src/components/sidebar-callout.tsx`, `packages/app/src/contexts/sidebar-callout-context.tsx`, `packages/app/src/components/worktree-setup-callout-source.tsx`, `packages/app/src/desktop/updates/rosetta-callout-source.tsx`, `packages/app/src/desktop/updates/update-callout-source.tsx` |
+| Searchable picker                                   | `packages/app/src/components/ui/combobox.tsx`, `packages/app/src/components/branch-switcher.tsx`                                                                                                                                                                                                         |
+| Trigger-anchored menu                               | `packages/app/src/components/ui/dropdown-menu.tsx` (used in `sidebar-workspace-list.tsx`, theme picker)                                                                                                                                                                                                  |
+| Right-click / long-press menu                       | `packages/app/src/components/ui/context-menu.tsx` (used in `sidebar-workspace-list.tsx`)                                                                                                                                                                                                                 |
+| Headers (back, screen, menu)                        | `packages/app/src/components/headers/back-header.tsx`, `screen-header.tsx`, `menu-header.tsx`                                                                                                                                                                                                            |

docs/development.md

@@ -0,0 +1,175 @@
+# Development
+
+## Prerequisites
+
+- Node.js (see `.tool-versions` for exact version)
+- npm workspaces (comes with Node)
+
+## Running the dev server
+
+```bash
+npm run dev
+```
+
+`scripts/dev.sh` runs the daemon and Expo together via `concurrently`, fronted by [`portless`](https://www.npmjs.com/package/portless) so each service is reachable at a stable name like `https://daemon.localhost` / `https://app.localhost` instead of a fixed port. The underlying TCP ports are ephemeral — never hardcode them. (Windows uses `scripts/dev.ps1`, which still binds the daemon to `localhost:6767` directly.)
+
+### PASEO_HOME
+
+`PASEO_HOME` is the directory that holds runtime state (agents, sockets, daemon log). Resolution rules:
+
+- The **server itself** (e.g. when launched by the desktop app or `npm run start`) defaults to `~/.paseo` (see `packages/server/src/server/paseo-home.ts`).
+- **`npm run dev` from a git worktree** derives a stable home like `~/.paseo-<worktree-name>` and, on first run, seeds it from `~/.paseo` by copying agent/project JSON metadata and `config.json`. Checkout/worktree directories are not copied.
+- **`npm run dev` from the main checkout** (not a worktree) uses a fresh `mktemp` directory under `$TMPDIR` and removes it on exit. Set `PASEO_HOME` explicitly to keep state across runs.
+
+Override knobs:
+
+```bash
+PASEO_HOME=~/.paseo-blue npm run dev          # explicit home
+PASEO_DEV_SEED_HOME=/path/to/home npm run dev # seed from a different source home
+PASEO_DEV_RESET_HOME=1 npm run dev            # clear and reseed the derived worktree home
+```
+
+### Daemon endpoints
+
+- Stable daemon launched by the desktop app: `localhost:6767`.
+- `npm run dev` (macOS/Linux): portless URLs only — read them from the `dev.sh` banner or `portless get daemon` / `portless get app`.
+- `npm run dev` (Windows): `localhost:6767` for the daemon.
+
+In any worktree-style or portless setup, never assume default ports.
+
+### Daemon logs
+
+Check `$PASEO_HOME/daemon.log` for trace-level logs.
+
+## paseo.json service scripts
+
+`worktree.setup` and `worktree.teardown` accept either a multiline shell script or an array
+of commands. Both run sequentially.
+
+```json
+{
+  "worktree": {
+    "setup": "npm ci\ncp \"$PASEO_SOURCE_CHECKOUT_PATH/.env\" .env\nnpm run db:migrate",
+    "teardown": "npm run db:drop || true"
+  }
+}
+```
+
+Every `scripts` entry with `"type": "service"` receives these environment variables:
+
+| Variable                    | Value                                                                                                                     |
+| --------------------------- | ------------------------------------------------------------------------------------------------------------------------- |
+| `PASEO_SERVICE_<NAME>_URL`  | Proxied daemon URL for a declared peer service. Prefer this for peer discovery; it survives peer restarts.                |
+| `PASEO_SERVICE_<NAME>_PORT` | Raw ephemeral port for a declared peer service. Use only as a bypass escape hatch; it can go stale if that peer restarts. |
+| `PASEO_URL`                 | Self alias for `PASEO_SERVICE_<SELF>_URL`.                                                                                |
+| `PASEO_PORT`                | Self alias for `PASEO_SERVICE_<SELF>_PORT`.                                                                               |
+| `HOST`                      | Bind host for the service process.                                                                                        |
+
+`<NAME>` is normalized from the script name by uppercasing it, replacing each run of non-`A-Z0-9` characters with `_`, and trimming leading or trailing `_`. For example, `app-server` and `app.server` both normalize to `APP_SERVER`; that collision fails at spawn time with an actionable error.
+
+`PORT` is not injected by default. If a framework requires `PORT`, set it in the command:
+
+```json
+{
+  "scripts": {
+    "web": {
+      "type": "service",
+      "command": "PORT=$PASEO_PORT npm run dev:web"
+    }
+  }
+}
+```
+
+## Build sync gotchas
+
+The daemon and CLI consume sibling workspaces from compiled `dist/` output, not `src/`. When you change a workspace that something else imports, rebuild the producer first or the consumer will speak a stale protocol and fail with handshake warnings, timeouts, or stale type errors.
+
+The fastest way to keep this consistent is to rebuild the whole daemon stack with one command:
+
+```bash
+npm run build:daemon
+```
+
+This rebuilds, in order, `@getpaseo/highlight` → `@getpaseo/relay` → `@getpaseo/server` → `@getpaseo/cli`. Use it whenever you have changed any of those four and need clean cross-package types or runtime behavior.
+
+For tighter loops, you can rebuild a single workspace:
+
+- Changed `packages/relay/src/*`: `npm run build --workspace=@getpaseo/relay` (server imports `@getpaseo/relay` from `dist/*`).
+- Changed `packages/server/src/client/*` (especially `daemon-client.ts`) or shared WS protocol types: `npm run build --workspace=@getpaseo/server` (CLI imports `@getpaseo/server` via package exports resolving to `dist/*`).
+- Changed `packages/highlight/src/*`: `npm run build --workspace=@getpaseo/highlight` (server depends on it).
+
+## CLI reference
+
+Use `npm run cli` to run the in-repo CLI from source (`npx tsx packages/cli/src/index.ts`). The globally installed `paseo` binary on macOS is a symlink into the installed Paseo desktop app, not this checkout — use it to drive the desktop's built-in daemon, but use `npm run cli` when you want to talk to the CLI you are editing.
+
+```bash
+npm run cli -- ls -a -g              # List all agents globally
+npm run cli -- ls -a -g --json       # Same, as JSON
+npm run cli -- inspect <id>          # Show detailed agent info
+npm run cli -- logs <id>             # View agent timeline
+npm run cli -- daemon status         # Check daemon status
+```
+
+Use `--host <host:port>` to point the CLI at a different daemon:
+
+```bash
+npm run cli -- --host localhost:7777 ls -a
+```
+
+## Agent state
+
+Agent data lives at:
+
+```
+$PASEO_HOME/agents/{cwd-with-dashes}/{agent-id}.json
+```
+
+Find an agent by ID:
+
+```bash
+find $PASEO_HOME/agents -name "{agent-id}.json"
+```
+
+Find by content:
+
+```bash
+rg -l "some title text" $PASEO_HOME/agents/
+```
+
+## Provider session files
+
+Get the session ID from the agent JSON (`persistence.sessionId`), then:
+
+**Claude:**
+
+```
+~/.claude/projects/{cwd-with-dashes}/{session-id}.jsonl
+```
+
+**Codex:**
+
+```
+~/.codex/sessions/{YYYY}/{MM}/{DD}/rollout-{timestamp}-{session-id}.jsonl
+```
+
+## Testing with Playwright MCP
+
+Point Playwright MCP at the running Expo web target. Under `npm run dev` (macOS/Linux) that is the portless URL printed in the dev banner — typically `https://app.localhost`. If you start Expo directly with `expo start --web` (no portless), Metro defaults to `http://localhost:8081`.
+
+Do NOT use browser history (back/forward). Always navigate by clicking UI elements or using `browser_navigate` with the full URL — the app uses client-side routing and browser history breaks state.
+
+## Expo troubleshooting
+
+```bash
+npx expo-doctor
+```
+
+Diagnoses version mismatches and native module issues.
+
+## Typecheck
+
+Always run typecheck after changes:
+
+```bash
+npm run typecheck
+```

docs/file-icons.md

@@ -0,0 +1,110 @@
+# File Icons
+
+The file explorer uses colored SVG icons from [`material-icon-theme`](https://github.com/material-extensions/vscode-material-icon-theme) (installed as a dev dependency in `packages/app`).
+
+Icons are inlined as SVG strings in:
+
+```
+packages/app/src/components/material-file-icons.ts
+```
+
+This file is auto-generated. Do not edit it by hand.
+
+## How it works
+
+- `SVG_ICONS` maps icon names (e.g. `"typescript"`) to raw SVG strings
+- `EXTENSION_TO_ICON` maps file extensions (e.g. `"ts"`) to icon names
+- `getFileIconSvg(fileName)` returns the SVG string for a given filename, falling back to a generic file icon
+- `packages/app/src/components/file-explorer-pane.tsx` is the only consumer; it renders the SVG with `SvgXml` from `react-native-svg`
+
+## Adding a new icon
+
+1. Find the icon name in the material-icon-theme manifest:
+
+```bash
+node -e "
+const m = require('./node_modules/material-icon-theme/dist/material-icons.json');
+console.log('fileExtensions:', m.fileExtensions['YOUR_EXT']);
+console.log('languageIds:', m.languageIds['YOUR_LANG']);
+"
+```
+
+2. Verify the SVG exists:
+
+```bash
+cat node_modules/material-icon-theme/icons/ICON_NAME.svg
+```
+
+3. Add two things to `material-file-icons.ts`:
+   - The SVG string in `SVG_ICONS`:
+
+     ```ts
+     "icon_name": `<svg ...>...</svg>`,
+     ```
+
+   - The extension mapping in `EXTENSION_TO_ICON`:
+     ```ts
+     "ext": "icon_name",
+     ```
+
+4. Run `npm run typecheck` to verify.
+
+## Currently included icons
+
+53 unique icons covering these extensions:
+
+| Extension(s)                               | Icon        |
+| ------------------------------------------ | ----------- |
+| `ts`                                       | typescript  |
+| `tsx`                                      | react_ts    |
+| `js`                                       | javascript  |
+| `jsx`                                      | react       |
+| `py`                                       | python      |
+| `go`                                       | go          |
+| `rs`                                       | rust        |
+| `rb`                                       | ruby        |
+| `java`                                     | java        |
+| `kt`                                       | kotlin      |
+| `c`                                        | c           |
+| `cpp`                                      | cpp         |
+| `h`                                        | h           |
+| `hpp`                                      | hpp         |
+| `cs`                                       | csharp      |
+| `swift`                                    | swift       |
+| `dart`                                     | dart        |
+| `ex`, `exs`                                | elixir      |
+| `erl`                                      | erlang      |
+| `hs`                                       | haskell     |
+| `clj`                                      | clojure     |
+| `scala`                                    | scala       |
+| `ml`                                       | ocaml       |
+| `r`                                        | r           |
+| `lua`                                      | lua         |
+| `zig`                                      | zig         |
+| `nix`                                      | nix         |
+| `php`                                      | php         |
+| `html`                                     | html        |
+| `css`                                      | css         |
+| `scss`                                     | sass        |
+| `less`                                     | less        |
+| `json`                                     | json        |
+| `yml`, `yaml`                              | yaml        |
+| `xml`                                      | xml         |
+| `toml`                                     | toml        |
+| `md`, `markdown`                           | markdown    |
+| `sql`                                      | database    |
+| `graphql`, `gql`                           | graphql     |
+| `sh`, `bash`                               | console     |
+| `tf`                                       | terraform   |
+| `hcl`                                      | hcl         |
+| `vue`                                      | vue         |
+| `svelte`                                   | svelte      |
+| `astro`                                    | astro       |
+| `wasm`                                     | webassembly |
+| `svg`                                      | svg         |
+| `png`, `jpg`, `jpeg`, `gif`, `webp`, `ico` | image       |
+| `txt`                                      | document    |
+| `conf`, `cfg`, `ini`                       | settings    |
+| `lock`                                     | lock        |
+| `groovy`                                   | groovy      |
+| `gradle`                                   | gradle      |

docs/glossary.md

@@ -0,0 +1,29 @@
+# Paseo Glossary
+
+Authoritative terminology. UI label wins. Don't invent synonyms; use what's here.
+
+- **Project** — Logical grouping of workspaces sharing a git remote (or main repo root). UI: "Project" / "Add project". Code: `ProjectSummary` (`packages/app/src/utils/projects.ts:22`), `projectKey` (`packages/server/src/server/workspace-registry-model.ts:16`). Forbidden: "Repo", "Repository" as UI label.
+- **Workspace** — One concrete `cwd` on one daemon, with git state; belongs to exactly one project. UI: "Workspace". Code: `WorkspaceDescriptorPayload` (`packages/server/src/shared/messages.ts:2178`). Don't confuse with: Branch (one branch can back many workspaces via worktrees). Forbidden: "Folder", "Directory" as UI label.
+- **Workspace kind** — `"directory" | "local_checkout" | "worktree"`. Code: `PersistedWorkspaceKind` (`packages/server/src/server/workspace-registry-model.ts:8`).
+- **Agent** — One AI coding agent run on a daemon (one provider, one model, one cwd, one timeline). UI: "Agent" / "New Agent". Code: `AgentSnapshotPayload` (`packages/server/src/shared/messages.ts:608`). Forbidden: "Task", "Job", "Run".
+- **Daemon** — Local Paseo server process; identified by `serverId`. UI: "Daemon" (system contexts only). Code: `serverId` in `ServerInfoStatusPayloadSchema` (`packages/server/src/shared/messages.ts:1936`), `DaemonClient` (`packages/server/src/client/daemon-client.ts`).
+- **Host** — Client-side connection profile pointing at a daemon; bundles one or more `HostConnection`s. UI: "Host" / "Add host" / "Switch host". Code: `HostProfile` (`packages/app/src/types/host-connection.ts:37`). Forbidden: "Connection" (means `HostConnection`, not host).
+- **Project host entry** — One row in a project for a single (project, daemon) pair, aggregating that daemon's workspaces in the project. Internal. Code: `ProjectHostEntry` (`packages/app/src/utils/projects.ts:11`). Don't introduce "Checkout" as a synonym.
+- **Placement** — One workspace's relationship to its project (projectKey, projectName, git checkout snapshot). Internal. Code: `ProjectPlacementPayload` (`packages/server/src/shared/messages.ts:2113`).
+- **Branch** — Plain git branch. UI: "Switch branch". Code: `currentBranch` in `WorkspaceGitRuntimePayloadSchema` (`packages/server/src/shared/messages.ts:2136`); `BranchSwitcher` (`packages/app/src/components/branch-switcher.tsx`).
+- **Worktree** — Paseo-managed git worktree (`~/.paseo/worktrees/{name}`); also a `workspaceKind` value. UI: CLI + `paseo.json` keys (`worktree.setup`, `worktree.teardown`) only. Code: `ProjectCheckoutLiteGitPaseoPayload` (`packages/server/src/shared/messages.ts:2092`); CLI `paseo worktree` (`packages/cli/src/commands/worktree/index.ts:8`). Forbidden: "Checkout" as a synonym.
+- **Repository / Remote** — Internal git inputs (`remoteUrl`, `mainRepoRoot`) used to derive `projectKey`. No UI label.
+- **Session** — Per-client connection to a daemon. Internal. Code: `Session` (`packages/server/src/server/session.ts`). Don't confuse with: provider-side agent session log.
+- **Profile** — Internal name for the persisted shape of a host. Code: `HostProfile` (`packages/app/src/types/host-connection.ts:37`). Never user-facing.
+- **Provider** — Agent backend (Claude Code, Codex, OpenCode). UI: "Provider". Code: `ProviderSnapshotEntry` (`packages/server/src/shared/messages.ts:198`).
+- **Model** — A specific LLM offered by a provider. UI: "Model" / "Select model". Code: `AgentModelDefinition` (`packages/server/src/shared/messages.ts:187`).
+- **Terminal** — Workspace-scoped PTY shell streamed over the binary mux channel. UI: "Terminal". Code: `TerminalStreamFrame` (`packages/server/src/shared/terminal-stream-protocol.ts`).
+- **Schedule** — Cron-style trigger that creates remote agents. UI: CLI only (`paseo schedule`). Code: `ScheduleCreateRequest` (re-exported from `packages/server/src/shared/messages.ts`). Don't confuse with: Loop (iterative re-execution of one agent).
+- **Mode** — Provider-specific operational mode (plan, default, full-access, …). UI: icon-only. Code: `modeId` in `AgentSessionConfig` (`packages/server/src/shared/messages.ts:257`).
+- **Attachment** — GitHub PR or Issue bound to an agent prompt. UI: "Attach issue or PR". Code: `AgentAttachment` (`packages/server/src/shared/messages.ts:782`).
+- **Conflict** — Two distinct senses; do NOT use the bare word in UI copy without qualifying which: (a) **stale-write conflict** on `paseo.json` ("Config changed on disk", code `stale_project_config`, `packages/app/src/screens/project-settings-screen.tsx:593`); (b) **git merge conflict** (no current UI string).
+
+## Inconsistencies (documented, not papered over)
+
+- CLI `--host <host>` description `"Daemon host target"` (`packages/cli/src/utils/command-options.ts:5`) blurs daemon/host; the app keeps them distinct.
+- `WorkspaceDescriptorPayloadSchema.workspaceKind` accepts legacy `"checkout"` on the wire (`packages/server/src/shared/messages.ts:2187`) while `PersistedWorkspaceKind` does not (`packages/server/src/server/workspace-registry-model.ts:8`).

docs/mobile-testing.md

@@ -0,0 +1,264 @@
+# Mobile Testing
+
+## Maestro
+
+Maestro flows live in `packages/app/maestro/`. Reusable sub-flows live in `packages/app/maestro/flows/`.
+
+Run a flow:
+
+```bash
+maestro test packages/app/maestro/my-flow.yaml
+```
+
+### Screenshots
+
+`takeScreenshot` writes to the **current working directory** — there's no way to configure the output path in the YAML. To keep screenshots out of the checkout, `cd` into a temp directory and use an absolute path for the flow:
+
+```bash
+FLOW="$(pwd)/packages/app/maestro/my-flow.yaml"
+mkdir -p /tmp/maestro-out
+cd /tmp/maestro-out && maestro test "$FLOW"
+```
+
+`packages/app/maestro/.gitignore` excludes `*.png` as a safety net.
+
+### Element targeting
+
+Use `testID` or `nativeID` on components, then target with `id:` in flows. Prefer this over text matching — text breaks on copy changes.
+
+```tsx
+// Component
+<Pressable testID="sidebar-sessions" onPress={onPress}>
+```
+
+```yaml
+# Flow
+- tapOn:
+    id: "sidebar-sessions"
+- assertVisible:
+    id: "sidebar-sessions"
+```
+
+### Conditional steps
+
+Use `runFlow:when:visible` for steps that should only execute when a specific element is on screen:
+
+```yaml
+- runFlow:
+    when:
+      visible:
+        id: "sidebar-sessions"
+    commands:
+      - swipe:
+          direction: LEFT
+          duration: 300
+```
+
+This is how `flows/dev-client.yaml` handles Expo dev client screens that only appear in dev builds.
+
+### Don't use launchApp against a running dev app
+
+`launchApp` kills and restarts the app, disrupting Expo dev client state and host connections. For flows that test against an already-running dev app, **omit launchApp entirely** — just interact with whatever is on screen.
+
+Use `launchApp` only in flows that need a clean start (e.g., onboarding tests).
+
+### Swipe gestures
+
+Use `start`/`end` with percentage coordinates for precise control:
+
+```yaml
+# Edge swipe from left to open sidebar
+- swipe:
+    start: "5%,50%"
+    end: "80%,50%"
+    duration: 300
+```
+
+`direction: RIGHT` is simpler but less precise — use it for generic swipes, use coordinates when the start position matters (edge gestures, avoiding specific UI regions).
+
+### Assertions
+
+`assertVisible` checks **actual screen visibility**, not just view tree presence. An element that exists in the tree but is off-screen (e.g., `translateX: -400`) will correctly fail `assertVisible`. This makes it reliable for catching animation bugs where state says "open" but the view is visually hidden.
+
+For async elements, use `extendedWaitUntil`:
+
+```yaml
+- extendedWaitUntil:
+    visible: ".*Online.*"
+    timeout: 90000
+```
+
+### Dev client handling
+
+Two reusable flows handle Expo dev client screens after launch:
+
+- `flows/launch.yaml` — handles dev launcher, dismisses dev menu, asserts "Welcome to Paseo"
+- `flows/dev-client.yaml` — same but without asserting a particular app route
+
+### Reach the composer
+
+`flows/land-in-chat.yaml` is the canonical "get into a chat" primitive. It `clearState`s, runs `launch.yaml`, taps the welcome screen's direct-connection option, types `127.0.0.1:6767`, submits, and waits for `message-input-root`. Compose any composer-level fixture on top of it:
+
+```yaml
+appId: sh.paseo
+---
+- runFlow: flows/land-in-chat.yaml
+# ...your scenario here, starting from a ready composer
+```
+
+See `image-picker-repro.yaml` for an example.
+
+**Prefer direct connection over relay pairing for local E2E.** Relay needs a 400+ character pairing URL typed into an input; direct needs `127.0.0.1:6767`. The daemon listens on 6767 and the simulator can reach it directly.
+
+### New Workspace Creation
+
+The Android workspace-creation regression has a dedicated harness:
+
+```bash
+bash packages/app/maestro/test-workspace-create-android-crash.sh
+```
+
+For a short recording that starts after launch/connection/sidebar setup:
+
+```bash
+bash packages/app/maestro/record-workspace-create-android-focus.sh
+```
+
+The flow details are documented in `packages/app/maestro/README.md`. The important rule is that a valid new-workspace assertion must prove the redirect completed: select a real model, tap `Create`, wait for `workspace-header-title`, wait for `message-input-root`, assert `New workspace` is gone, and assert the Android redbox strings are absent. Waiting for the composer alone is too weak because it can still be the `/new` route after a validation error.
+
+New workspace scenarios should compose the reusable subflows in `packages/app/maestro/flows/`:
+
+- `android-dev-client.yaml`
+- `connect-direct-if-welcome.yaml`
+- `open-prepared-project-sidebar.yaml`
+- `new-workspace-open-from-sidebar.yaml`
+- `new-workspace-select-codex-gpt54.yaml`
+- `new-workspace-submit-and-assert-created.yaml`
+
+The workspace-create shell scripts render those subflows into a temp directory before running Maestro, which keeps nested `runFlow` paths and `${PASEO_MAESTRO_*}` placeholders working together.
+
+### Inputs that Maestro types into
+
+Maestro `inputText` fires one character at a time. React Native's **controlled** `TextInput` re-renders per keystroke; if a controlled input's state update lags or re-mounts mid-type, characters are dropped silently — the final value on screen is a truncated/scrambled version of what was "typed."
+
+For inputs that E2E flows type into (host endpoint, pairing URL, etc.), use an **uncontrolled ref-backed input**: `defaultValue` + `onChangeText` writes into a `useRef`, reads via the ref on submit. No per-keystroke re-render, no dropped characters.
+
+See `pair-link-modal.tsx` for the pattern (`useRef`-backed `onChangeText`, no `value=` prop). Always pair the source change with a Maestro `assertVisible` on the input's `id + text` after `inputText`, so regressions are caught immediately.
+
+### Dropdowns that launch native presenters (iOS)
+
+On iOS, when a dropdown menu (`DropdownMenu` / RN `Modal`) item needs to launch a native presenter like `PHPickerViewController` (image picker) or a `UIDocumentPicker`, the callback **must not fire while the `Modal` is still dismissing**. UIKit dismissal completion spans multiple frames beyond React unmount; launching a native presenter mid-dismissal leaves an invisible backdrop mounted that traps every subsequent touch.
+
+`DropdownMenu` handles this by deferring the selected item's `onSelect` until `Modal.onDismiss` fires (UIKit-level dismissal complete), then adds a small extra buffer before invoking it. See `components/ui/dropdown-menu.tsx`'s `selectItem` / `flushPendingSelect`.
+
+When building a new component that composes a dropdown with a native presenter, reuse this dropdown — do not invent a new timing shim.
+
+## Self-verification loops
+
+Maestro can only interact with the app UI — it can't toggle iOS appearance, change locale, or simulate network conditions. For bugs that depend on system-level state, wrap Maestro in a bash script that handles the system changes between Maestro runs.
+
+This pattern also lets agents self-verify fixes without manual user testing.
+
+### Pattern
+
+1. Run baseline Maestro flow (confirm feature works)
+2. Make system-level change via `xcrun simctl` (toggle appearance, etc.)
+3. Re-run Maestro flow (confirm feature still works)
+4. Repeat N iterations to catch intermittent failures
+
+Scripts run `maestro test` from inside a temp directory so screenshots don't dirty the checkout.
+
+See `packages/app/maestro/test-sidebar-theme.sh` for the canonical example:
+
+```bash
+bash packages/app/maestro/test-sidebar-theme.sh 6 1
+# Args: iterations=6, wait_seconds=1 between toggle and test
+```
+
+Key elements of the script pattern:
+
+```bash
+set -euo pipefail
+ITERATIONS="${1:-3}"
+
+for i in $(seq 1 "$ITERATIONS"); do
+  # Toggle system state
+  xcrun simctl ui booted appearance light
+
+  # Wait for change to propagate
+  sleep 1
+
+  # Run Maestro flow and capture result
+  if maestro test "$FLOW" 2>&1 | tee "$ITER_DIR/test.log"; then
+    echo "PASS"
+  else
+    echo "FAIL"
+    xcrun simctl io booted screenshot "$ITER_DIR/failure-state.png"
+  fi
+done
+```
+
+## Unistyles + Reanimated
+
+### The crash
+
+Applying Unistyles theme-reactive styles (`StyleSheet.create((theme) => ...)`) directly to `Animated.View` causes **"Unable to find node on an unmounted component"** on theme change.
+
+Unistyles wraps styled components in `<UnistylesComponent>` and patches native view properties via C++. Reanimated also manages the same native node for animated transforms. When the theme changes, both systems try to update the node simultaneously and the view crashes.
+
+### The fix
+
+Use plain React Native `StyleSheet.create` for static positioning on `Animated.View`. Pass theme-dependent values as inline styles from `useUnistyles()`:
+
+```tsx
+// BAD: Unistyles dynamic style on Animated.View
+const styles = StyleSheet.create((theme) => ({
+  sidebar: {
+    position: "absolute",
+    top: 0,
+    left: 0,
+    bottom: 0,
+    backgroundColor: theme.colors.surfaceSidebar, // theme-reactive
+    overflow: "hidden",
+  },
+}));
+
+<Animated.View style={[styles.sidebar, animatedStyle]} />;
+```
+
+```tsx
+// GOOD: static stylesheet + inline theme values
+import { StyleSheet as RNStyleSheet } from "react-native";
+
+const staticStyles = RNStyleSheet.create({
+  sidebar: {
+    position: "absolute",
+    top: 0,
+    left: 0,
+    bottom: 0,
+    overflow: "hidden",
+  },
+});
+
+const { theme } = useUnistyles();
+
+<Animated.View
+  style={[staticStyles.sidebar, animatedStyle, { backgroundColor: theme.colors.surfaceSidebar }]}
+/>;
+```
+
+Regular `View` components can safely use Unistyles dynamic styles — the conflict is specific to `Animated.View`.
+
+## iOS Simulator
+
+```bash
+# Screenshot
+xcrun simctl io booted screenshot /tmp/screenshot.png
+
+# Dark/light mode
+xcrun simctl ui booted appearance          # check current
+xcrun simctl ui booted appearance dark     # set dark
+xcrun simctl ui booted appearance light    # set light
+```
+
+Expo dev server logs are in the tmux pane running `npm run dev`. Daemon logs are at `$PASEO_HOME/daemon.log` (see [development.md](development.md)).

docs/product.md

@@ -30,6 +30,7 @@ Each project opens as a workspace. For git projects, the default workspace is th
 ### Inside a workspace
 
 A workspace is a flexible canvas:
+
 - Launch multiple agents side by side in split panes
 - Open terminals alongside agents
 - Mix and match providers within the same workspace
@@ -39,6 +40,7 @@ A workspace is a flexible canvas:
 Paseo is a client-server system. The daemon (Node.js) runs on your machine, manages agent processes, and streams output in real time over WebSocket. Clients connect to the daemon — locally or remotely.
 
 This architecture means:
+
 - The daemon can run on any machine: laptop, VM, remote server
 - Multiple clients can connect simultaneously
 - Agents keep running when you close the app
@@ -65,10 +67,14 @@ Anyone who builds software:
 3. **The daemon as infrastructure.** Server/client architecture enables deployment anywhere.
 4. **Open source outlasts funding.** Open source communities are resilient. Contributors become advocates.
 
-## Current state (March 2026)
+## Current state (May 2026)
 
 - Desktop (Electron), mobile (iOS/Android), web, CLI
-- Providers: Claude Code (Agent SDK), Codex (app-server), OpenCode
-- Daily releases
-- Community contributions starting (packaging, bug fixes)
-- Key UX: split panes, keybinding customization, workspace model
+- Built-in providers: Claude Code (Agent SDK), Codex (app-server), GitHub Copilot (ACP), OpenCode, Pi
+- One-click ACP provider catalog: Cursor, Hermes, Qwen Coder, Kimi Code, and others — plus custom ACP providers
+- Voice mode: dictate prompts or talk through problems hands-free
+- MCP server exposes the daemon to other agents (create_agent, send_agent_prompt, schedules, terminals, worktrees)
+- Scheduled agents (cron-style triggers) via app, CLI, and MCP
+- Frequent releases (multiple per week)
+- Community contributions across packaging, providers, and bug fixes
+- Key UX: split panes, keybinding customization, workspace model, in-app browser

docs/providers.md

@@ -0,0 +1,380 @@
+# Adding a New Provider to Paseo
+
+This guide walks through adding a new agent provider end-to-end. There are two integration patterns, and this doc covers both.
+
+## Two Integration Patterns
+
+### ACP (Agent Client Protocol) -- recommended
+
+Extend `ACPAgentClient` from `packages/server/src/server/agent/providers/acp-agent.ts`. The base class handles process spawning, stdio transport, session lifecycle, streaming, permissions, and model discovery. You provide configuration (command, modes, capabilities) and optionally override `isAvailable()` for auth checks.
+
+The only built-in ACP provider today is `copilot` (`copilot-acp-agent.ts`). `GenericACPAgentClient` (`generic-acp-agent.ts`) is also ACP-based but is used for user-defined custom providers configured via `extends: "acp"` overrides — see [docs/custom-providers.md](custom-providers.md).
+
+### Direct
+
+Implement the `AgentClient` and `AgentSession` interfaces from `agent-sdk-types.ts` yourself. This gives full control but requires you to handle process management, streaming, permissions, and session persistence from scratch.
+
+Existing direct providers: `claude` (in `providers/claude/agent.ts`), `codex` (`codex-app-server-agent.ts`), `opencode` (`opencode-agent.ts`), `pi` (`pi-direct-agent.ts`). The dev-only `mock` provider (`mock-load-test-agent.ts`) is also direct.
+
+---
+
+## ACP Provider Checklist
+
+### 1. Create the provider class
+
+Create `packages/server/src/server/agent/providers/{name}-agent.ts`.
+
+Define capabilities, modes, and a thin subclass of `ACPAgentClient`:
+
+```ts
+import type { Logger } from "pino";
+import type { AgentCapabilityFlags, AgentMode } from "../agent-sdk-types.js";
+import type { ProviderRuntimeSettings } from "../provider-launch-config.js";
+import { ACPAgentClient } from "./acp-agent.js";
+
+const MY_PROVIDER_CAPABILITIES: AgentCapabilityFlags = {
+  supportsStreaming: true,
+  supportsSessionPersistence: true,
+  supportsDynamicModes: true,
+  supportsMcpServers: true,
+  supportsReasoningStream: true,
+  supportsToolInvocations: true,
+};
+
+const MY_PROVIDER_MODES: AgentMode[] = [
+  {
+    id: "default",
+    label: "Default",
+    description: "Standard agent mode",
+  },
+  // Add more modes as needed
+];
+
+type MyProviderClientOptions = {
+  logger: Logger;
+  runtimeSettings?: ProviderRuntimeSettings;
+};
+
+export class MyProviderACPAgentClient extends ACPAgentClient {
+  constructor(options: MyProviderClientOptions) {
+    super({
+      provider: "my-provider", // Must match the ID used everywhere else
+      logger: options.logger,
+      runtimeSettings: options.runtimeSettings,
+      defaultCommand: ["my-agent-binary", "--acp"], // CLI command to spawn
+      defaultModes: MY_PROVIDER_MODES,
+      capabilities: MY_PROVIDER_CAPABILITIES,
+    });
+  }
+
+  // Override isAvailable() if the provider needs specific auth/env vars
+  override async isAvailable(): Promise<boolean> {
+    if (!(await super.isAvailable())) {
+      return false; // Binary not found
+    }
+    return Boolean(process.env["MY_PROVIDER_API_KEY"]);
+  }
+}
+```
+
+The `super.isAvailable()` call checks that the binary from `defaultCommand` is on `$PATH`. Override only to add credential checks on top.
+
+For reference, here is how Copilot does it -- no auth override needed because the CLI handles auth itself:
+
+```ts
+export class CopilotACPAgentClient extends ACPAgentClient {
+  constructor(options: CopilotACPAgentClientOptions) {
+    super({
+      provider: "copilot",
+      logger: options.logger,
+      runtimeSettings: options.runtimeSettings,
+      defaultCommand: ["copilot", "--acp"],
+      defaultModes: COPILOT_MODES,
+      capabilities: COPILOT_CAPABILITIES,
+    });
+  }
+
+  override async isAvailable(): Promise<boolean> {
+    return super.isAvailable();
+  }
+}
+```
+
+### 2. Add to the provider manifest
+
+In `packages/server/src/server/agent/provider-manifest.ts`, add mode definitions with UI metadata (icons, color tiers) and a provider definition entry.
+
+First, define the modes with visual metadata:
+
+```ts
+const MY_PROVIDER_MODES: AgentProviderModeDefinition[] = [
+  {
+    id: "default",
+    label: "Default",
+    description: "Standard agent mode",
+    icon: "ShieldCheck",
+    colorTier: "safe",
+  },
+  {
+    id: "autonomous",
+    label: "Autonomous",
+    description: "Runs without prompting",
+    icon: "ShieldOff",
+    colorTier: "dangerous",
+  },
+];
+```
+
+Available `colorTier` values: `"safe"`, `"moderate"`, `"dangerous"`, `"planning"`.
+Available `icon` values: `"ShieldCheck"`, `"ShieldAlert"`, `"ShieldOff"`.
+
+Then add to the `AGENT_PROVIDER_DEFINITIONS` array:
+
+```ts
+export const AGENT_PROVIDER_DEFINITIONS: AgentProviderDefinition[] = [
+  // ... existing providers ...
+  {
+    id: "my-provider",
+    label: "My Provider",
+    description: "Short description of the provider",
+    defaultModeId: "default",
+    modes: MY_PROVIDER_MODES,
+    // Optional: enable voice
+    voice: {
+      enabled: true,
+      defaultModeId: "default",
+      defaultModel: "some-model",
+    },
+  },
+];
+```
+
+### 3. Add the factory to the provider registry
+
+In `packages/server/src/server/agent/provider-registry.ts`, import your class and add a factory entry to `PROVIDER_CLIENT_FACTORIES`:
+
+```ts
+import { MyProviderACPAgentClient } from "./providers/my-provider-agent.js";
+
+const PROVIDER_CLIENT_FACTORIES: Record<string, ProviderClientFactory> = {
+  // ... existing factories ...
+  "my-provider": (logger, runtimeSettings) =>
+    new MyProviderACPAgentClient({
+      logger,
+      runtimeSettings,
+    }),
+};
+```
+
+The factory is invoked with `(logger, runtimeSettings, options)`; `options.workspaceGitService` is also available if you need it (see the `codex` factory for an example). The registry already passes the per-provider runtime settings slice through, so you don't index into the map yourself.
+
+### 4. Add a provider icon (app)
+
+Create `packages/app/src/components/icons/my-provider-icon.tsx` following the pattern from existing icons (e.g., `claude-icon.tsx`):
+
+```tsx
+import Svg, { Path } from "react-native-svg";
+
+interface MyProviderIconProps {
+  size?: number;
+  color?: string;
+}
+
+export function MyProviderIcon({ size = 16, color = "currentColor" }: MyProviderIconProps) {
+  return (
+    <Svg width={size} height={size} viewBox="0 0 24 24" fill={color}>
+      <Path d="..." />
+    </Svg>
+  );
+}
+```
+
+Then register it in `packages/app/src/components/provider-icons.ts` by adding an entry to the existing `PROVIDER_ICONS` map (which already covers the built-in providers):
+
+```ts
+import { MyProviderIcon } from "@/components/icons/my-provider-icon";
+
+const PROVIDER_ICONS: Record<string, typeof Bot> = {
+  // ... existing entries ...
+  "my-provider": MyProviderIcon as unknown as typeof Bot,
+};
+```
+
+If no icon is registered, `getProviderIcon()` falls back to a generic `Bot` icon from lucide.
+
+### 5. Add E2E test config
+
+In `packages/server/src/server/daemon-e2e/agent-configs.ts`, add your provider:
+
+```ts
+export const agentConfigs = {
+  // ... existing configs ...
+  "my-provider": {
+    provider: "my-provider",
+    model: "default-model-id",
+    modes: {
+      full: "autonomous", // Mode with no permission prompts
+      ask: "default", // Mode that requires permission approval
+    },
+  },
+} as const satisfies Record<string, AgentTestConfig>;
+```
+
+Add an availability check in `isProviderAvailable()`. Note `isCommandAvailable` is async, so all branches `await` it:
+
+```ts
+case "my-provider":
+  return (
+    (await isCommandAvailable("my-agent-binary")) &&
+    Boolean(process.env.MY_PROVIDER_API_KEY)
+  );
+```
+
+Add to the `allProviders` array (current built-ins are `claude`, `codex`, `copilot`, `opencode`, `pi`):
+
+```ts
+export const allProviders: AgentProvider[] = [
+  "claude",
+  "codex",
+  "copilot",
+  "opencode",
+  "pi",
+  "my-provider",
+];
+```
+
+### 6. Run typecheck
+
+```bash
+npm run typecheck
+```
+
+This is required after every change per project rules.
+
+---
+
+## Direct Provider Checklist
+
+If your agent does not speak ACP, implement the interfaces from `agent-sdk-types.ts` directly.
+
+### Interfaces to implement
+
+The interfaces below are abridged signatures — read `agent-sdk-types.ts` for the full source of truth (option bag types, generics, etc.).
+
+**`AgentClient`** -- factory for sessions and model/mode listing:
+
+```ts
+interface AgentClient {
+  readonly provider: AgentProvider;
+  readonly capabilities: AgentCapabilityFlags;
+  createSession(
+    config: AgentSessionConfig,
+    launchContext?: AgentLaunchContext,
+    options?: AgentCreateSessionOptions,
+  ): Promise<AgentSession>;
+  resumeSession(
+    handle: AgentPersistenceHandle,
+    overrides?: Partial<AgentSessionConfig>,
+    launchContext?: AgentLaunchContext,
+  ): Promise<AgentSession>;
+  listModels(options: ListModelsOptions): Promise<AgentModelDefinition[]>;
+  isAvailable(): Promise<boolean>;
+  // Optional:
+  listModes?(options: ListModesOptions): Promise<AgentMode[]>;
+  listPersistedAgents?(options?: ListPersistedAgentsOptions): Promise<PersistedAgentDescriptor[]>;
+  getDiagnostic?(): Promise<{ diagnostic: string }>;
+}
+```
+
+**`AgentSession`** -- a running agent conversation:
+
+```ts
+interface AgentSession {
+  readonly provider: AgentProvider;
+  readonly id: string | null;
+  readonly capabilities: AgentCapabilityFlags;
+  readonly features?: AgentFeature[];
+  run(prompt: AgentPromptInput, options?: AgentRunOptions): Promise<AgentRunResult>;
+  startTurn(prompt: AgentPromptInput, options?: AgentRunOptions): Promise<{ turnId: string }>;
+  subscribe(callback: (event: AgentStreamEvent) => void): () => void;
+  streamHistory(): AsyncGenerator<AgentStreamEvent>;
+  getRuntimeInfo(): Promise<AgentRuntimeInfo>;
+  getAvailableModes(): Promise<AgentMode[]>;
+  getCurrentMode(): Promise<string | null>;
+  setMode(modeId: string): Promise<void>;
+  getPendingPermissions(): AgentPermissionRequest[];
+  respondToPermission(
+    requestId: string,
+    response: AgentPermissionResponse,
+  ): Promise<AgentPermissionResult | void>;
+  describePersistence(): AgentPersistenceHandle | null;
+  interrupt(): Promise<void>;
+  close(): Promise<void>;
+  // Optional:
+  listCommands?(): Promise<AgentSlashCommand[]>;
+  setModel?(modelId: string | null): Promise<void>;
+  setThinkingOption?(thinkingOptionId: string | null): Promise<void>;
+  setFeature?(featureId: string, value: unknown): Promise<void>;
+  tryHandleOutOfBand?(prompt: AgentPromptInput): {
+    run(ctx: { emit: (event: AgentStreamEvent) => void }): Promise<void>;
+  } | null;
+}
+```
+
+### Steps
+
+1. Create `packages/server/src/server/agent/providers/{name}-agent.ts` implementing both interfaces
+2. Add to the provider manifest (same as ACP step 2 above)
+3. Add factory to the registry (same as ACP step 3 above)
+4. Add icon (same as ACP step 4 above)
+5. Add E2E config (same as ACP step 5 above)
+6. Run typecheck
+
+---
+
+## Testing
+
+### Manual testing with the CLI
+
+Start the daemon if not already running, then:
+
+```bash
+# Launch an agent with your provider
+paseo run --provider my-provider
+
+# Launch with a specific model and mode
+paseo run --provider my-provider --model some-model --mode default
+
+# List running agents
+paseo ls -a -g
+
+# Check if the provider reports models
+paseo models --provider my-provider
+```
+
+### E2E test patterns
+
+The E2E configs in `agent-configs.ts` expose two helpers:
+
+- `getFullAccessConfig(provider)` -- returns config for a session with no permission prompts
+- `getAskModeConfig(provider)` -- returns config for a session that triggers permission requests
+
+Tests use `isProviderAvailable(provider)` to skip when the binary or credentials are missing, so CI will not fail for providers that are not installed.
+
+---
+
+## Gotchas
+
+**Mode IDs can be URIs.** ACP providers like Copilot use full URIs as mode IDs (e.g., `"https://agentclientprotocol.com/protocol/session-modes#agent"`). Never assume mode IDs are simple strings. The manifest `defaultModeId` must match exactly.
+
+**Models and modes are discovered dynamically.** ACP providers report available models and modes at runtime via the protocol. The static definitions in `provider-manifest.ts` are used for UI scaffolding (icons, color tiers) but the runtime values from the agent process are the source of truth.
+
+**`AgentProvider` is always `string`.** The type alias is `type AgentProvider = string`. Provider IDs are validated against the manifest at runtime, not at the type level.
+
+**Auth patterns vary.** Some providers need API keys in env vars (`ANTHROPIC_API_KEY`, `OPENAI_API_KEY`), some use OAuth tokens (`CLAUDE_CODE_OAUTH_TOKEN`), some use auth files (`~/.codex/auth.json`), and some handle auth entirely in their CLI binary (Copilot). Your `isAvailable()` method should check whatever is needed.
+
+**The manifest mode list and the agent class mode list are separate.** The manifest in `provider-manifest.ts` includes UI metadata (`icon`, `colorTier`). The agent class defines modes without UI metadata (just `id`, `label`, `description`). Keep them in sync.
+
+**`defaultCommand` is a tuple.** The first element is the binary name, the rest are default arguments. The base class uses this to find the executable and spawn the process.
+
+**Runtime settings can override the command.** Users can configure custom binary paths or environment variables per provider via `ProviderRuntimeSettings`. Your factory in the registry should pass `runtimeSettings?.["your-provider"]` through to the constructor.

docs/release.md

@@ -0,0 +1,322 @@
+# Release
+
+All workspaces share one version and release together.
+
+## Two paths
+
+There are two supported ways to ship from `main`:
+
+1. **Direct stable release**: you are ready to ship the current `main` commit to everyone immediately.
+2. **Beta flow**: you want public test builds first, but you are not ready for the website, npm, or production mobile release flows to move yet.
+
+## Standard release (patch)
+
+Before running any stable patch release command:
+
+- Make sure the intended release commit is already committed to `main` and the working tree is clean.
+- **Run `npm run format`, `npm run lint`, and `npm run typecheck` and commit any resulting changes BEFORE you start any `release:*` command.** `release:check` runs `npm install --workspaces --include-workspace-root` as part of `release:prepare`, which can mutate `package-lock.json` (e.g. churning `"dev": true` markers on optional deps). The next step, `version:all:*`, runs `npm version` which aborts when the working tree is dirty. If this happens mid-flight you have to commit the lockfile churn before retrying — and the pre-commit format hook will reject a lockfile-only commit because oxfmt internally skips `package-lock.json` while lefthook's glob still matches it. Avoid the whole mess by running format/lint/typecheck first, then `release:prepare` once on its own to absorb any lockfile churn into a normal commit, then start the release.
+- Do not use `npm run release:patch` as a substitute for checking whether the current commit is actually ready.
+
+```bash
+npm run release:patch
+```
+
+This bumps the version across all workspaces, runs checks, publishes to npm, and pushes the branch + tag (triggering `Desktop Release`, `Android APK Release`, EAS `release-mobile.yml`, and `Release Notes Sync` workflows).
+
+If asked to "release paseo" without specifying major/minor, treat it as a patch release.
+
+Use the direct stable path when the current `main` changes are ready to become the public release immediately.
+
+## Manual step-by-step
+
+```bash
+npm run typecheck            # Verify the exact commit you intend to release
+npm run release:check        # Typecheck, build, dry-run pack
+npm run version:all:patch    # Bump version, create commit + tag
+npm run release:publish      # Publish to npm
+npm run release:push         # Push HEAD + tag (triggers CI workflows)
+```
+
+## Beta flow
+
+```bash
+npm run release:beta:patch       # Bump to X.Y.Z-beta.1, push commit + tag
+# ... test desktop and APK prerelease assets from GitHub Releases ...
+npm run release:beta:next        # Optional: cut X.Y.Z-beta.2, beta.3, ...
+npm run release:promote          # Promote X.Y.Z-beta.N to stable X.Y.Z
+```
+
+- Beta tags are published GitHub prereleases like `v0.1.41-beta.1`
+- Betas publish desktop assets and APKs for testing, but they do not publish npm packages and do not trigger the production web/mobile release flows
+- `release:promote` creates a fresh stable tag like `v0.1.41`; the final release never reuses the beta tag
+- Desktop assets now come from the Electron package at `packages/desktop`
+- Beta releases use Electron's `beta` update channel. Users on the stable channel only receive stable releases; users on the beta channel receive beta releases and the final stable release when it is published.
+- **Do create a changelog entry for betas.** The beta entry is temporary and gets updated in place until promotion.
+
+Use the beta path when you need to:
+
+- test a build manually in a Linux or Windows VM
+- send a build to a user who is hitting a specific problem
+- iterate on `beta.1`, `beta.2`, `beta.3`, and so on before deciding to ship broadly
+
+## Staged rollout (stable channel)
+
+Stable desktop releases go out via a linear time-based rollout: 0% admitted when the updater manifests appear, 100% admitted 24 hours later, linear ramp in between. Beta releases bypass the rollout entirely — beta users always receive updates immediately.
+
+The rollout is driven by a `rolloutHours` field stamped into the GitHub Release manifests (`latest-mac.yml`, `latest-linux.yml`, `latest.yml`) by the `finalize-rollout` job in `desktop-release.yml`.
+
+Desktop release builds now publish in two phases:
+
+- Platform build jobs upload the installers/packages (`.dmg`, `.zip`, `.exe`, `.AppImage`, etc.) to the GitHub release.
+- The final job merges/stamps the manifests and uploads all `.yml` files only after they already contain the final `releaseDate` and `rolloutHours`.
+
+Updater clients only discover a release through those `.yml` manifests, so there is no silent 100% admission window before rollout metadata is present.
+
+### Default behavior
+
+`npm run release:patch` → tag push → 24h ramp. No extra action needed.
+
+The `rollout_hours` input on `desktop-release.yml` is **only read on `workflow_dispatch`** — tag-push runs always default to 24. To get any other rollout duration on a fresh release, use the post-publish flip below.
+
+### Instant-admit release (rollout_hours=0 from publish)
+
+For a fresh release that should admit everyone immediately (low-risk change, doc-only, hotfix, or just a release you want out fast), cut the release normally and queue the rollout flip immediately after:
+
+```bash
+# 1. Cut and publish (default 24h ramp from tag push).
+npm run release:patch
+
+# 2. Immediately queue the flip — runs as soon as finalize-rollout completes.
+gh workflow run desktop-rollout.yml \
+  -f tag=v0.1.64 \
+  -f rollout_hours=0
+```
+
+**Why this is gap-free:** `desktop-release.yml`'s `finalize-rollout` job and `desktop-rollout.yml` share the concurrency group `desktop-rollout-<tag>`. Dispatching `desktop-rollout.yml` while the tag-push pipeline is still running queues it safely behind `finalize-rollout`. The first public manifests already carry `rolloutHours=24`, then `desktop-rollout.yml` flips them to `rolloutHours=0` shortly afterward. The renderer polls every 30 minutes, so active stable users pick up the new manifest on their next check.
+
+Run the dispatch right after `release:patch` returns. Don't wait for the tag-push CI to finish.
+
+### Adjusting an already-published release
+
+To change the rollout duration on a release that's already shipped — e.g. flip a hotfix to instant admit, or slow a release down — use the dedicated `desktop-rollout.yml` workflow. It edits the manifests in place on the GitHub release without rebuilding anything. It only rewrites `rolloutHours`; `releaseDate` is preserved, so the rollout clock keeps ticking from the original publish time.
+
+**Hotfix (instant admit) on an already-shipped release:**
+
+```bash
+gh workflow run desktop-rollout.yml \
+  -f tag=v0.1.42 \
+  -f rollout_hours=0
+```
+
+`rollout_hours=0` admits 100% of stable users on their next update check (within ~30 min for active clients).
+
+**Slow a rollout down** (e.g. extend total duration to 72h since the original release):
+
+```bash
+gh workflow run desktop-rollout.yml \
+  -f tag=v0.1.42 \
+  -f rollout_hours=72
+```
+
+`rollout_hours` is **total duration since the original release date**, not "extend by N more hours from now." If `v0.1.42` was published 2h ago and you set `rollout_hours=72`, the ramp finishes 70h from now.
+
+The dispatch is idempotent and shares the `desktop-rollout-<tag>` concurrency group with `desktop-release.yml`'s `finalize-rollout` job, so it serializes safely against an in-flight tag-push pipeline targeting the same release.
+
+### Custom ramp on a manually-dispatched build
+
+`desktop-release.yml` accepts `rollout_hours` only on `workflow_dispatch`, which is the path used to **rebuild an existing tag** (retry a failed release, force a rebuild on a different ref). When you go that route, you can stamp a non-default ramp directly:
+
+```bash
+gh workflow run desktop-release.yml \
+  -f tag=v0.1.43 \
+  -f rollout_hours=6
+```
+
+This does **not** apply to fresh releases cut via `npm run release:patch` — that path always tag-pushes and stamps 24. For a fresh release with a custom ramp, cut normally and then dispatch `desktop-rollout.yml` (same pattern as the instant-admit flow above, with your chosen `rollout_hours`).
+
+### Releasing during an active rollout
+
+If you ship N+1 while N is still ramping, N+1 starts a fresh rollout from its own publish timestamp. N's rollout effectively ends — the newer manifest supersedes it.
+
+If N+1 is a hotfix for a bug in N, dispatch `desktop-rollout.yml -f tag=v0.1.<N+1> -f rollout_hours=0` after N+1 publishes so the users who already got N reach the fix fast.
+
+### Limitations
+
+- **No pause / kill switch.** Once a stable user is admitted, they will install the update on next quit (`autoInstallOnAppQuit = true`). To stop new admissions, ship a superseding release. To "recall" already-admitted users, ship a hotfix `+1` patch.
+- **No rollback.** `allowDowngrade = false`. Bad release = ship a hotfix.
+- **Bootstrap caveat.** Clients running a build older than the rollout feature ignore `rolloutHours` and admit immediately. Rollout protection only applies to clients running the rollout-aware version or later.
+- **Up to ~30 min admission latency.** Renderer polls every 30 minutes, so a stable user may take up to that long to be evaluated against the rollout window.
+
+## Release notes on GitHub
+
+The GitHub Release body is populated automatically by the `Release Notes Sync` workflow (`.github/workflows/release-notes-sync.yml`). It triggers on every `v*` tag push and on any push to `main` that touches `CHANGELOG.md`, then runs `scripts/sync-release-notes-from-changelog.mjs` to mirror the matching changelog entry into the release body. You don't need to write release notes on GitHub manually — keep `CHANGELOG.md` correct and the workflow will sync it. To force a re-sync, dispatch the workflow with the tag input.
+
+## Website behavior
+
+- The website download page points to GitHub's latest published **stable** release.
+- Published beta prereleases are public on GitHub Releases, but they do **not** become the website download target.
+- The website only moves when you publish the final stable release tag like `v0.1.41`.
+- The website itself is deployed by `Deploy Website` (Cloudflare Workers), which redeploys on `release: published` for non-prerelease releases and on pushes to `main` that touch `CHANGELOG.md` or `packages/website/**`.
+
+## Fixing a failed release build
+
+**NEVER bump the version to fix a build problem.** New versions are reserved for meaningful product changes (features, fixes, improvements). Build/CI failures are fixed on the current version.
+
+**Do not rely on `workflow_dispatch` for tagged code fixes.** The `workflow_dispatch` trigger runs the workflow file from the default branch but checks out the code at the tag ref (`ref: ${{ inputs.tag }}`). That means fixes committed to `main` won't change the tagged source tree being built. `workflow_dispatch` only helps when the fix lives in the workflow file itself.
+
+To retry a failed workflow, **always push a retry tag** on the commit you want to build. Reusing the same tag name is expected: move it with `git tag -f ...` and push it with `--force` so the workflow rebuilds the commit you actually want.
+
+Prefer a tag push over `workflow_dispatch` whenever you are rebuilding release code or release assets.
+
+The retry tag patterns below still work and remain the supported way to rebuild specific release targets:
+
+```bash
+# Desktop (all platforms)
+git tag -f desktop-v0.1.28 HEAD && git push origin desktop-v0.1.28 --force
+
+# Desktop (single platform)
+git tag -f desktop-macos-v0.1.28 HEAD && git push origin desktop-macos-v0.1.28 --force
+git tag -f desktop-linux-v0.1.28 HEAD && git push origin desktop-linux-v0.1.28 --force
+git tag -f desktop-windows-v0.1.28 HEAD && git push origin desktop-windows-v0.1.28 --force
+
+# Android APK
+git tag -f android-v0.1.28 HEAD && git push origin android-v0.1.28 --force
+
+# Beta
+git tag -f v0.1.29-beta.2 HEAD && git push origin v0.1.29-beta.2 --force
+```
+
+This ensures the checkout ref matches the actual code on `main` with the fix included.
+
+- `vX.Y.Z` or `vX.Y.Z-beta.N` rebuilds the full tagged release
+- `desktop-vX.Y.Z` rebuilds desktop for all desktop platforms only
+- `desktop-macos-vX.Y.Z`, `desktop-linux-vX.Y.Z`, and `desktop-windows-vX.Y.Z` rebuild only that desktop platform
+- `android-vX.Y.Z` rebuilds the Android APK release only
+
+## Notes
+
+- `version:all:*` bumps root + syncs workspace versions and `@getpaseo/*` dependency versions
+- `release:prepare` refreshes workspace `node_modules` links to prevent stale types
+- `npm run dev:desktop` and `npm run build:desktop` target the Electron desktop package in `packages/desktop`
+- If `release:publish` partially fails, re-run it — npm skips already-published versions
+- The website uses GitHub's latest published release API for download links, so published beta prereleases do not replace the stable download target.
+
+## Changelog format
+
+Release notes depend on the changelog heading format. The heading **must** be strictly followed:
+
+```
+## X.Y.Z - YYYY-MM-DD
+## X.Y.Z-beta.N - YYYY-MM-DD
+```
+
+No prefix (`v`), no extra text. The parser matches the first `## X.Y.Z` line to extract the version. A malformed heading will break download links on the homepage.
+
+## Changelog policy
+
+- `CHANGELOG.md` includes stable releases and the current beta line.
+- The first beta inserts a top entry like `## 0.1.60-beta.1 - YYYY-MM-DD`.
+- The next beta updates that same top entry in place, for example from `0.1.60-beta.1` to `0.1.60-beta.2`.
+- Stable promotion updates that same entry in place, for example from `0.1.60-beta.2` to `0.1.60`.
+- Do not create duplicate entries for each beta on the same version line.
+
+## Changelog ownership
+
+- **Only Claude should write changelog entries.**
+- If you are Codex and a stable release needs a changelog entry, launch a Claude agent with Paseo to draft it, then review and commit the result.
+
+## Changelog voice
+
+The changelog is shown on the Paseo homepage. Write it for **end users**, not developers.
+
+- **Frame everything from the user's perspective.** Describe what changed in the app, not what changed in the code. Users care that "workspaces load instantly" — not that a component no longer remounts.
+- **Never mention component names, internal modules, or implementation details.** No `WorkingIndicator`, no `accumulatedUsage`, no `reconcileAndEmitWorkspaceUpdates`.
+- **Collapse internal iterations.** If a feature was added and then fixed within the same release, just list the feature as working. Users never saw the broken version.
+- **Only list changes relative to the previous stable release.** The diff is `v(previous)..HEAD`. If something was introduced and fixed between those two tags, it never shipped — don't mention the fix.
+  - **Common trap:** when drafting from `git log`, every commit looks like a separate bullet — including the "fix X" commits that landed on top of a brand-new feature in the same release window. Before listing a Fixed entry, check whether the thing being fixed was itself added in this same release. If so, drop the fix and fold it into the feature bullet.
+  - **Example:** if the release adds an in-app browser and also contains a commit "fix: browser pane keyboard handling no longer steals shortcuts", do **not** list the keyboard fix under Fixed. The browser is shipping for the first time, so users will only ever see the working version. The Added entry covers it.
+- **Cut low-signal entries.** "Toolbar buttons have consistent sizing" is too granular. Combine small polish items or drop them.
+
+## Changelog conciseness
+
+Every bullet must be scannable at a glance. The changelog is not release documentation — it's a list.
+
+- **One line per bullet.** If a bullet wraps to three lines in a narrow column, it's too long.
+- **Split bullets that pack multiple distinct changes.** If a bullet uses "and", "plus", a comma list, or an em-dash to chain several independent improvements, break them into separate bullets — even when they share a theme or author. One bullet = one user-facing change.
+- **Trim qualifying clauses.** Drop "with a hint shown when…", "matching the CLI's behaviour", "across common install shapes". If the detail doesn't change whether a user cares, cut it.
+- **Lead with the outcome.** "Windows: agents launch reliably from npm `.cmd` shims…" is better than "Windows: agents launch reliably across common install shapes. Claude, Codex, and OpenCode now start correctly…".
+- **Attribution follows the split.** When you split a dense bullet, move each PR/author to the bullet it belongs to. Never duplicate the same PR across multiple bullets.
+
+## Changelog attribution
+
+Every changelog bullet must credit contributors and link to the PR(s) that delivered the change. This is not one-PR-per-line — a single bullet describes a user-facing change and may reference multiple PRs.
+
+Format: append `([#123](https://github.com/getpaseo/paseo/pull/123) by [@user](https://github.com/user))` at the end of each bullet. For changes spanning multiple PRs or contributors:
+
+```markdown
+- Voice mode now works on tablets with proper microphone permissions. ([#210](https://github.com/getpaseo/paseo/pull/210), [#215](https://github.com/getpaseo/paseo/pull/215) by [@alice](https://github.com/alice), [@bob](https://github.com/bob))
+```
+
+Rules:
+
+- **Always link the PR number** as `[#N](https://github.com/getpaseo/paseo/pull/N)`.
+- **Always link the contributor's GitHub profile** as `[@user](https://github.com/user)`.
+- **One bullet = one user-facing change**, regardless of how many PRs went into it. Group related PRs on the same bullet.
+- **De-duplicate contributors.** If the same person authored multiple PRs in one bullet, list them once.
+- **Only credit external contributors.** Skip attribution for [@boudra](https://github.com/boudra). The changelog credits community contributions — core team work is the default.
+- **Credit the commit author, not the PR opener.** A maintainer often opens a PR that lands work authored by someone else (cherry-pick, rebase of a contributor's branch, manual extraction from a stacked PR). The squash commit preserves the original commit's author, but `gh pr view N --json author` returns the PR opener — using that field will silently mis-credit the work to the maintainer (and then the "skip @boudra" rule drops the attribution entirely). Always resolve attribution from commit authors.
+
+  Use this command to get the GitHub logins for each PR:
+
+  ```bash
+  gh pr view N --json commits --jq '[.commits[].authors[].login] | unique | .[]'
+  ```
+
+  This returns every distinct GitHub login that authored or co-authored a commit in the PR. Use those logins for attribution. Fall back to `gh pr view N --json author` only if the commits command returns nothing (which should not happen for merged PRs).
+
+  When listing PR numbers, `git log --format='%H %s' v<previous>..HEAD | grep -E '\(#[0-9]+\)$'` pulls the PR number out of squash commit subjects.
+
+## Changelog ordering
+
+Entries within each section (Added, Improved, Fixed) are ordered by user impact:
+
+1. **User-facing features and changes first** — things users will notice, want to try, or that change their workflow.
+2. **Quality-of-life improvements** — polish, performance, smoother interactions.
+3. **Internal/infra changes last** — only include if they have a tangible user benefit (e.g. "faster startup" is user-facing even if the fix was internal).
+
+## Pre-release sanity check
+
+Before cutting any release (beta or stable), run a Codex review of the diff as a last line of defence against shipping bugs.
+
+Load the `paseo` skill and launch a **Codex 5.4** agent with a prompt like:
+
+> Review the diff between the latest release tag and HEAD. Focus on:
+>
+> 1. **Breaking changes** — especially in the WebSocket protocol, agent lifecycle, and any server↔client contract.
+> 2. **Backward compatibility** — the important direction is old app clients talking to newly updated daemons. Users update desktop and daemon first, then keep running the old app for a while. Flag anything that breaks old clients against new daemons or requires both sides to update in lockstep.
+> 3. **Regressions** — anything that looks like it could break existing functionality.
+>
+> Diff: `git diff <latest-release-tag>..HEAD`
+
+The agent's job is a deep sanity check, not a full code review. If it flags anything, investigate before proceeding.
+
+## Changelog scope
+
+The changelog always covers **stable-to-HEAD**:
+
+- **Beta release**: the diff and release notes cover `latest stable tag -> HEAD`. The current beta changelog entry is updated in place.
+- **Stable release**: the same changelog entry is promoted in place. It still captures the full delta from the previous stable release, not just what changed since the last beta.
+
+In other words, betas are checkpoints along the way; the changelog entry remains the single record for the final jump from one stable version to the next.
+
+## Completion checklist
+
+- [ ] Run the pre-release sanity check (see above) and address any findings
+- [ ] Ensure the intended release commit is already committed and the git worktree is clean before running any `release:*` patch/promote command
+- [ ] Ensure local `npm run typecheck` passes on that exact commit before running any `release:*` patch/promote command
+- [ ] Update `CHANGELOG.md` with user-facing release notes (features, fixes — not refactors)
+- [ ] Verify the changelog heading follows strict `## X.Y.Z - YYYY-MM-DD` format
+- [ ] `npm run release:patch` or `npm run release:promote` completes successfully
+- [ ] GitHub `Desktop Release` workflow for the `v*` tag is green
+- [ ] GitHub `Android APK Release` workflow for the same tag is green
+- [ ] EAS `release-mobile.yml` workflow for the same tag is green

docs/testing.md

@@ -79,7 +79,9 @@ const realSender: EmailSender = { send: sendgrid.send };
 function createTestEmailSender() {
   const sent: Array<{ to: string; body: string }> = [];
   return {
-    send: async (to: string, body: string) => { sent.push({ to, body }); },
+    send: async (to: string, body: string) => {
+      sent.push({ to, body });
+    },
     sent,
   };
 }
@@ -96,6 +98,37 @@ When a test is labeled end-to-end, it calls the real service. No environment var
 - Test bodies should read like plain English
 - Build a vocabulary of test helpers that make complex flows simple
 
+### File naming
+
+Vitest picks up tests by suffix. The suffix tells the runner which category it belongs to.
+
+| Suffix                | What it is                                                                                         | Where it runs                                                                        |
+| --------------------- | -------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------ |
+| `*.test.ts(x)`        | Unit test — pure, fast, no daemon                                                                  | `npm run test:unit`                                                                  |
+| `*.posix.test.ts`     | Unit test that needs POSIX-only behavior                                                           | unit, skipped on Windows                                                             |
+| `*.browser.test.ts`   | App test that needs a real browser (DOM)                                                           | `npm run test:browser` (Vitest browser mode, Playwright provider, headless Chromium) |
+| `*.e2e.test.ts`       | End-to-end against a real daemon                                                                   | `npm run test:e2e`                                                                   |
+| `*.real.e2e.test.ts`  | E2E that hits a real provider (Claude/Codex/OpenCode) — needs creds in `packages/server/.env.test` | `npm run test:integration:real` / `test:e2e:real`                                    |
+| `*.local.e2e.test.ts` | E2E that needs a local-only resource                                                               | `npm run test:integration:local` / `test:e2e:local`                                  |
+
+App-level Playwright browser E2E lives in `packages/app/e2e/*.spec.ts` and runs via `npm run test:e2e --workspace=@getpaseo/app` (separate from Vitest E2E).
+
+### Test setup
+
+- Server: `packages/server/src/test-utils/vitest-setup.ts` loads `.env.test`, sets `PASEO_SUPERVISED=0`, and disables Git/SSH prompts. Add new global env shims here, not in individual tests.
+- App: `packages/app/vitest.setup.ts` provides `expo`/`__DEV__` shims and stubs a few native-only modules (`react-native-unistyles`, `react-native-svg`, `expo-linking`, `@xterm/addon-ligatures`). Stubbing here is for modules that have no meaningful Node behavior — not a license to mock app code.
+
+## Running tests locally
+
+Test suites in this repo are heavy. Running them in bulk freezes the machine, especially with multiple agents in parallel.
+
+- Run only the file you changed: `npx vitest run <path> --bail=1`
+- Never run `npm run test` for a whole workspace unless asked.
+- For a broad sweep, redirect to a file and read it after: `npx vitest run <path> --bail=1 > /tmp/test-output.txt 2>&1`
+- Never re-run a suite another agent already reported green.
+- For full-suite confidence, push to CI and check GitHub Actions.
+- Never run Playwright E2E (`packages/app/e2e/*.spec.ts`) locally — defer to CI.
+
 ## Agent authentication in tests
 
 Agent providers handle their own auth. Do not add auth checks, environment variable gates, or conditional skips to tests. If auth fails, report it.
@@ -115,6 +148,7 @@ The test output is the source of truth, not your reading of the code.
 ## Design for testability
 
 If code isn't testable, refactor it. Signs:
+
 - You want to reach for a mock
 - You can't inject a dependency
 - You need to test private internals

docs/unistyles.md

@@ -0,0 +1,327 @@
+# Unistyles Gotchas
+
+This app uses [`react-native-unistyles` v3](https://www.unistyl.es/) for theme-aware styles. Unistyles is fast because most style updates do not go through React renders: the [Babel plugin](https://www.unistyl.es/v3/other/babel-plugin) rewrites React Native component imports, attaches style metadata, and lets the native ShadowRegistry update tracked views when theme or runtime dependencies change.
+
+That model is powerful, but it has sharp edges. Use this note when adding theme-dependent styles.
+
+## STOP — `useUnistyles()` Is Forbidden
+
+**Do not call `useUnistyles()` unless every alternative below has been ruled out and you can explain in a code comment why.** The library authors themselves [strongly advise against it](https://www.unistyl.es/v3/references/use-unistyles):
+
+> We strongly recommend **not using** this hook, as it will re-render your component on every change. This hook was created to simplify the migration process and should only be used when other methods fail.
+
+We have hit this gotcha repeatedly in Paseo. It manifests as periodic, lockstep re-renders of warm subtrees (agent streams, panels, sidebars) even when nothing the user can see has changed — confirmed in profiling: `AgentStreamView` re-rendering constantly with `theme` showing as the only changed input on every render. The hook subscribes the component to **all** Unistyles runtime changes (theme, breakpoint, insets, color scheme, scale) and returns a fresh object reference each call, which also breaks every downstream `useMemo`/`memo` boundary that includes a derived theme value.
+
+Before reaching for `useUnistyles()`, work down this list of alternatives in order:
+
+### 1. `StyleSheet.create((theme) => ...)` — default
+
+Most theme-aware styling needs nothing else. The Babel plugin tracks theme dependencies inside the factory and updates the native ShadowTree without any React re-render.
+
+```tsx
+const styles = StyleSheet.create((theme) => ({
+  container: {
+    backgroundColor: theme.colors.surface0,
+    padding: theme.spacing[4],
+  },
+}));
+
+<View style={styles.container} />;
+```
+
+If you are reading a theme value just to feed it back into a `style` prop, you almost certainly want this and not the hook.
+
+### 2. Hard-coded constants for genuinely static values
+
+If you only need a number that happens to live on the theme (e.g. a fixed spacing value used to compute a gap or animation distance), use a literal constant or import a static module. Static reads do not need a subscription. See the "Static Theme Imports" section below — importing `baseColors`, theme-name constants, or `type Theme` is fine when the value is intentionally static.
+
+### 3. `withUnistyles(Component)` for third-party props
+
+When a third-party component takes a non-`style` prop that must be theme-reactive (e.g. `BlurView.tint`, `Image.tintColor`, navigator option props, bottom-sheet `backgroundStyle`), wrap that single component with `withUnistyles`. Only the wrapper re-renders, not the surrounding tree.
+
+```tsx
+const ThemedBlur = withUnistyles(BlurView);
+<ThemedBlur tint={theme.colors.surface0} />;
+```
+
+(Mind the `> *` child-selector leak documented further down.)
+
+### 4. Lift the read into a tiny leaf component
+
+If only one prop in a large component needs a theme value at runtime, extract a small leaf component that calls `useUnistyles()` and accept its re-renders in isolation. Never let a whole stream / panel / sidebar / virtualized list subscribe.
+
+### 5. (Last resort) `useUnistyles()`
+
+Only acceptable when both of:
+
+- (a) The value is consumed by a 3rd-party library that cannot be wrapped with `withUnistyles` (per the upstream "When to use it?" list), AND
+- (b) The component is small, leaf-level, and not on a hot render path.
+
+If you add a new `useUnistyles()` call, leave a comment on the line explaining which of (a)/(b) applies and why each higher-priority alternative was ruled out.
+
+### Hot-path forbidden list
+
+Do not introduce `useUnistyles()` in or above any of these subtrees — re-renders here are observably expensive:
+
+- `AgentStreamView` and anything it renders (message rows, tool calls, plan card, todo list, activity log, compaction marker, copy buttons)
+- `AgentPanel` body (`packages/app/src/panels/agent-panel.tsx`)
+- `Composer` and `MessageInput`
+- `WorkspaceScreen` shell, `WorkspaceDesktopTabsRow`, deck wrapper
+- `LeftSidebar` row items, `SidebarWorkspaceList`, `CommandCenter`
+- Anything inside a virtualized list (`@tanstack/react-virtual`, `FlashList`)
+
+Reviewers must reject PRs that add `useUnistyles()` calls in these areas without a written justification matching the last-resort criteria above.
+
+## How Updates Propagate
+
+For standard React Native components, the [Unistyles Babel plugin](https://www.unistyl.es/v3/other/babel-plugin) rewrites imports such as `View`, `Text`, `Pressable`, and `ScrollView` to Unistyles-aware component factories. On native, those factories borrow the component ref and register the `style` prop with the ShadowRegistry. The upstream ["Why my view doesn't update?"](https://www.unistyl.es/v3/guides/why-my-view-doesnt-update) guide describes this as the ShadowTree update path that avoids unnecessary React re-renders.
+
+The important detail: the automatic native path tracks `props.style`. It does not generally track every prop that happens to carry style-like values.
+
+[`useUnistyles()`](https://www.unistyl.es/v3/references/use-unistyles) is different. It gives React access to the current theme/runtime and can make a component re-render when those values change. Use it for values that must be rendered through React props, such as icon colors or small escape hatches. Do not expect direct reads from `UnistylesRuntime` to re-render a component; [issue #817](https://github.com/jpudysz/react-native-unistyles/issues/817) is a useful reminder of that invariant.
+
+## Main Gotcha: `contentContainerStyle`
+
+`ScrollView.contentContainerStyle` is the canonical trap. It looks like a style prop, but it is not the same prop that Unistyles' remapped native component registers by default. The upstream tutorial calls this out directly in its [ScrollView Background Issue](https://www.unistyl.es/v3/tutorial/settings-screen#scrollview-background-issue) section.
+
+Avoid this pattern when the style depends on the theme:
+
+```tsx
+<ScrollView contentContainerStyle={styles.container} />;
+
+const styles = StyleSheet.create((theme) => ({
+  container: {
+    flexGrow: 1,
+    backgroundColor: theme.colors.surface0,
+  },
+}));
+```
+
+On first mount this can paint with the current adaptive or initial theme. If app settings later load a persisted theme and call [`UnistylesRuntime.setTheme`](https://www.unistyl.es/v3/guides/theming#change-theme), the JS-side style proxy may report the new theme while the native content container keeps the old background. That is how the welcome screen ended up with a light background and dark foreground/buttons.
+
+This applies broadly to non-`style` props that carry theme-dependent values, such as component props named `color`, `trackColor`, `tintColor`, `backgroundStyle`, `handleIndicatorStyle`, and other library-specific style props. The [3rd-party view decision algorithm](https://www.unistyl.es/v3/references/3rd-party-views) recommends explicit handling for these cases, and [issue #1030](https://github.com/jpudysz/react-native-unistyles/issues/1030) shows a related native-prop update edge case around `Image.tintColor`. Treat these values as React props unless wrapped with `withUnistyles`.
+
+## Fix Patterns
+
+Preferred pattern: put themed backgrounds on a normal wrapper view, and keep `contentContainerStyle` theme-free.
+
+```tsx
+<View style={styles.container}>
+  <ScrollView contentContainerStyle={styles.contentContainer}>{children}</ScrollView>
+</View>;
+
+const styles = StyleSheet.create((theme) => ({
+  container: {
+    flex: 1,
+    backgroundColor: theme.colors.surface0,
+  },
+  contentContainer: {
+    flexGrow: 1,
+    padding: theme.spacing[4],
+  },
+}));
+```
+
+This is the pattern used by the settings screen: the screen background lives on a normal `View style={styles.container}`, while the scroll content container only carries layout.
+
+In practice the wrapper-`View` pattern is the one we use. Across the app, `withUnistyles` is now reserved for wrapping leaf components — mostly lucide icons (`ThemedActivityIndicator`, `ThemedChevronDown`, …) and small third-party components like `MarkdownWithStableRenderer` — so they pick up theme-reactive `color`/`tintColor` props without re-rendering their parent.
+
+In principle, [`withUnistyles`](https://www.unistyl.es/v3/references/with-unistyles) can also wrap a `ScrollView` to make `contentContainerStyle` theme-reactive via its [auto-mapping behavior for `style` and `contentContainerStyle`](https://www.unistyl.es/v3/references/with-unistyles#auto-mapping-for-style-and-contentcontainerstyle-props). We previously did this on the welcome screen and hit the `> *` child-selector leak documented below; we have since moved the welcome screen to the wrapper-`View` pattern. If you find yourself reaching for `withUnistyles(ScrollView)`, treat it as a smell and check whether a wrapper view works first.
+
+The smallest escape hatch is to use `useUnistyles()` and pass an inline value through React:
+
+```tsx
+const { theme } = useUnistyles();
+
+<ScrollView
+  contentContainerStyle={[styles.contentContainer, { backgroundColor: theme.colors.surface0 }]}
+/>;
+```
+
+Use this sparingly. It works because React re-renders the prop, but it gives up the main Unistyles native-update path for that value.
+
+## `withUnistyles` And The `> *` Child-Selector Leak
+
+`withUnistyles` on a component with a theme-dependent `style` prop works by wrapping the component in a `<div style={{display: 'contents'}} className={hash}>` and emitting the style under a `.hash > *` child selector so the styles cascade onto the wrapped component. This is how auto-mapping for `style` and `contentContainerStyle` works on web.
+
+The sharp edge: Unistyles hashes styles by value. If `withUnistyles` receives a style whose value is **identical** to a style used elsewhere in the app on a plain `View`, both usages get the same hash — and both CSS rules (the element rule and the `> *` child rule) are emitted under the same class name. The `> *` rule then leaks onto the direct children of every `View` that shares the hash.
+
+Concrete regression we hit: `welcome-screen.tsx` had `const ThemedScrollView = withUnistyles(ScrollView)` with `style={{ flex: 1, backgroundColor: theme.colors.surface0 }}`. `panels/agent-panel.tsx` had `root` and `container` styles with the exact same value. All three collided on class `unistyles_j2k2iilhfz`, so the browser stylesheet contained:
+
+```css
+.unistyles_j2k2iilhfz {
+  flex: 1 1 0%;
+  background-color: var(--colors-surface0);
+}
+.unistyles_j2k2iilhfz > * {
+  flex: 1 1 0%;
+  background-color: var(--colors-surface0);
+}
+```
+
+The child-selector rule forced `flex:1` and `background-color: surface0` onto the Composer's outer `Animated.View` (a direct child of `container`), stretching it to fill remaining space and leaving a large empty gap between the composer UI and the bottom of the screen. It also painted a `surface0` band behind the scroll-to-bottom button. The bug only appeared in the browser — Electron skips `WelcomeScreen` after pairing, so the `> *` rule was never injected there.
+
+Symptoms to watch for:
+
+- A sibling of a themed panel-background `View` stretches unexpectedly on web only.
+- Random direct children of a `{ flex: 1, backgroundColor: surface0 }` `View` pick up an unexpected background.
+- DevTools shows a `.unistyles_xxx > *` rule you did not write.
+
+Quick confirmation in DevTools console:
+
+```js
+[...document.styleSheets]
+  .flatMap((s) => [...(s.cssRules || [])])
+  .map((r) => r.cssText)
+  .filter((t) => t.includes("unistyles") && t.includes("> *"));
+```
+
+Any match beyond benign `r-pointerEvents-* > *` rules from react-native-web is a leak.
+
+Avoid the bug by preferring the wrapper-`View` pattern from the previous section whenever possible: put `{ flex: 1, backgroundColor: surface0 }` on a plain `View` and give the `ScrollView` a theme-free `style`/`contentContainerStyle`. That keeps `withUnistyles` off the hot path and avoids the hash collision. Only reach for `withUnistyles(ScrollView)` when a wrapper view is genuinely awkward, and when you do, give the wrapped style a distinctive shape (extra key, different layout) so it does not hash-collide with a common panel background used elsewhere.
+
+## Hidden Sheet Content
+
+`@gorhom/bottom-sheet` can keep `BottomSheetModal` content mounted while the sheet is hidden. That matters during Paseo's startup theme transition: a header node can be created under the initial adaptive theme, stay hidden, then appear later with stale native style values even though surrounding content has re-rendered correctly.
+
+We saw this in `AdaptiveModalSheet`: the body text and buttons were dark-theme-correct, but the shared sheet title opened with the initial light-theme text color on a dark sheet background. For tiny values in a reusable sheet header, prefer the inline escape hatch:
+
+```tsx
+const { theme } = useUnistyles();
+
+<Text style={[styles.title, { color: theme.colors.foreground }]}>{title}</Text>;
+```
+
+Keep layout and typography in `StyleSheet.create`; move only the stale theme-dependent value through React. If a larger subtree shows the same behavior, consider remounting the sheet on theme changes or moving the themed paint onto a wrapper that is mounted with the visible content.
+
+The same rule applies to bottom-sheet component props such as `backgroundStyle` and `handleIndicatorStyle`: they are library props, not the direct React Native `style` prop Unistyles registers. Prefer a custom `backgroundComponent` that calls `useUnistyles()`, or pass a small inline object from the hook theme.
+
+## Memoized Style Objects
+
+When a third-party library receives a plain style object, it is outside Unistyles' native tracking path. Make sure any memo that builds that style object depends on the actual theme values it reads.
+
+Avoid indirect keys like this:
+
+```tsx
+const { theme, rt } = useUnistyles();
+const markdownStyles = useMemo(() => createMarkdownStyles(theme), [rt.themeName]);
+```
+
+On adaptive system-theme changes, the hook can provide a light/dark theme update while an indirect runtime key is not the value that invalidates the memo. That leaves the library rendering stale colors. Assistant markdown hit this exact failure: the workspace shell switched to light, but assistant text and code spans kept the old dark-theme markdown style object.
+
+Prefer the hook theme itself, or explicit theme tokens, as the dependency:
+
+```tsx
+const { theme } = useUnistyles();
+const markdownStyles = useMemo(() => createMarkdownStyles(theme), [theme]);
+```
+
+If a style factory is cheap, skipping `useMemo` entirely is also fine.
+
+## Static Theme Imports
+
+Do not import `theme` from `@/styles/theme` for live UI colors. That export is a dark-theme compatibility default, so using it in render code leaves icons, placeholders, or third-party props pinned to dark colors in light mode.
+
+Wrap the icon (or other leaf component) with `withUnistyles` instead, so only that node re-renders when the theme changes:
+
+```tsx
+import { ChevronDown } from "lucide-react-native";
+import { StyleSheet, withUnistyles } from "react-native-unistyles";
+
+const ThemedChevronDown = withUnistyles(ChevronDown);
+
+const styles = StyleSheet.create((theme) => ({
+  icon: { color: theme.colors.foregroundMuted },
+}));
+
+<ThemedChevronDown size={theme.iconSize.md} style={styles.icon} />;
+```
+
+This is the dominant pattern in the app today (see `sidebar-workspace-list.tsx`, `message.tsx`, the workspace screens). Reserve `useUnistyles()` for the last-resort cases described at the top of this file. Importing `baseColors`, theme-name constants, or `type Theme` is fine when the value is intentionally static or type-only.
+
+## Reanimated `Animated.View` + Dynamic Styles Crashes
+
+Do not apply `StyleSheet.create((theme) => ...)` styles to a Reanimated `Animated.View`. Unistyles wraps styled components in a `<UnistylesComponent>` and patches native view props from C++ via the ShadowRegistry. Reanimated also reaches into the same native node from its worklet runtime. When a theme change fires, both systems try to mutate the same node and the app crashes with `Unable to find node on an unmounted component.` This was a real iOS sidebar crash on theme toggle (commit `4896cfe9`).
+
+Fix: keep static positioning on the `Animated.View` in plain React Native `StyleSheet`, and pass theme-dependent values (e.g. `backgroundColor`) as inline style from `useUnistyles()` — the inline path is acceptable here because no other escape works:
+
+```tsx
+import { StyleSheet as RNStyleSheet } from "react-native";
+import Animated from "react-native-reanimated";
+import { useUnistyles } from "react-native-unistyles";
+
+const positionStyles = RNStyleSheet.create({
+  sidebar: { position: "absolute", inset: 0, width: 280 },
+});
+
+function Sidebar() {
+  const { theme } = useUnistyles();
+  return (
+    <Animated.View
+      style={[positionStyles.sidebar, animatedStyle, { backgroundColor: theme.colors.surface1 }]}
+    />
+  );
+}
+```
+
+This is one of the rare places `useUnistyles()` is the right tool: there is no `withUnistyles(Animated.View)` equivalent, the affected component is small, and the alternative is a crash.
+
+## Adaptive Themes And Persisted Settings
+
+Unistyles [`initialTheme`](https://www.unistyl.es/v3/guides/theming#select-theme) and [`adaptiveThemes`](https://www.unistyl.es/v3/guides/theming#adaptive-themes) are mutually exclusive. `initialTheme` can be a string or a synchronous function, but it cannot wait on async storage.
+
+Paseo currently stores app settings in AsyncStorage and loads them through react-query. That means the app can mount under adaptive/system theme first, then switch after settings load:
+
+1. Unistyles config starts with `adaptiveThemes: true`.
+2. The device may report system light.
+3. Settings load a persisted non-auto preference, such as dark.
+4. The app calls `setAdaptiveThemes(false)` and `setTheme("dark")`.
+
+That brief transition is expected with the current storage model. It makes tracking-compatible styles important: anything mounted during the initial adaptive theme must update correctly after the persisted preference applies. [Issue #550](https://github.com/jpudysz/react-native-unistyles/issues/550) was a separate ScrollView sticky-header bug, but it is still useful context for why ScrollView theme updates deserve extra suspicion.
+
+If we ever need to avoid the transition entirely, store at least the theme preference in synchronous storage and configure Unistyles with `initialTheme`.
+
+## Debugging
+
+To inspect what the Babel plugin sees, temporarily enable [`debug: true`](https://www.unistyl.es/v3/other/babel-plugin#debug) in `packages/app/babel.config.js`:
+
+```js
+[
+  "react-native-unistyles/plugin",
+  {
+    root: "src",
+    debug: true,
+  },
+],
+```
+
+Then rebuild the bundle and look for lines such as:
+
+```text
+src/components/welcome-screen.tsx: styles.container: [Theme]
+```
+
+This only confirms that the stylesheet dependency was detected. The upstream debugging guide makes the same distinction: dependency detection is only one failure mode. It does not prove the style prop is registered on the native view you care about.
+
+For paint-layer bugs, use high-contrast probes:
+
+1. Paint each candidate layer a distinct color, such as root wrapper cyan, `ScrollView.style` yellow, and `contentContainerStyle` magenta.
+2. Cold-restart the app, not just Fast Refresh.
+3. Screenshot the simulator and sample pixels to see which color fills the area.
+4. Remove the probes before committing.
+
+The welcome-screen investigation used this approach to prove the white layer was the `ScrollView` content container. Deep-dive evidence is in [welcome-theme-split-research.md](/Users/moboudra/.paseo/notes/welcome-theme-split-research.md).
+
+## References
+
+- [Unistyles v3 documentation](https://www.unistyl.es/)
+- [Theming: initial theme, adaptive themes, and runtime theme changes](https://www.unistyl.es/v3/guides/theming)
+- [ScrollView Background Issue](https://www.unistyl.es/v3/tutorial/settings-screen#scrollview-background-issue)
+- [withUnistyles reference](https://www.unistyl.es/v3/references/with-unistyles)
+- [3rd-party view decision algorithm](https://www.unistyl.es/v3/references/3rd-party-views)
+- [Babel plugin debug option](https://www.unistyl.es/v3/other/babel-plugin#debug)
+- [Why my view doesn't update?](https://www.unistyl.es/v3/guides/why-my-view-doesnt-update)
+- [GitHub issue #550: ScrollView sticky-header theme updates](https://github.com/jpudysz/react-native-unistyles/issues/550)
+- [GitHub issue #817: `UnistylesRuntime.themeName` does not re-render](https://github.com/jpudysz/react-native-unistyles/issues/817)
+- [GitHub issue #1030: `Image.tintColor` and native style update edge case](https://github.com/jpudysz/react-native-unistyles/issues/1030)
+- [Local research note: welcome theme split](/Users/moboudra/.paseo/notes/welcome-theme-split-research.md)

knip.json

@@ -0,0 +1,96 @@
+{
+  "$schema": "./node_modules/knip/schema.json",
+  "workspaces": {
+    ".": {
+      "entry": ["scripts/**/*.{js,mjs,cjs,ts}"],
+      "project": ["scripts/**/*.{js,mjs,cjs,ts}"]
+    },
+    "packages/server": {
+      "entry": [
+        "src/server/index.ts",
+        "src/server/exports.ts",
+        "src/utils/tool-call-parsers.ts",
+        "src/shared/**/*.ts",
+        "src/client/**/*.ts",
+        "src/server/agent/agent-sdk-types.ts",
+        "src/server/agent/provider-manifest.ts",
+        "scripts/**/*.{ts,mts,mjs,cjs,js}",
+        "src/**/*.test.ts",
+        "src/**/*.test.tsx",
+        "src/**/*.e2e.ts",
+        "src/**/*.e2e.tsx"
+      ],
+      "project": ["src/**/*.{ts,tsx}", "scripts/**/*.{ts,mts,mjs,cjs,js}"]
+    },
+    "packages/app": {
+      "entry": [
+        "index.ts",
+        "app.config.js",
+        "babel.config.js",
+        "app/**/*.{ts,tsx}",
+        "src/**/*.test.{ts,tsx}",
+        "src/**/*.e2e.{ts,tsx}",
+        "src/**/*.native.{ts,tsx}",
+        "e2e/**/*.{ts,tsx}",
+        "playwright.config.{ts,js}",
+        "vitest.config.{ts,js}",
+        "test-stubs/**/*.ts"
+      ],
+      "project": ["**/*.{ts,tsx,js,jsx}"],
+      "paths": {
+        "@server/*": ["../server/src/*"]
+      },
+      "ignore": ["android/**", "ios/**", ".expo/**", "dist/**", "scripts/reset-project.js"]
+    },
+    "packages/cli": {
+      "entry": ["src/index.ts", "bin/paseo", "src/**/*.test.{ts,tsx}", "tests/**/*.{ts,tsx}"],
+      "project": ["src/**/*.{ts,tsx}", "tests/**/*.{ts,tsx}"]
+    },
+    "packages/relay": {
+      "entry": ["src/index.ts", "src/e2ee.ts", "src/cloudflare-adapter.ts", "src/**/*.test.ts"],
+      "project": ["src/**/*.ts"]
+    },
+    "packages/website": {
+      "entry": ["src/router.tsx", "vite.config.ts", "src/routes/**/*.tsx"],
+      "project": ["src/**/*.{ts,tsx}"],
+      "ignore": ["src/routeTree.gen.ts"]
+    },
+    "packages/desktop": {
+      "entry": ["src/main.ts", "src/preload.ts", "src/**/*.test.ts"],
+      "project": ["src/**/*.ts"]
+    },
+    "packages/highlight": {
+      "entry": ["src/index.ts", "src/**/*.test.ts"],
+      "project": ["src/**/*.ts"]
+    },
+    "packages/expo-two-way-audio": {
+      "entry": ["src/index.ts"],
+      "project": ["src/**/*.ts"]
+    }
+  },
+  "ignoreDependencies": [
+    "sherpa-onnx-node",
+    "@playwright/test",
+    "material-icon-theme",
+    "eas-cli",
+    "wait-on",
+    "concurrently",
+    "get-port-cli",
+    "patch-package",
+    "cross-env",
+    "expo-module-scripts",
+    "buffer",
+    "metro-config"
+  ],
+  "ignoreBinaries": [
+    "expo-module",
+    "xed",
+    "eas",
+    "playwright",
+    "wrangler",
+    "powershell",
+    "tsx",
+    "vitest",
+    "open"
+  ]
+}

lefthook.yml

@@ -0,0 +1,11 @@
+pre-commit:
+  parallel: true
+  jobs:
+    - name: format
+      glob: "*.{css,js,json,jsonc,jsx,md,ts,tsx,yaml,yml}"
+      run: npm run format:check:files -- {staged_files}
+    - name: lint
+      glob: "*.{js,jsx,ts,tsx}"
+      run: npm run lint -- {staged_files}
+    - name: typecheck
+      run: npm run typecheck

nix/module.nix

@@ -9,6 +9,10 @@ let
   cfg = config.services.paseo;
 in
 {
+  imports = [
+    (lib.mkRenamedOptionModule [ "services" "paseo" "allowedHosts" ] [ "services" "paseo" "hostnames" ])
+  ];
+
   options.services.paseo = {
     enable = lib.mkEnableOption "Paseo, a self-hosted daemon for AI coding agents";
 
@@ -58,12 +62,12 @@ in
       description = "Whether to open the firewall for the Paseo daemon port.";
     };
 
-    allowedHosts = lib.mkOption {
+    hostnames = lib.mkOption {
       type = lib.types.either (lib.types.enum [ true ]) (lib.types.listOf lib.types.str);
       default = [ ];
       example = [ ".example.com" "myhost.local" ];
       description = ''
-        Hosts allowed to connect to the Paseo daemon (DNS rebinding protection).
+        Hostnames the Paseo daemon accepts in the Host header (DNS rebinding protection).
         Localhost and IP addresses are always allowed by default.
 
         Use a leading dot to match a domain and all its subdomains
@@ -141,10 +145,10 @@ in
           "/run/wrappers/bin"
           "/nix/var/nix/profiles/default/bin"
         ]);
-      } // lib.optionalAttrs (cfg.allowedHosts == true) {
-        PASEO_ALLOWED_HOSTS = "true";
-      } // lib.optionalAttrs (lib.isList cfg.allowedHosts && cfg.allowedHosts != [ ]) {
-        PASEO_ALLOWED_HOSTS = lib.concatStringsSep "," cfg.allowedHosts;
+      } // lib.optionalAttrs (cfg.hostnames == true) {
+        PASEO_HOSTNAMES = "true";
+      } // lib.optionalAttrs (lib.isList cfg.hostnames && cfg.hostnames != [ ]) {
+        PASEO_HOSTNAMES = lib.concatStringsSep "," cfg.hostnames;
       } // cfg.environment;
 
       serviceConfig = {

nix/package.nix

@@ -42,7 +42,7 @@ buildNpmPackage rec {
 
   # To update: run `nix build` with lib.fakeHash, copy the `got:` hash.
   # CI auto-updates this when package-lock.json changes (see .github/workflows/).
-  npmDepsHash = "sha256-Cz3xidzBIWER4ktn3wWzT9PDm9PnipVA7XnsTQC440U=";
+  npmDepsHash = "sha256-ocBSTgfzPlucgcCaPRi3iYhxpZD6hTzD3b2lphSTtb4=";
 
   # Prevent onnxruntime-node's install script from running during automatic
   # npm rebuild (it tries to download from api.nuget.org, which fails in the sandbox).
@@ -123,7 +123,7 @@ buildNpmPackage rec {
     # Create wrapper for the server entry point (for systemd / direct use)
     mkdir -p $out/bin
     makeWrapper ${nodejs}/bin/node $out/bin/paseo-server \
-      --add-flags "$out/lib/paseo/packages/server/dist/server/server/index.js" \
+      --add-flags "$out/lib/paseo/packages/server/dist/scripts/supervisor-entrypoint.js" \
       --set NODE_ENV production
 
     # Create wrapper for the CLI

package.json

@@ -1,7 +1,26 @@
 {
   "name": "paseo",
-  "version": "0.1.38",
+  "version": "0.1.71",
   "private": true,
+  "description": "Paseo: voice-controlled development environment with OpenAI Realtime API",
+  "keywords": [
+    "development",
+    "mcp",
+    "openai",
+    "realtime",
+    "voice",
+    "voice-assistant"
+  ],
+  "homepage": "https://paseo.sh",
+  "license": "AGPL-3.0-or-later",
+  "author": {
+    "name": "Mohamed Boudra",
+    "email": "hello@moboudra.com"
+  },
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/getpaseo/paseo.git"
+  },
   "workspaces": [
     "packages/expo-two-way-audio",
     "packages/highlight",
@@ -14,18 +33,25 @@
   ],
   "scripts": {
     "dev": "./scripts/dev.sh",
+    "dev:win": "powershell ./scripts/dev.ps1",
     "dev:server": "npm run dev --workspace=@getpaseo/server",
     "dev:app": "npm run start --workspace=@getpaseo/app",
     "dev:website": "npm run dev --workspace=@getpaseo/website",
     "postinstall": "node scripts/postinstall-patches.mjs",
+    "prepare": "lefthook install --force",
     "build": "npm run build --workspaces --if-present",
     "build:highlight": "npm run build --workspace=@getpaseo/highlight",
     "build:daemon": "npm run build --workspace=@getpaseo/highlight && npm run build --workspace=@getpaseo/relay && npm run build --workspace=@getpaseo/server && npm run build --workspace=@getpaseo/cli",
     "typecheck": "npm run typecheck --workspaces --if-present",
     "typecheck:daemon": "npm run typecheck --workspace=@getpaseo/relay && npm run typecheck --workspace=@getpaseo/server && npm run typecheck --workspace=@getpaseo/cli",
     "test": "npm run test --workspaces --if-present",
-    "format": "biome format --write .",
-    "format:check": "biome format .",
+    "format": "oxfmt .",
+    "format:files": "oxfmt",
+    "format:check": "oxfmt --check .",
+    "format:check:files": "oxfmt --check",
+    "lint": "oxlint",
+    "lint:fix": "oxlint --fix",
+    "knip": "knip",
     "start": "npm run start --workspace=@getpaseo/server",
     "android": "npm run android --workspace=@getpaseo/app",
     "android:development": "npm run android:development --workspace=@getpaseo/app",
@@ -36,63 +62,54 @@
     "ios": "npm run ios --workspace=@getpaseo/app",
     "web": "npm run web --workspace=@getpaseo/app",
     "dev:desktop": "npm run dev --workspace=@getpaseo/desktop",
-    "build:desktop": "npm run version:sync-internal && npm run build:web --workspace=@getpaseo/app && npm run build --workspace=@getpaseo/desktop",
+    "dev:win:desktop": "npm run dev:win --workspace=@getpaseo/desktop",
+    "build:desktop": "npm run build:workspace-deps --workspace=@getpaseo/app && cd packages/app && cross-env PASEO_WEB_PLATFORM=electron npx expo export --platform web && cd ../.. && npm run build --workspace=@getpaseo/desktop --",
+    "db:query": "npm run db:query --workspace=@getpaseo/server --",
     "cli": "npx tsx packages/cli/src/index.js",
     "version": "npm run version:sync-internal && npm run release:prepare && git add -A",
     "version:sync-internal": "node scripts/sync-workspace-versions.mjs",
     "release:prepare": "npm install --workspaces --include-workspace-root",
-    "version:all:patch": "npm version patch --include-workspace-root --message \"chore(release): cut %s\"",
-    "version:all:minor": "npm version minor --include-workspace-root --message \"chore(release): cut %s\"",
-    "version:all:major": "npm version major --include-workspace-root --message \"chore(release): cut %s\"",
-    "release:check": "npm run release:prepare && npm run typecheck --workspace=@getpaseo/highlight && npm run typecheck --workspace=@getpaseo/relay && npm run typecheck --workspace=@getpaseo/server && npm run typecheck --workspace=@getpaseo/cli && npm run build --workspace=@getpaseo/highlight && npm run build --workspace=@getpaseo/relay && npm run build --workspace=@getpaseo/server && npm run build --workspace=@getpaseo/cli && npm pack --dry-run --workspace=@getpaseo/highlight && npm pack --dry-run --workspace=@getpaseo/relay && npm pack --dry-run --workspace=@getpaseo/server && npm pack --dry-run --workspace=@getpaseo/cli",
+    "version:all:patch": "node scripts/set-release-version.mjs --mode patch",
+    "version:all:minor": "node scripts/set-release-version.mjs --mode minor",
+    "version:all:major": "node scripts/set-release-version.mjs --mode major",
+    "version:all:beta:patch": "node scripts/set-release-version.mjs --mode beta-patch",
+    "version:all:beta:minor": "node scripts/set-release-version.mjs --mode beta-minor",
+    "version:all:beta:major": "node scripts/set-release-version.mjs --mode beta-major",
+    "version:all:beta:next": "node scripts/set-release-version.mjs --mode beta-next",
+    "version:all:promote": "node scripts/set-release-version.mjs --mode promote",
+    "release:check": "npm run release:prepare && npm run typecheck --workspace=@getpaseo/highlight && npm run build --workspace=@getpaseo/highlight && npm run typecheck --workspace=@getpaseo/relay && npm run build --workspace=@getpaseo/relay && npm run typecheck --workspace=@getpaseo/server && npm run build --workspace=@getpaseo/server && npm run typecheck --workspace=@getpaseo/cli && npm run build --workspace=@getpaseo/cli && npm pack --dry-run --workspace=@getpaseo/highlight && npm pack --dry-run --workspace=@getpaseo/relay && npm pack --dry-run --workspace=@getpaseo/server && npm pack --dry-run --workspace=@getpaseo/cli",
     "release:publish:dry-run": "npm publish --dry-run --workspace=@getpaseo/highlight --access public && npm publish --dry-run --workspace=@getpaseo/relay --access public && npm publish --dry-run --workspace=@getpaseo/server --access public && npm publish --dry-run --workspace=@getpaseo/cli --access public",
     "release:publish": "npm publish --workspace=@getpaseo/highlight --access public && npm publish --workspace=@getpaseo/relay --access public && npm publish --workspace=@getpaseo/server --access public && npm publish --workspace=@getpaseo/cli --access public",
     "release:push": "node scripts/push-current-release-tag.mjs",
-    "draft-release:push": "node scripts/push-current-release-tag.mjs --draft-release",
-    "draft-release:patch": "npm run release:check && npm run version:all:patch && npm run draft-release:push",
-    "draft-release:minor": "npm run release:check && npm run version:all:minor && npm run draft-release:push",
-    "draft-release:major": "npm run release:check && npm run version:all:major && npm run draft-release:push",
-    "release:finalize": "node scripts/finalize-current-release.mjs",
+    "release:beta:patch": "npm run release:check && npm run version:all:beta:patch && npm run release:push",
+    "release:beta:minor": "npm run release:check && npm run version:all:beta:minor && npm run release:push",
+    "release:beta:major": "npm run release:check && npm run version:all:beta:major && npm run release:push",
+    "release:beta:next": "npm run release:check && npm run version:all:beta:next && npm run release:push",
+    "release:promote": "npm run release:check && npm run version:all:promote && npm run release:publish && npm run release:push",
     "release:patch": "npm run release:check && npm run version:all:patch && npm run release:publish && npm run release:push",
     "release:minor": "npm run release:check && npm run version:all:minor && npm run release:publish && npm run release:push",
     "release:major": "npm run release:check && npm run version:all:major && npm run release:publish && npm run release:push"
   },
   "devDependencies": {
-    "@biomejs/biome": "^2.4.8",
+    "@types/ws": "^8.5.14",
+    "@typescript/native-preview": "7.0.0-dev.20260423.1",
     "concurrently": "^9.2.1",
+    "cross-env": "^10.1.0",
     "get-port-cli": "^3.0.0",
+    "js-yaml": "^4.1.1",
     "knip": "^5.82.1",
+    "lefthook": "^2.1.6",
+    "oxfmt": "0.46.0",
+    "oxlint": "1.61.0",
+    "oxlint-tsgolint": "^0.22.1",
     "patch-package": "^8.0.1",
-    "react": "19.1.4",
-    "react-dom": "19.1.4",
-    "typescript": "^5.9.3"
+    "playwright": "^1.56.1",
+    "typescript": "^5.9.3",
+    "ws": "^8.20.0"
   },
-  "description": "Paseo: voice-controlled development environment with OpenAI Realtime API",
-  "homepage": "https://paseo.sh",
-  "keywords": [
-    "openai",
-    "realtime",
-    "voice",
-    "voice-assistant",
-    "development",
-    "mcp"
-  ],
-  "author": {
-    "name": "Mohamed Boudra",
-    "email": "hello@moboudra.com"
-  },
-  "repository": {
-    "type": "git",
-    "url": "https://github.com/getpaseo/paseo.git"
-  },
-  "license": "AGPL-3.0-or-later",
   "overrides": {
     "lightningcss": "1.30.1",
-    "react": "19.1.4",
-    "react-dom": "19.1.4"
-  },
-  "dependencies": {
-    "@anthropic-ai/claude-agent-sdk": "^0.2.11",
-    "@modelcontextprotocol/sdk": "^1.27.1"
+    "react": "19.1.0",
+    "react-dom": "19.1.0"
   }
 }

packages/app/.eas/workflows/release-mobile.yml

@@ -4,6 +4,8 @@ on:
   push:
     tags:
       - "v*"
+      - "!v*-rc.*"
+      - "!v*-beta.*"
   workflow_dispatch: {}
 
 jobs:
@@ -36,3 +38,15 @@ jobs:
     params:
       build_id: ${{ needs.build_android.outputs.build_id }}
       profile: production
+
+  submit_ios_for_review:
+    name: Submit iOS for App Store review
+    needs: [submit_ios]
+    environment: production
+    runs_on: macos-medium
+    steps:
+      - uses: eas/checkout
+      - name: Install fastlane
+        run: bundle install
+      - name: Submit for review
+        run: bundle exec fastlane ios submit_review

packages/app/.eas/workflows/resubmit-ios-review.yml

@@ -0,0 +1,21 @@
+name: Resubmit iOS for App Store review
+
+# Standalone re-trigger for the App Store review submission step. The iOS
+# binary is already uploaded to TestFlight via the main release-mobile flow;
+# this workflow just runs the fastlane submit_review lane against the latest
+# TestFlight build, without rebuilding or re-uploading.
+
+on:
+  workflow_dispatch: {}
+
+jobs:
+  submit_ios_for_review:
+    name: Submit iOS for App Store review
+    environment: production
+    runs_on: macos-medium
+    steps:
+      - uses: eas/checkout
+      - name: Install fastlane
+        run: bundle install
+      - name: Submit for review
+        run: bundle exec fastlane ios submit_review

packages/app/.gitignore

@@ -37,6 +37,9 @@ yarn-error.*
 # typescript
 *.tsbuildinfo
 
+# vitest browser failure screenshots
+.vitest-screenshots/
+
 app-example
 
 # generated native folders

packages/app/Gemfile

@@ -0,0 +1,3 @@
+source "https://rubygems.org"
+
+gem "fastlane"

packages/app/e2e/00-sessions-empty.spec.ts

@@ -0,0 +1,37 @@
+// "00-" prefix is intentional: this file must sort before every other spec.
+// Sessions history is daemon-global — any agent created by a prior spec hides the empty state.
+// If the beforeAll probe below fails, a spec sorted before this file is creating agents.
+import { test } from "./fixtures";
+import {
+  connectArchiveTabDaemonClient,
+  expectSessionsEmptyState,
+  openSessions,
+} from "./helpers/archive-tab";
+
+test.describe("Sessions screen empty state", () => {
+  test.beforeAll(async () => {
+    const client = await connectArchiveTabDaemonClient();
+    try {
+      const history = await client.fetchAgentHistory({ page: { limit: 1 } });
+      if (history.entries.length > 0) {
+        throw new Error(
+          `Sessions empty-state precondition failed: daemon already has ${history.entries.length} agent(s). ` +
+            `Either a spec that sorts before 00-sessions-empty.spec.ts created agents, ` +
+            `or the daemon has stale history from a previous run.`,
+        );
+      }
+    } finally {
+      await client.close().catch(() => undefined);
+    }
+  });
+
+  test("shows empty placeholder when there is no session history", async ({
+    page,
+    withWorkspace,
+  }) => {
+    const workspace = await withWorkspace({ prefix: "sessions-empty-" });
+    await workspace.navigateTo();
+    await openSessions(page);
+    await expectSessionsEmptyState(page);
+  });
+});

packages/app/e2e/acp-provider-catalog.spec.ts

@@ -0,0 +1,33 @@
+import { test } from "./fixtures";
+import { gotoAppShell, openSettings } from "./helpers/app";
+import {
+  expectProviderInstalledInSettings,
+  installAcpCatalogProvider,
+  openAddProviderModal,
+  openSettingsHost,
+} from "./helpers/settings";
+
+const ACP_PROVIDER = {
+  id: "hermes",
+  name: "Hermes",
+};
+
+function getSeededServerId(): string {
+  const serverId = process.env.E2E_SERVER_ID;
+  if (!serverId) {
+    throw new Error("E2E_SERVER_ID is not set (expected from Playwright globalSetup).");
+  }
+  return serverId;
+}
+
+test.describe("ACP provider catalog", () => {
+  test("adds a catalog provider from settings", async ({ page }) => {
+    await gotoAppShell(page);
+    await openSettings(page);
+    await openSettingsHost(page, getSeededServerId());
+    await openAddProviderModal(page);
+
+    await installAcpCatalogProvider(page, ACP_PROVIDER.name);
+    await expectProviderInstalledInSettings(page, ACP_PROVIDER.name);
+  });
+});

packages/app/e2e/agent-stream-ui.spec.ts

@@ -0,0 +1,45 @@
+import { test } from "./fixtures";
+import {
+  awaitAssistantMessage,
+  expectAgentIdle,
+  expectInlineWorkingIndicator,
+  expectTurnCopyButton,
+  expectScrollFollowsNewContent,
+} from "./helpers/agent-stream";
+import { startRunningMockAgent } from "./helpers/composer";
+
+test.describe("Agent stream UI", () => {
+  test("auto-scroll sticks to bottom across token bursts", async ({ page }) => {
+    test.setTimeout(120_000);
+    const { client, repo } = await startRunningMockAgent(page, {
+      prefix: "stream-scroll-",
+      model: "one-minute-stream",
+      prompt: "Stream for auto-scroll test.",
+    });
+    try {
+      await awaitAssistantMessage(page);
+      await expectScrollFollowsNewContent(page);
+    } finally {
+      await client.close();
+      await repo.cleanup();
+    }
+  });
+
+  test("working-indicator transitions to copy-button when stream ends", async ({ page }) => {
+    test.setTimeout(60_000);
+    const { client, repo } = await startRunningMockAgent(page, {
+      prefix: "stream-indicator-",
+      model: "ten-second-stream",
+      prompt: "Stream briefly for indicator transition test.",
+    });
+    try {
+      await awaitAssistantMessage(page);
+      await expectInlineWorkingIndicator(page);
+      await expectAgentIdle(page, 30_000);
+      await expectTurnCopyButton(page);
+    } finally {
+      await client.close();
+      await repo.cleanup();
+    }
+  });
+});

packages/app/e2e/archive-tab.spec.ts

@@ -0,0 +1,129 @@
+import { randomUUID } from "node:crypto";
+import { test } from "./fixtures";
+import { createTempGitRepo } from "./helpers/workspace";
+import {
+  archiveAgentFromDaemon,
+  archiveAgentFromSessions,
+  clickSessionRow,
+  closeWorkspaceAgentTab,
+  connectArchiveTabDaemonClient,
+  createIdleAgent,
+  expectArchivedAgentFocused,
+  expectSessionRowArchived,
+  expectSessionRowVisible,
+  expectWorkspaceArchiveOutcome,
+  expectWorkspaceTabHidden,
+  openSessions,
+  openWorkspaceWithAgents,
+  primeAdditionalPage,
+  resetSeededPageState,
+  reloadWorkspace,
+} from "./helpers/archive-tab";
+
+test.describe("Archive tab reconciliation", () => {
+  let client: Awaited<ReturnType<typeof connectArchiveTabDaemonClient>>;
+  let tempRepo: { path: string; cleanup: () => Promise<void> };
+
+  test.describe.configure({ timeout: 300_000 });
+
+  test.beforeAll(async () => {
+    tempRepo = await createTempGitRepo("archive-tab-");
+    client = await connectArchiveTabDaemonClient();
+  });
+
+  test.afterAll(async () => {
+    await client?.close().catch(() => undefined);
+    await tempRepo?.cleanup();
+  });
+
+  test("non-UI archive prunes the archived tab across open pages and reload", async ({ page }) => {
+    const archived = await createIdleAgent(client, {
+      cwd: tempRepo.path,
+      title: `cli-archive-${randomUUID().slice(0, 8)}`,
+    });
+    const surviving = await createIdleAgent(client, {
+      cwd: tempRepo.path,
+      title: `cli-control-${randomUUID().slice(0, 8)}`,
+    });
+    const passivePage = await page.context().newPage();
+
+    try {
+      await primeAdditionalPage(passivePage);
+      await resetSeededPageState(page);
+      await resetSeededPageState(passivePage);
+      await openSessions(page);
+      await expectSessionRowVisible(page, archived.title);
+      await expectSessionRowVisible(page, surviving.title);
+      await openSessions(passivePage);
+      await expectSessionRowVisible(passivePage, archived.title);
+      await expectSessionRowVisible(passivePage, surviving.title);
+      await openWorkspaceWithAgents(page, [archived, surviving]);
+      await openWorkspaceWithAgents(passivePage, [archived, surviving]);
+      await archiveAgentFromDaemon(client, archived.id);
+      await expectWorkspaceArchiveOutcome(page, {
+        archivedAgentId: archived.id,
+        survivingAgentId: surviving.id,
+      });
+      await expectWorkspaceArchiveOutcome(passivePage, {
+        archivedAgentId: archived.id,
+        survivingAgentId: surviving.id,
+      });
+      await reloadWorkspace(passivePage, tempRepo.path);
+      await expectWorkspaceTabHidden(passivePage, archived.id);
+    } finally {
+      await passivePage.close();
+    }
+  });
+
+  test("Sessions archive prunes the archived tab across open pages", async ({ page }) => {
+    const archived = await createIdleAgent(client, {
+      cwd: tempRepo.path,
+      title: `ui-archive-${randomUUID().slice(0, 8)}`,
+    });
+    const surviving = await createIdleAgent(client, {
+      cwd: tempRepo.path,
+      title: `ui-control-${randomUUID().slice(0, 8)}`,
+    });
+    const passivePage = await page.context().newPage();
+
+    try {
+      await primeAdditionalPage(passivePage);
+      await resetSeededPageState(page);
+      await resetSeededPageState(passivePage);
+      await openWorkspaceWithAgents(page, [archived, surviving]);
+      await openWorkspaceWithAgents(passivePage, [archived, surviving]);
+      await openSessions(page);
+      await archiveAgentFromSessions(page, { agentId: archived.id, title: archived.title });
+      await reloadWorkspace(page, tempRepo.path);
+      await expectWorkspaceTabHidden(page, archived.id);
+      await expectWorkspaceArchiveOutcome(passivePage, {
+        archivedAgentId: archived.id,
+        survivingAgentId: surviving.id,
+      });
+    } finally {
+      await passivePage.close();
+    }
+  });
+
+  test("clicking an archived session reopens its closed tab focused", async ({ page }) => {
+    const archived = await createIdleAgent(client, {
+      cwd: tempRepo.path,
+      title: `reopen-archived-${randomUUID().slice(0, 8)}`,
+    });
+    const surviving = await createIdleAgent(client, {
+      cwd: tempRepo.path,
+      title: `reopen-control-${randomUUID().slice(0, 8)}`,
+    });
+
+    await resetSeededPageState(page);
+    await openWorkspaceWithAgents(page, [archived, surviving]);
+    await closeWorkspaceAgentTab(page, archived.id);
+    await archiveAgentFromDaemon(client, archived.id);
+    await openSessions(page);
+    await expectSessionRowArchived(page, archived.title);
+
+    await clickSessionRow(page, archived.title);
+
+    await expectArchivedAgentFocused(page, archived.id);
+  });
+});

packages/app/e2e/codex-plan-approval.spec.ts

@@ -0,0 +1,61 @@
+import { expect, test } from "./fixtures";
+import { buildHostWorkspaceRoute } from "@/utils/host-routes";
+import { allowPermission, waitForPermissionPrompt } from "./helpers/permissions";
+import { connectTerminalClient } from "./helpers/terminal-perf";
+import { createTempGitRepo } from "./helpers/workspace";
+
+function getServerId(): string {
+  const serverId = process.env.E2E_SERVER_ID;
+  if (!serverId) {
+    throw new Error("E2E_SERVER_ID is not set.");
+  }
+  return serverId;
+}
+
+test.describe("Codex plan approval", () => {
+  test("shows a single actionable plan panel and removes it after implementation starts", async ({
+    page,
+  }) => {
+    test.setTimeout(180_000);
+
+    const repo = await createTempGitRepo("codex-plan-approval-");
+    const client = await connectTerminalClient();
+
+    try {
+      const workspaceResult = await client.openProject(repo.path);
+      if (!workspaceResult.workspace) {
+        throw new Error(workspaceResult.error ?? `Failed to open project ${repo.path}`);
+      }
+
+      const agent = await client.createAgent({
+        provider: "mock",
+        cwd: repo.path,
+        title: "Codex plan approval e2e",
+        modeId: "load-test",
+        model: "ten-second-stream",
+        initialPrompt: "Emit synthetic plan approval.",
+      });
+
+      const agentUrl = `${buildHostWorkspaceRoute(
+        getServerId(),
+        repo.path,
+      )}?open=${encodeURIComponent(`agent:${agent.id}`)}`;
+      await page.goto(agentUrl);
+
+      await waitForPermissionPrompt(page, 120_000);
+
+      await expect(page.getByTestId("permission-plan-card")).toHaveCount(1);
+      await expect(page.getByTestId("timeline-plan-card")).toHaveCount(0);
+
+      await allowPermission(page);
+
+      await expect(page.getByTestId("permission-plan-card")).toHaveCount(0, {
+        timeout: 30_000,
+      });
+      await expect(page.getByTestId("timeline-plan-card")).toHaveCount(0);
+    } finally {
+      await client.close();
+      await repo.cleanup();
+    }
+  });
+});

packages/app/e2e/composer-attachments.spec.ts

@@ -0,0 +1,284 @@
+import { expect, test } from "./fixtures";
+import { clickNewChat } from "./helpers/launcher";
+import { expectComposerVisible } from "./helpers/composer";
+import { expectAgentIdle } from "./helpers/agent-stream";
+import {
+  openAttachmentMenu,
+  openGithubPickerFromMenu,
+  attachImageFromMenu,
+  expectAttachmentPill,
+  removeAttachmentPill,
+  openImageLightbox,
+  closeImageLightbox,
+  pressInterruptShortcut,
+  expectComposerDraft,
+  expectComposerDisabled,
+  expectComposerEditable,
+  expectAttachButtonDisabled,
+  fillComposerDraft,
+  sendDraftToQueue,
+  expectQueuedMessageButton,
+  startRunningMockAgent,
+  selectGithubOption,
+  expectGithubAttachmentPill,
+  openGithubWorkspace,
+} from "./helpers/composer";
+import {
+  connectNewWorkspaceDaemonClient,
+  delayBrowserAgentCreatedStatus,
+  openNewWorkspaceComposer,
+  openProjectViaDaemon,
+} from "./helpers/new-workspace";
+import { gotoAppShell } from "./helpers/app";
+import { waitForSidebarHydration, switchWorkspaceViaSidebar } from "./helpers/workspace-ui";
+import { createTempGitRepo } from "./helpers/workspace";
+import { hasGithubAuth, createTempGithubRepo } from "./helpers/github-fixtures";
+
+const MINIMAL_PNG = Buffer.from(
+  "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR42mNk+M9QDwADhgGAWjR9awAAAABJRU5ErkJggg==",
+  "base64",
+);
+
+const TEST_IMAGE = { name: "test.png", mimeType: "image/png", buffer: MINIMAL_PNG };
+
+test.describe("Composer attachments", () => {
+  test("Plus menu shows image and GitHub options", async ({ page, withWorkspace }) => {
+    test.setTimeout(60_000);
+    const workspace = await withWorkspace({ prefix: "attach-plus-" });
+    await workspace.navigateTo();
+    await clickNewChat(page);
+    await expectComposerVisible(page);
+
+    await openAttachmentMenu(page);
+
+    await expect(page.getByTestId("message-input-attachment-menu-item-image")).toBeVisible();
+    await expect(page.getByTestId("message-input-attachment-menu-item-github")).toBeVisible();
+  });
+
+  test("GitHub combobox does not render until the picker is opened", async ({
+    page,
+    withWorkspace,
+  }) => {
+    test.setTimeout(60_000);
+    const workspace = await withWorkspace({ prefix: "attach-gh-lazy-" });
+    await workspace.navigateTo();
+    await clickNewChat(page);
+    await expectComposerVisible(page);
+
+    await expect(page.getByTestId("combobox-desktop-container")).not.toBeVisible();
+
+    await openGithubPickerFromMenu(page);
+
+    await expect(page.getByPlaceholder("Search issues and PRs...")).toBeVisible({ timeout: 5_000 });
+    await expect(
+      page.getByTestId("combobox-empty-text").or(page.getByText("Searching...")),
+    ).toBeVisible({ timeout: 10_000 });
+  });
+
+  test("GitHub issue attachment pill visible after search and selection", async ({ page }) => {
+    test.setTimeout(120_000);
+    if (!hasGithubAuth()) {
+      test.skip(true, "GitHub auth not available in this environment");
+    }
+
+    const ghRepo = await createTempGithubRepo({
+      category: "attach-issue",
+      issues: [{ title: "fix: attach-issue-unique-alpha" }],
+      prs: [{ title: "feat: attach-issue-dummy-pr", state: "open" }],
+    });
+    const handle = await openGithubWorkspace(page, ghRepo.prs[0].localPath);
+    try {
+      await clickNewChat(page);
+      await expectComposerVisible(page);
+
+      await selectGithubOption(
+        page,
+        "attach-issue-unique-alpha",
+        `issue:${ghRepo.issues[0].number}`,
+      );
+
+      await expectGithubAttachmentPill(page, {
+        number: ghRepo.issues[0].number,
+        title: ghRepo.issues[0].title,
+      });
+    } finally {
+      await handle.cleanup();
+      await ghRepo.cleanup();
+    }
+  });
+
+  test("GitHub PR attachment pill visible after search and selection", async ({ page }) => {
+    test.setTimeout(120_000);
+    if (!hasGithubAuth()) {
+      test.skip(true, "GitHub auth not available in this environment");
+    }
+
+    const ghRepo = await createTempGithubRepo({
+      category: "attach-pr",
+      prs: [{ title: "feat: attach-pr-unique-beta", state: "open" }],
+    });
+    const handle = await openGithubWorkspace(page, ghRepo.prs[0].localPath);
+    try {
+      await clickNewChat(page);
+      await expectComposerVisible(page);
+
+      await selectGithubOption(page, "attach-pr-unique-beta", `pr:${ghRepo.prs[0].number}`);
+
+      await expectGithubAttachmentPill(page, {
+        number: ghRepo.prs[0].number,
+        title: ghRepo.prs[0].title,
+      });
+    } finally {
+      await handle.cleanup();
+      await ghRepo.cleanup();
+    }
+  });
+
+  test.fixme("workspace-review pill suppresses on X-click and reappears after send", async () => {
+    // The review attachment is created via InlineReviewEditor in surface.tsx (addComment action).
+    // Automating this requires: a workspace with staged changes, navigating to the diff panel,
+    // hovering the gutter "+" button, typing a comment, and submitting. A dedicated
+    // helpers/review.ts with addInlineReviewComment(page, filePath, lineNumber, comment) is
+    // needed before this can be exercised end-to-end.
+  });
+
+  test.fixme("browser-element attachment pill is created from Electron webview selection", async () => {
+    // The browser-element attachment is only created in browser-pane.electron.tsx via DOM
+    // element selection in the Electron webview. It is not exercisable in headless Chromium E2E.
+  });
+
+  test("image lightbox opens on pill click and closes on Escape", async ({
+    page,
+    withWorkspace,
+  }) => {
+    test.setTimeout(60_000);
+    const workspace = await withWorkspace({ prefix: "attach-lightbox-" });
+    await workspace.navigateTo();
+    await clickNewChat(page);
+    await expectComposerVisible(page);
+
+    await attachImageFromMenu(page, TEST_IMAGE);
+    await expectAttachmentPill(page, "composer-image-attachment-pill");
+
+    await openImageLightbox(page);
+    await closeImageLightbox(page);
+  });
+
+  test("image attachment pill renders after file is selected", async ({ page, withWorkspace }) => {
+    test.setTimeout(60_000);
+    const workspace = await withWorkspace({ prefix: "attach-pill-" });
+    await workspace.navigateTo();
+    await clickNewChat(page);
+    await expectComposerVisible(page);
+
+    await attachImageFromMenu(page, TEST_IMAGE);
+
+    await expectAttachmentPill(page, "composer-image-attachment-pill");
+  });
+
+  test("clicking the X on an image pill removes it", async ({ page, withWorkspace }) => {
+    test.setTimeout(60_000);
+    const workspace = await withWorkspace({ prefix: "attach-remove-" });
+    await workspace.navigateTo();
+    await clickNewChat(page);
+    await expectComposerVisible(page);
+
+    await attachImageFromMenu(page, TEST_IMAGE);
+    await expectAttachmentPill(page, "composer-image-attachment-pill");
+
+    await removeAttachmentPill(page, "composer-image-attachment-pill", "Remove image attachment");
+
+    await expect(page.getByTestId("composer-image-attachment-pill")).toHaveCount(0, {
+      timeout: 5_000,
+    });
+  });
+
+  test("submitting while agent is running queues the message and clears the draft", async ({
+    page,
+  }) => {
+    test.setTimeout(120_000);
+    const { client, repo } = await startRunningMockAgent(page, {
+      prefix: "attach-queue-",
+      model: "one-minute-stream",
+      prompt: "Stay running for queue test.",
+    });
+    try {
+      await fillComposerDraft(page, "queued draft text");
+      await sendDraftToQueue(page);
+
+      await expectQueuedMessageButton(page);
+      await expectComposerDraft(page, "");
+    } finally {
+      await client.close();
+      await repo.cleanup();
+    }
+  });
+
+  test("Escape interrupt cancels the running agent and preserves composer draft", async ({
+    page,
+  }) => {
+    test.setTimeout(120_000);
+    const { client, repo } = await startRunningMockAgent(page, {
+      prefix: "attach-interrupt-",
+      model: "ten-second-stream",
+      prompt: "Stay running for interrupt test.",
+    });
+    try {
+      await fillComposerDraft(page, "preserve me");
+      await pressInterruptShortcut(page);
+
+      await expectAgentIdle(page, 15_000);
+      await expectComposerDraft(page, "preserve me");
+    } finally {
+      await client.close();
+      await repo.cleanup();
+    }
+  });
+
+  test("composer is locked while new workspace agent is being created", async ({ page }) => {
+    test.setTimeout(120_000);
+    const serverId = process.env.E2E_SERVER_ID;
+    if (!serverId) throw new Error("E2E_SERVER_ID is not set.");
+
+    const repo = await createTempGitRepo("attach-lock-");
+    const agentCreatedDelay = await delayBrowserAgentCreatedStatus(page);
+    const daemonClient = await connectNewWorkspaceDaemonClient();
+
+    try {
+      const openedProject = await openProjectViaDaemon(daemonClient, repo.path);
+
+      await gotoAppShell(page);
+      await waitForSidebarHydration(page);
+      await switchWorkspaceViaSidebar({
+        page,
+        serverId,
+        targetWorkspacePath: openedProject.workspaceId,
+      });
+
+      await openNewWorkspaceComposer(page, {
+        projectKey: openedProject.projectKey,
+        projectDisplayName: openedProject.projectDisplayName,
+      });
+      await fillComposerDraft(page, "lock test prompt");
+      const createButton = page
+        .getByTestId("message-input-root")
+        .getByRole("button", { name: "Create" });
+      await expect(createButton).toBeVisible({ timeout: 30_000 });
+      await createButton.click();
+
+      await agentCreatedDelay.waitForCreateRequest();
+      await agentCreatedDelay.waitForDelayedCreatedStatus();
+
+      await expectComposerDisabled(page);
+      await expectAttachButtonDisabled(page);
+
+      agentCreatedDelay.release();
+
+      await expectComposerEditable(page);
+    } finally {
+      agentCreatedDelay.release();
+      await daemonClient.close().catch(() => undefined);
+      await repo.cleanup();
+    }
+  });
+});

packages/app/e2e/desktop-updates.spec.ts

@@ -0,0 +1,151 @@
+import { test } from "./fixtures";
+import { gotoAppShell } from "./helpers/app";
+import {
+  loadRealDaemonState,
+  injectDesktopBridge,
+  openDesktopSettings,
+  expectUpdateBanner,
+  clickInstallUpdate,
+  expectInstallInProgress,
+  interceptDaemonManagementConfirmDialog,
+  toggleDaemonManagement,
+  expectDaemonManagementConfirmDialog,
+  expectDaemonManagementEnabled,
+  expectDaemonManagementDisabled,
+  expectDaemonStatusPid,
+  expectDaemonStatusLogPath,
+  expectDaemonStatusVersion,
+} from "./helpers/desktop-updates";
+
+function getSeededServerId(): string {
+  const serverId = process.env.E2E_SERVER_ID;
+  if (!serverId) {
+    throw new Error("E2E_SERVER_ID is not set (expected from Playwright globalSetup).");
+  }
+  return serverId;
+}
+
+// No Playwright Electron runner exists; we simulate the desktop bridge via
+// addInitScript so Electron-gated UI activates without a real Electron process.
+test.describe("Desktop updates", () => {
+  test("update banner appears in the sidebar when an app update is available", async ({ page }) => {
+    await injectDesktopBridge(page, {
+      serverId: getSeededServerId(),
+      updateAvailable: true,
+      latestVersion: "1.2.3",
+    });
+    await gotoAppShell(page);
+
+    await expectUpdateBanner(page, "1.2.3");
+  });
+
+  test("clicking install shows the installing state on the callout", async ({ page }) => {
+    await injectDesktopBridge(page, {
+      serverId: getSeededServerId(),
+      updateAvailable: true,
+      latestVersion: "1.2.3",
+      slowInstall: true,
+    });
+    await gotoAppShell(page);
+
+    await expectUpdateBanner(page, "1.2.3");
+    await clickInstallUpdate(page);
+    await expectInstallInProgress(page);
+  });
+});
+
+test.describe("Desktop daemon management", () => {
+  test("disabling built-in daemon management shows confirm dialog with correct copy", async ({
+    page,
+  }) => {
+    const serverId = getSeededServerId();
+    await injectDesktopBridge(page, {
+      serverId,
+      manageBuiltInDaemon: true,
+      confirmShouldAccept: false,
+    });
+    await gotoAppShell(page);
+    await openDesktopSettings(page, serverId);
+
+    const dialogArgs = await interceptDaemonManagementConfirmDialog(page);
+    expectDaemonManagementConfirmDialog(dialogArgs);
+
+    await expectDaemonManagementEnabled(page);
+  });
+
+  test("cancelling the confirm dialog leaves the daemon management toggle on", async ({ page }) => {
+    const serverId = getSeededServerId();
+    await injectDesktopBridge(page, {
+      serverId,
+      manageBuiltInDaemon: true,
+      confirmShouldAccept: false,
+    });
+    await gotoAppShell(page);
+    await openDesktopSettings(page, serverId);
+
+    await expectDaemonManagementEnabled(page);
+    await toggleDaemonManagement(page, "disable");
+    await expectDaemonManagementEnabled(page);
+  });
+
+  test("confirming the dialog disables built-in daemon management", async ({ page }) => {
+    const serverId = getSeededServerId();
+    await injectDesktopBridge(page, {
+      serverId,
+      manageBuiltInDaemon: true,
+      confirmShouldAccept: true,
+    });
+    await gotoAppShell(page);
+    await openDesktopSettings(page, serverId);
+
+    await toggleDaemonManagement(page, "disable");
+
+    await expectDaemonManagementDisabled(page);
+  });
+
+  test("daemon status panel renders version, PID, and log path from the real daemon", async ({
+    page,
+  }) => {
+    const serverId = getSeededServerId();
+    const realState = await loadRealDaemonState();
+    await injectDesktopBridge(page, {
+      serverId,
+      manageBuiltInDaemon: false,
+      daemonPid: realState.pid,
+      daemonVersion: realState.version,
+      daemonLogPath: realState.logPath,
+    });
+    await gotoAppShell(page);
+    await openDesktopSettings(page, serverId);
+
+    await expectDaemonStatusVersion(page, realState.version);
+    await expectDaemonStatusPid(page, realState.pid);
+    await expectDaemonStatusLogPath(page, realState.logPath);
+  });
+
+  test("stopping and restarting the daemon updates the PID", async ({ page }) => {
+    const serverId = getSeededServerId();
+    const realState = await loadRealDaemonState();
+    await injectDesktopBridge(page, {
+      serverId,
+      manageBuiltInDaemon: true,
+      daemonPid: realState.pid,
+      daemonVersion: realState.version,
+      daemonLogPath: realState.logPath,
+      confirmShouldAccept: true,
+    });
+    await gotoAppShell(page);
+    await openDesktopSettings(page, serverId);
+
+    await expectDaemonStatusPid(page, realState.pid);
+
+    await toggleDaemonManagement(page, "disable");
+    await expectDaemonManagementDisabled(page);
+    await expectDaemonStatusPid(page, null);
+
+    await toggleDaemonManagement(page, "enable");
+    await expectDaemonManagementEnabled(page);
+    const newPid = realState.pid !== null ? realState.pid + 1000 : 11000;
+    await expectDaemonStatusPid(page, newPid);
+  });
+});

packages/app/e2e/file-explorer-collapse.spec.ts

@@ -0,0 +1,61 @@
+import { test } from "./fixtures";
+import {
+  collapseFolder,
+  expandFolder,
+  expectExplorerEntryHidden,
+  expectExplorerEntryVisible,
+  expectFileTabOpen,
+  openFileExplorer,
+  openFileFromExplorer,
+} from "./helpers/file-explorer";
+import { gotoWorkspace } from "./helpers/launcher";
+import { createTempGitRepo } from "./helpers/workspace";
+import {
+  connectWorkspaceSetupClient,
+  type WorkspaceSetupDaemonClient,
+} from "./helpers/workspace-setup";
+
+let tempRepo: { path: string; cleanup: () => Promise<void> };
+let workspaceId: string;
+let seedClient: WorkspaceSetupDaemonClient;
+
+test.beforeAll(async () => {
+  tempRepo = await createTempGitRepo("file-explorer-collapse-", {
+    files: [
+      { path: "assets/logo.png", content: "image bytes for explorer e2e\n" },
+      { path: "docs/guide.md", content: "# Guide\n" },
+    ],
+  });
+  seedClient = await connectWorkspaceSetupClient();
+  const result = await seedClient.openProject(tempRepo.path);
+  if (!result.workspace) {
+    throw new Error(result.error ?? "Failed to seed workspace");
+  }
+  workspaceId = result.workspace.id;
+});
+
+test.afterAll(async () => {
+  await seedClient?.close();
+  await tempRepo?.cleanup();
+});
+
+test.describe("File explorer collapse", () => {
+  test("collapses an opened image file parent folder and still expands other folders", async ({
+    page,
+  }) => {
+    await gotoWorkspace(page, workspaceId);
+    await openFileExplorer(page);
+
+    await expandFolder(page, "assets");
+    await expectExplorerEntryVisible(page, "logo.png");
+
+    await openFileFromExplorer(page, "logo.png");
+    await expectFileTabOpen(page, "assets/logo.png");
+
+    await collapseFolder(page, "assets");
+    await expectExplorerEntryHidden(page, "logo.png");
+
+    await expandFolder(page, "docs");
+    await expectExplorerEntryVisible(page, "guide.md");
+  });
+});

packages/app/e2e/fixtures.ts

@@ -1,111 +1,113 @@
 import { test as base, expect, type Page } from "@playwright/test";
 import { buildCreateAgentPreferences, buildSeededHost } from "./helpers/daemon-registry";
+import { createWithWorkspace, type WithWorkspace } from "./helpers/with-workspace";
 
-// Extend base test to provide dynamic baseURL from global-setup
-const test = base.extend({
-  baseURL: async ({}, use) => {
+// Test setup is wired through an `auto: true` fixture rather than `test.beforeEach`.
+// `test.beforeEach` declared at the top level of a non-test fixture file is unreliable
+// across spec-file boundaries — Playwright sometimes skips it for the first test of a
+// subsequent spec when multiple specs run in the same worker. Auto fixtures run
+// reliably for every test that uses this `test` object.
+const test = base.extend<{ paseoE2ESetup: void; withWorkspace: WithWorkspace }>({
+  baseURL: async ({}, provide) => {
     const metroPort = process.env.E2E_METRO_PORT;
     if (!metroPort) {
       throw new Error("E2E_METRO_PORT not set - globalSetup must run first");
     }
-    await use(`http://localhost:${metroPort}`);
+    await provide(`http://localhost:${metroPort}`);
+  },
+  paseoE2ESetup: [
+    async ({ page }, provide, testInfo) => {
+      const daemonPort = process.env.E2E_DAEMON_PORT;
+      const metroPort = process.env.E2E_METRO_PORT;
+      if (!daemonPort) {
+        throw new Error(
+          "E2E_DAEMON_PORT is not set. Refusing to run e2e against the default daemon (e.g. localhost:6767). " +
+            "Ensure Playwright `globalSetup` starts the e2e daemon and exports E2E_DAEMON_PORT.",
+        );
+      }
+      if (daemonPort === "6767") {
+        throw new Error(
+          "E2E_DAEMON_PORT is 6767. Refusing to run e2e against the default local daemon. " +
+            "Fix Playwright globalSetup to start an isolated test daemon and export its port.",
+        );
+      }
+      if (!metroPort) {
+        throw new Error(
+          "E2E_METRO_PORT is not set. Ensure Playwright `globalSetup` starts Metro and exports E2E_METRO_PORT.",
+        );
+      }
+
+      // Hard guardrail: never allow tests to hit the developer's default daemon.
+      // This blocks both HTTP and WS attempts to :6767 (before any navigation).
+      await page.route(/:(6767)\b/, (route) => route.abort());
+      await page.routeWebSocket(/:(6767)\b/, async (ws) => {
+        await ws.close({ code: 1008, reason: "Blocked connection to localhost:6767 during e2e." });
+      });
+
+      const entries: string[] = [];
+
+      page.on("console", (message) => {
+        entries.push(`[console:${message.type()}] ${message.text()}`);
+      });
+
+      page.on("pageerror", (error) => {
+        entries.push(`[pageerror] ${error.message}`);
+      });
+
+      const nowIso = new Date().toISOString();
+      const seedNonce = Math.random().toString(36).slice(2);
+      const serverId = process.env.E2E_SERVER_ID;
+      if (!serverId) {
+        throw new Error("E2E_SERVER_ID is not set - expected from Playwright globalSetup.");
+      }
+      const testDaemon = buildSeededHost({
+        serverId,
+        endpoint: `127.0.0.1:${daemonPort}`,
+        nowIso,
+      });
+      const createAgentPreferences = buildCreateAgentPreferences(testDaemon.serverId);
+
+      await page.addInitScript(
+        ({ daemon, preferences, seedNonce: nonce }) => {
+          // `addInitScript` runs on every navigation (including reloads). Some tests intentionally
+          // override storage and reload; they can opt out of seeding for the *next* navigation by
+          // setting this flag before the reload.
+          const disableOnceKey = "@paseo:e2e-disable-default-seed-once";
+          const disableValue = localStorage.getItem(disableOnceKey);
+          if (disableValue) {
+            localStorage.removeItem(disableOnceKey);
+            if (disableValue === nonce) {
+              return;
+            }
+          }
+
+          localStorage.setItem("@paseo:e2e", "1");
+          localStorage.setItem("@paseo:e2e-seed-nonce", nonce);
+
+          // Hard-reset anything that could point to a developer's real daemon.
+          localStorage.setItem("@paseo:daemon-registry", JSON.stringify([daemon]));
+          localStorage.removeItem("@paseo:settings");
+          localStorage.setItem("@paseo:create-agent-preferences", JSON.stringify(preferences));
+        },
+        { daemon: testDaemon, preferences: createAgentPreferences, seedNonce },
+      );
+
+      await provide();
+
+      if (entries.length > 0 && testInfo.status !== testInfo.expectedStatus) {
+        await testInfo.attach("browser-console", {
+          body: entries.join("\n"),
+          contentType: "text/plain",
+        });
+      }
+    },
+    { auto: true },
+  ],
+  withWorkspace: async ({ page }, provide) => {
+    const handle = createWithWorkspace(page);
+    await provide(handle.withWorkspace);
+    await handle.cleanup();
   },
 });
 
-const consoleEntries = new WeakMap<Page, string[]>();
-
-test.beforeEach(async ({ page }) => {
-  const daemonPort = process.env.E2E_DAEMON_PORT;
-  const metroPort = process.env.E2E_METRO_PORT;
-  if (!daemonPort) {
-    throw new Error(
-      "E2E_DAEMON_PORT is not set. Refusing to run e2e against the default daemon (e.g. localhost:6767). " +
-        "Ensure Playwright `globalSetup` starts the e2e daemon and exports E2E_DAEMON_PORT.",
-    );
-  }
-  if (daemonPort === "6767") {
-    throw new Error(
-      "E2E_DAEMON_PORT is 6767. Refusing to run e2e against the default local daemon. " +
-        "Fix Playwright globalSetup to start an isolated test daemon and export its port.",
-    );
-  }
-  if (!metroPort) {
-    throw new Error(
-      "E2E_METRO_PORT is not set. Ensure Playwright `globalSetup` starts Metro and exports E2E_METRO_PORT.",
-    );
-  }
-
-  // Hard guardrail: never allow tests to hit the developer's default daemon.
-  // This blocks both HTTP and WS attempts to :6767 (before any navigation).
-  await page.route(/:(6767)\b/, (route) => route.abort());
-  await page.routeWebSocket(/:(6767)\b/, async (ws) => {
-    await ws.close({ code: 1008, reason: "Blocked connection to localhost:6767 during e2e." });
-  });
-
-  const entries: string[] = [];
-  consoleEntries.set(page, entries);
-
-  page.on("console", (message) => {
-    entries.push(`[console:${message.type()}] ${message.text()}`);
-  });
-
-  page.on("pageerror", (error) => {
-    entries.push(`[pageerror] ${error.message}`);
-  });
-
-  const nowIso = new Date().toISOString();
-  const seedNonce = Math.random().toString(36).slice(2);
-  const serverId = process.env.E2E_SERVER_ID;
-  if (!serverId) {
-    throw new Error("E2E_SERVER_ID is not set - expected from Playwright globalSetup.");
-  }
-  const testDaemon = buildSeededHost({
-    serverId,
-    endpoint: `127.0.0.1:${daemonPort}`,
-    nowIso,
-  });
-  const createAgentPreferences = buildCreateAgentPreferences(testDaemon.serverId);
-
-  await page.addInitScript(
-    ({ daemon, preferences, seedNonce }) => {
-      // `addInitScript` runs on every navigation (including reloads). Some tests intentionally
-      // override storage and reload; they can opt out of seeding for the *next* navigation by
-      // setting this flag before the reload.
-      const disableOnceKey = "@paseo:e2e-disable-default-seed-once";
-      const disableValue = localStorage.getItem(disableOnceKey);
-      if (disableValue) {
-        localStorage.removeItem(disableOnceKey);
-        if (disableValue === seedNonce) {
-          return;
-        }
-      }
-
-      localStorage.setItem("@paseo:e2e", "1");
-      localStorage.setItem("@paseo:e2e-seed-nonce", seedNonce);
-
-      // Hard-reset anything that could point to a developer's real daemon.
-      localStorage.setItem("@paseo:daemon-registry", JSON.stringify([daemon]));
-      localStorage.removeItem("@paseo:settings");
-      localStorage.setItem("@paseo:create-agent-preferences", JSON.stringify(preferences));
-    },
-    { daemon: testDaemon, preferences: createAgentPreferences, seedNonce },
-  );
-});
-
-test.afterEach(async ({ page }, testInfo) => {
-  const entries = consoleEntries.get(page);
-  if (!entries || entries.length === 0) {
-    return;
-  }
-
-  if (testInfo.status === testInfo.expectedStatus) {
-    return;
-  }
-
-  await testInfo.attach("browser-console", {
-    body: entries.join("\n"),
-    contentType: "text/plain",
-  });
-});
-
 export { test, expect, type Page };

packages/app/e2e/global-setup.ts

@@ -1,19 +1,20 @@
-import { spawn, type ChildProcess, execSync } from "node:child_process";
+import { spawn, type ChildProcess, execFileSync, execSync } from "node:child_process";
 import { existsSync } from "node:fs";
-import { mkdtemp, rm } from "node:fs/promises";
+import { chmod, mkdtemp, rm, writeFile } from "node:fs/promises";
 import { tmpdir } from "node:os";
 import path from "node:path";
 import net from "node:net";
 import { Buffer } from "node:buffer";
 import dotenv from "dotenv";
+import { forkPaseoHomeMetadata, resolvePaseoHomePath } from "./helpers/paseo-home-fork";
 
-type WaitForServerOptions = {
+interface WaitForServerOptions {
   host?: string;
   timeoutMs?: number;
   label: string;
   childProcess?: ChildProcess | null;
   getRecentOutput?: () => string;
-};
+}
 
 async function getAvailablePort(): Promise<number> {
   return new Promise((resolve, reject) => {
@@ -124,16 +125,21 @@ async function stopProcess(child: ChildProcess | null): Promise<void> {
   }
   child.kill("SIGTERM");
   await new Promise<void>((resolve) => {
+    let pendingResolve: (() => void) | null = resolve;
+    const settle = () => {
+      if (!pendingResolve) return;
+      const fn = pendingResolve;
+      pendingResolve = null;
+      clearTimeout(timeout);
+      fn();
+    };
     const timeout = setTimeout(() => {
       if (child.exitCode === null && child.signalCode === null) {
         child.kill("SIGKILL");
       }
-      resolve();
+      settle();
     }, 5000);
-    child.once("exit", () => {
-      clearTimeout(timeout);
-      resolve();
-    });
+    child.once("exit", settle);
   });
 }
 
@@ -182,17 +188,109 @@ async function isOpenAiApiKeyUsable(apiKey: string | undefined): Promise<boolean
 let daemonProcess: ChildProcess | null = null;
 let metroProcess: ChildProcess | null = null;
 let paseoHome: string | null = null;
+let fakeGhBinDir: string | null = null;
 let relayProcess: ChildProcess | null = null;
 
-type OfferPayload = {
+function resolveOptionalPaseoHomeEnv(value: string | undefined): string | null {
+  const trimmed = value?.trim();
+  if (!trimmed) {
+    return null;
+  }
+  if (trimmed === "current") {
+    return resolvePaseoHomePath("~/.paseo");
+  }
+  return resolvePaseoHomePath(trimmed);
+}
+
+interface OfferPayload {
   v: 2;
   serverId: string;
   daemonPublicKeyB64: string;
   relay: { endpoint: string };
-};
+}
+
+async function createFakeGhBin(): Promise<string> {
+  const binDir = await mkdtemp(path.join(tmpdir(), "paseo-e2e-gh-bin-"));
+  const ghPath = path.join(binDir, "gh");
+  await writeFile(
+    ghPath,
+    `#!/usr/bin/env node
+const { spawnSync } = require("child_process");
+const fs = require("fs");
+const path = require("path");
+const args = process.argv.slice(2);
+
+function findRealGh() {
+  const fakeBinDir = __dirname;
+  for (const dir of (process.env.PATH || "").split(path.delimiter)) {
+    if (dir === fakeBinDir) continue;
+    const candidate = path.join(dir, "gh");
+    try { fs.accessSync(candidate, fs.constants.X_OK); return candidate; } catch {}
+  }
+  return null;
+}
+
+function forwardToRealGh() {
+  const realGh = findRealGh();
+  if (!realGh) { console.error("[fake-gh] real gh not found in PATH"); process.exit(1); }
+  const result = spawnSync(realGh, process.argv.slice(2), { stdio: "inherit", env: process.env });
+  process.exit(result.status ?? 1);
+}
+
+if (args[0] === "auth" && args[1] === "status") {
+  process.exit(0);
+}
+
+if (args[0] === "pr" && args[1] === "list") {
+  console.log(JSON.stringify([
+    {
+      number: 515,
+      title: "Review selected start ref",
+      url: "https://github.com/getpaseo/paseo/pull/515",
+      state: "OPEN",
+      body: "Fixture pull request for app e2e.",
+      labels: [],
+      baseRefName: "main",
+      headRefName: "feature/start-from-pr"
+    }
+  ]));
+  process.exit(0);
+}
+
+if (args[0] === "pr" && args[1] === "view" && args[2] === "--json" && args[3]) {
+  const fixture = path.join(process.cwd(), ".paseo-e2e-pr.json");
+  if (fs.existsSync(fixture)) {
+    console.log(fs.readFileSync(fixture, "utf8"));
+    process.exit(0);
+  }
+  forwardToRealGh();
+}
+
+if (args[0] === "api" && args[1] === "graphql") {
+  const fixture = path.join(process.cwd(), ".paseo-e2e-timeline.json");
+  if (fs.existsSync(fixture)) {
+    console.log(fs.readFileSync(fixture, "utf8"));
+    process.exit(0);
+  }
+  forwardToRealGh();
+}
+
+if (args[0] === "issue" && args[1] === "list") {
+  console.log("[]");
+  process.exit(0);
+}
+
+forwardToRealGh();
+`,
+  );
+  await chmod(ghPath, 0o755);
+  return binDir;
+}
+
+const ANSI_PATTERN = new RegExp(`${String.fromCharCode(0x1b)}\\[[0-9;]*m`, "g");
 
 function stripAnsi(input: string): string {
-  return input.replace(/\u001b\[[0-9;]*m/g, "");
+  return input.replace(ANSI_PATTERN, "");
 }
 
 function ensureRelayBuildArtifact(repoRoot: string): void {
@@ -224,37 +322,89 @@ function decodeOfferFromFragmentUrl(url: string): OfferPayload {
   return offer as OfferPayload;
 }
 
-export default async function globalSetup() {
-  const repoRoot = path.resolve(__dirname, "../../..");
-  ensureRelayBuildArtifact(repoRoot);
+function loadPairingOfferFromCli(repoRoot: string, paseoHomePath: string): OfferPayload {
+  const stdout = execFileSync(
+    process.execPath,
+    ["--import", "tsx", "packages/cli/src/index.ts", "daemon", "pair", "--json"],
+    {
+      cwd: repoRoot,
+      env: {
+        ...process.env,
+        PASEO_HOME: paseoHomePath,
+      },
+      encoding: "utf8",
+    },
+  );
+  const payload = JSON.parse(stdout) as { relayEnabled?: boolean; url?: string | null };
+  if (payload.relayEnabled !== true || typeof payload.url !== "string") {
+    throw new Error(`Unexpected daemon pair response: ${stdout}`);
+  }
+  return decodeOfferFromFragmentUrl(payload.url);
+}
+
+async function waitForPairingOfferFromCli(args: {
+  repoRoot: string;
+  paseoHome: string;
+  timeoutMs?: number;
+}): Promise<OfferPayload> {
+  const timeoutMs = args.timeoutMs ?? 15000;
+  const start = Date.now();
+  let lastError: unknown = null;
+
+  while (Date.now() - start < timeoutMs) {
+    try {
+      return loadPairingOfferFromCli(args.repoRoot, args.paseoHome);
+    } catch (error) {
+      lastError = error;
+      await sleep(100);
+    }
+  }
+
+  throw new Error(
+    `Timed out waiting for \`paseo daemon pair --json\` to produce a pairing offer: ${
+      lastError instanceof Error ? lastError.message : String(lastError)
+    }`,
+  );
+}
+
+interface DictationConfig {
+  openAiUsable: boolean;
+  localModelsDir: string | null;
+}
+
+async function loadEnvTestFile(repoRoot: string): Promise<void> {
   const envTestPath = path.join(repoRoot, ".env.test");
   if (existsSync(envTestPath)) {
     dotenv.config({ path: envTestPath });
   }
+}
 
-  const port = await getAvailablePort();
-  let relayPort = 0;
-  const metroPort = await getAvailablePort();
-  paseoHome = await mkdtemp(path.join(tmpdir(), "paseo-e2e-home-"));
-  let relayLineBuffer = createLineBuffer();
-  const metroLineBuffer = createLineBuffer();
-  const daemonLineBuffer = createLineBuffer();
-
-  const cleanup = async () => {
-    await Promise.all([
-      stopProcess(daemonProcess),
-      stopProcess(metroProcess),
-      stopProcess(relayProcess),
-    ]);
-    daemonProcess = null;
-    metroProcess = null;
-    relayProcess = null;
-    if (paseoHome) {
-      await rm(paseoHome, { recursive: true, force: true });
-      paseoHome = null;
-    }
-  };
+async function applyPaseoHomeFork(targetHome: string): Promise<void> {
+  const forkSourceHome = resolveOptionalPaseoHomeEnv(process.env.E2E_FORK_PASEO_HOME_FROM);
+  if (!forkSourceHome) {
+    return;
+  }
+  const forkResult = await forkPaseoHomeMetadata({
+    sourceHome: forkSourceHome,
+    targetHome,
+  });
+  process.env.E2E_FORK_SOURCE_PASEO_HOME = forkResult.sourceHome;
+  process.env.E2E_FORK_TARGET_PASEO_HOME = forkResult.targetHome;
+  process.env.E2E_FORK_COPIED_FILES = String(forkResult.copiedFiles);
+  process.env.E2E_FORK_COPIED_BYTES = String(forkResult.copiedBytes);
+  console.log(
+    `[e2e] Forked Paseo metadata from ${forkResult.sourceHome} to ${forkResult.targetHome} ` +
+      `(${forkResult.agentFiles} agent files, ${forkResult.projectFiles} project registry files, ` +
+      `${forkResult.copiedBytes} bytes)`,
+  );
+  if (forkResult.skippedMissing.length > 0) {
+    console.warn(
+      `[e2e] Paseo metadata fork skipped missing paths: ${forkResult.skippedMissing.join(", ")}`,
+    );
+  }
+}
 
+async function resolveDictationConfig(): Promise<DictationConfig> {
   const openAiUsable = await isOpenAiApiKeyUsable(process.env.OPENAI_API_KEY);
   const defaultLocalModelsDir = path.join(
     process.env.HOME ?? "",
@@ -277,238 +427,292 @@ export default async function globalSetup() {
   console.log(
     `[e2e] Dictation STT provider: ${dictationProvider}${openAiUsable ? "" : " (OpenAI probe failed)"}`,
   );
+  return { openAiUsable, localModelsDir };
+}
+
+interface RelayStreamState {
+  failureLine: string | null;
+  readyForSelectedPort: boolean;
+}
+
+function attachRelayStreamHandlers(
+  child: ChildProcess,
+  relayPort: number,
+  buffer: ReturnType<typeof createLineBuffer>,
+  state: RelayStreamState,
+): void {
+  function handleChunk(data: Buffer, streamTag: "stdout" | "stderr") {
+    const lines = data
+      .toString()
+      .split("\n")
+      .filter((line) => line.trim());
+    for (const line of lines) {
+      buffer.add(`[${streamTag}] ${line}`);
+      const failure = parseRelayStartupFailure(line);
+      if (failure) {
+        state.failureLine = failure;
+      }
+      const clean = stripAnsi(line);
+      const readyMatch = clean.match(/Ready on .*:(\d+)\b/i);
+      if (readyMatch && Number(readyMatch[1]) === relayPort) {
+        state.readyForSelectedPort = true;
+      }
+      if (streamTag === "stdout") {
+        console.log(`[relay] ${line}`);
+      } else {
+        console.error(`[relay] ${line}`);
+      }
+    }
+  }
+
+  child.stdout?.on("data", (data: Buffer) => handleChunk(data, "stdout"));
+  child.stderr?.on("data", (data: Buffer) => handleChunk(data, "stderr"));
+}
+
+async function awaitRelayReady(
+  child: ChildProcess,
+  relayPort: number,
+  state: RelayStreamState,
+  buffer: ReturnType<typeof createLineBuffer>,
+): Promise<void> {
+  await waitForServer(relayPort, {
+    label: "Relay dev server",
+    timeoutMs: 30000,
+    childProcess: child,
+    getRecentOutput: buffer.dump,
+  });
+
+  const readyDeadline = Date.now() + 5000;
+  function isRelayReadyCheckPending(): boolean {
+    if (state.readyForSelectedPort) return false;
+    if (state.failureLine !== null) return false;
+    if (child.exitCode !== null) return false;
+    if (child.signalCode !== null) return false;
+    if (Date.now() >= readyDeadline) return false;
+    return true;
+  }
+  while (isRelayReadyCheckPending()) await sleep(100);
+
+  if (state.failureLine) {
+    throw new Error(`Relay startup failed: ${state.failureLine}`);
+  }
+  if (!state.readyForSelectedPort) {
+    throw new Error(
+      `Relay process did not report ready for selected port ${relayPort}.${formatRecentOutput(
+        buffer.dump,
+      )}`,
+    );
+  }
+  if (child.exitCode !== null || child.signalCode !== null) {
+    throw new Error(
+      `Relay process exited before startup completed (exit code ${child.exitCode}, signal ${child.signalCode}).${formatRecentOutput(
+        buffer.dump,
+      )}`,
+    );
+  }
+}
+
+async function startRelay(): Promise<number> {
+  const relayDir = path.resolve(__dirname, "..", "..", "relay");
+  const maxRelayStartupAttempts = 5;
+  let lastRelayStartupError: unknown = null;
+
+  for (let attempt = 1; attempt <= maxRelayStartupAttempts; attempt += 1) {
+    const relayPort = await getAvailablePort();
+    const buffer = createLineBuffer();
+    const state: RelayStreamState = { failureLine: null, readyForSelectedPort: false };
+
+    relayProcess = spawn(
+      "npx",
+      ["wrangler", "dev", "--local", "--ip", "127.0.0.1", "--port", String(relayPort)],
+      {
+        cwd: relayDir,
+        env: { ...process.env },
+        stdio: ["ignore", "pipe", "pipe"],
+        detached: false,
+      },
+    );
+    attachRelayStreamHandlers(relayProcess, relayPort, buffer, state);
+
+    try {
+      await awaitRelayReady(relayProcess, relayPort, state, buffer);
+      return relayPort;
+    } catch (error) {
+      lastRelayStartupError = error;
+      await stopProcess(relayProcess);
+      relayProcess = null;
+    }
+  }
+
+  const message =
+    lastRelayStartupError instanceof Error
+      ? lastRelayStartupError.message
+      : String(lastRelayStartupError);
+  throw new Error(
+    `Failed to start relay dev server after ${maxRelayStartupAttempts} attempts. ${message}`,
+  );
+}
+
+function startMetro(metroPort: number, buffer: ReturnType<typeof createLineBuffer>): ChildProcess {
+  const appDir = path.resolve(__dirname, "..");
+  const child = spawn("npx", ["expo", "start", "--web", "--port", String(metroPort)], {
+    cwd: appDir,
+    env: {
+      ...process.env,
+      BROWSER: "none",
+    },
+    stdio: ["ignore", "pipe", "pipe"],
+    detached: false,
+  });
+
+  child.stdout?.on("data", (data: Buffer) => {
+    const lines = data
+      .toString()
+      .split("\n")
+      .filter((line) => line.trim());
+    for (const line of lines) {
+      buffer.add(`[stdout] ${line}`);
+      console.log(`[metro] ${line}`);
+    }
+  });
+
+  child.stderr?.on("data", (data: Buffer) => {
+    const lines = data
+      .toString()
+      .split("\n")
+      .filter((line) => line.trim());
+    for (const line of lines) {
+      buffer.add(`[stderr] ${line}`);
+      console.error(`[metro] ${line}`);
+    }
+  });
+
+  return child;
+}
+
+interface DaemonSpawnArgs {
+  port: number;
+  relayPort: number;
+  metroPort: number;
+  paseoHome: string;
+  fakeGhBinDir: string;
+  dictation: DictationConfig;
+  buffer: ReturnType<typeof createLineBuffer>;
+}
+
+function startDaemon(args: DaemonSpawnArgs): ChildProcess {
+  const serverDir = path.resolve(__dirname, "../../..", "packages/server");
+  const tsxBin = execSync("which tsx").toString().trim();
+  const { openAiUsable, localModelsDir } = args.dictation;
+
+  const child = spawn(tsxBin, ["scripts/supervisor-entrypoint.ts", "--dev"], {
+    cwd: serverDir,
+    env: {
+      ...process.env,
+      PATH: `${args.fakeGhBinDir}${path.delimiter}${process.env.PATH ?? ""}`,
+      PASEO_HOME: args.paseoHome,
+      PASEO_SERVER_ID: "srv_e2e_test_daemon",
+      PASEO_LISTEN: `0.0.0.0:${args.port}`,
+      PASEO_RELAY_ENDPOINT: `127.0.0.1:${args.relayPort}`,
+      PASEO_CORS_ORIGINS: `http://localhost:${args.metroPort}`,
+      PASEO_DICTATION_ENABLED: openAiUsable ? "1" : "0",
+      PASEO_VOICE_MODE_ENABLED: openAiUsable ? "1" : "0",
+      PASEO_NODE_ENV: "development",
+      ...(openAiUsable
+        ? {
+            PASEO_DICTATION_STT_PROVIDER: "openai",
+            PASEO_VOICE_STT_PROVIDER: "openai",
+            PASEO_VOICE_TTS_PROVIDER: "openai",
+          }
+        : {}),
+      ...(localModelsDir ? { PASEO_LOCAL_MODELS_DIR: localModelsDir } : {}),
+      NODE_ENV: "development",
+    },
+    stdio: ["ignore", "pipe", "pipe"],
+    detached: false,
+  });
+
+  let stdoutBuffer = "";
+  child.stdout?.on("data", (data: Buffer) => {
+    stdoutBuffer += data.toString("utf8");
+    const lines = stdoutBuffer.split("\n");
+    stdoutBuffer = lines.pop() ?? "";
+    for (const line of lines) {
+      const trimmed = line.trim();
+      if (!trimmed) continue;
+      args.buffer.add(`[stdout] ${trimmed}`);
+      console.log(`[daemon] ${trimmed}`);
+    }
+  });
+
+  child.stderr?.on("data", (data: Buffer) => {
+    const lines = data
+      .toString()
+      .split("\n")
+      .filter((line) => line.trim());
+    for (const line of lines) {
+      args.buffer.add(`[stderr] ${line}`);
+      console.error(`[daemon] ${line}`);
+    }
+  });
+
+  return child;
+}
+
+async function performCleanup(shouldRemovePaseoHome: boolean): Promise<void> {
+  await Promise.all([
+    stopProcess(daemonProcess),
+    stopProcess(metroProcess),
+    stopProcess(relayProcess),
+  ]);
+  daemonProcess = null;
+  metroProcess = null;
+  relayProcess = null;
+  if (paseoHome && shouldRemovePaseoHome) {
+    await rm(paseoHome, { recursive: true, force: true });
+    paseoHome = null;
+  } else if (paseoHome) {
+    console.log(`[e2e] Preserving PASEO_HOME: ${paseoHome}`);
+  }
+  if (fakeGhBinDir) {
+    await rm(fakeGhBinDir, { recursive: true, force: true });
+    fakeGhBinDir = null;
+  }
+}
+
+export default async function globalSetup() {
+  const repoRoot = path.resolve(__dirname, "../../..");
+  ensureRelayBuildArtifact(repoRoot);
+  await loadEnvTestFile(repoRoot);
+
+  const port = await getAvailablePort();
+  const metroPort = await getAvailablePort();
+  const requestedPaseoHome = resolveOptionalPaseoHomeEnv(process.env.E2E_PASEO_HOME);
+  const shouldRemovePaseoHome = !requestedPaseoHome && process.env.E2E_KEEP_PASEO_HOME !== "1";
+  paseoHome = requestedPaseoHome ?? (await mkdtemp(path.join(tmpdir(), "paseo-e2e-home-")));
+  fakeGhBinDir = await createFakeGhBin();
+  const metroLineBuffer = createLineBuffer();
+  const daemonLineBuffer = createLineBuffer();
+
+  await applyPaseoHomeFork(paseoHome);
+
+  const cleanup = () => performCleanup(shouldRemovePaseoHome);
+
+  const dictation = await resolveDictationConfig();
 
   try {
-    const relayDir = path.resolve(__dirname, "..", "..", "relay");
-    const maxRelayStartupAttempts = 5;
-    let relayStarted = false;
-    let lastRelayStartupError: unknown = null;
-
-    for (let attempt = 1; attempt <= maxRelayStartupAttempts; attempt += 1) {
-      relayPort = await getAvailablePort();
-      relayLineBuffer = createLineBuffer();
-      let relayStartupFailureLine: string | null = null;
-      let relayReadyForSelectedPort = false;
-
-      relayProcess = spawn(
-        "npx",
-        ["wrangler", "dev", "--local", "--ip", "127.0.0.1", "--port", String(relayPort)],
-        {
-          cwd: relayDir,
-          env: { ...process.env },
-          stdio: ["ignore", "pipe", "pipe"],
-          detached: false,
-        },
-      );
-
-      relayProcess.stdout?.on("data", (data: Buffer) => {
-        const lines = data
-          .toString()
-          .split("\n")
-          .filter((line) => line.trim());
-        for (const line of lines) {
-          relayLineBuffer.add(`[stdout] ${line}`);
-          const failure = parseRelayStartupFailure(line);
-          if (failure) {
-            relayStartupFailureLine = failure;
-          }
-          const clean = stripAnsi(line);
-          const readyMatch = clean.match(/Ready on .*:(\d+)\b/i);
-          if (readyMatch && Number(readyMatch[1]) === relayPort) {
-            relayReadyForSelectedPort = true;
-          }
-          console.log(`[relay] ${line}`);
-        }
-      });
-      relayProcess.stderr?.on("data", (data: Buffer) => {
-        const lines = data
-          .toString()
-          .split("\n")
-          .filter((line) => line.trim());
-        for (const line of lines) {
-          relayLineBuffer.add(`[stderr] ${line}`);
-          const failure = parseRelayStartupFailure(line);
-          if (failure) {
-            relayStartupFailureLine = failure;
-          }
-          const clean = stripAnsi(line);
-          const readyMatch = clean.match(/Ready on .*:(\d+)\b/i);
-          if (readyMatch && Number(readyMatch[1]) === relayPort) {
-            relayReadyForSelectedPort = true;
-          }
-          console.error(`[relay] ${line}`);
-        }
-      });
-
-      try {
-        await waitForServer(relayPort, {
-          label: "Relay dev server",
-          timeoutMs: 30000,
-          childProcess: relayProcess,
-          getRecentOutput: relayLineBuffer.dump,
-        });
-
-        const readyDeadline = Date.now() + 5000;
-        while (
-          !relayReadyForSelectedPort &&
-          relayStartupFailureLine === null &&
-          relayProcess?.exitCode === null &&
-          relayProcess?.signalCode === null &&
-          Date.now() < readyDeadline
-        ) {
-          await sleep(100);
-        }
-
-        if (relayStartupFailureLine) {
-          throw new Error(`Relay startup failed: ${relayStartupFailureLine}`);
-        }
-        if (!relayReadyForSelectedPort) {
-          throw new Error(
-            `Relay process did not report ready for selected port ${relayPort}.${formatRecentOutput(
-              relayLineBuffer.dump,
-            )}`,
-          );
-        }
-        if (relayProcess.exitCode !== null || relayProcess.signalCode !== null) {
-          throw new Error(
-            `Relay process exited before startup completed (exit code ${relayProcess.exitCode}, signal ${relayProcess.signalCode}).${formatRecentOutput(
-              relayLineBuffer.dump,
-            )}`,
-          );
-        }
-
-        relayStarted = true;
-        break;
-      } catch (error) {
-        lastRelayStartupError = error;
-        await stopProcess(relayProcess);
-        relayProcess = null;
-      }
-    }
-
-    if (!relayStarted) {
-      const message =
-        lastRelayStartupError instanceof Error
-          ? lastRelayStartupError.message
-          : String(lastRelayStartupError);
-      throw new Error(
-        `Failed to start relay dev server after ${maxRelayStartupAttempts} attempts. ${message}`,
-      );
-    }
-
-    // Start Metro bundler on dynamic port
-    const appDir = path.resolve(__dirname, "..");
-    metroProcess = spawn("npx", ["expo", "start", "--web", "--port", String(metroPort)], {
-      cwd: appDir,
-      env: {
-        ...process.env,
-        BROWSER: "none", // Don't auto-open browser
-      },
-      stdio: ["ignore", "pipe", "pipe"],
-      detached: false,
+    const relayPort = await startRelay();
+    metroProcess = startMetro(metroPort, metroLineBuffer);
+    daemonProcess = startDaemon({
+      port,
+      relayPort,
+      metroPort,
+      paseoHome,
+      fakeGhBinDir,
+      dictation,
+      buffer: daemonLineBuffer,
     });
 
-    metroProcess.stdout?.on("data", (data: Buffer) => {
-      const lines = data
-        .toString()
-        .split("\n")
-        .filter((line) => line.trim());
-      for (const line of lines) {
-        metroLineBuffer.add(`[stdout] ${line}`);
-        console.log(`[metro] ${line}`);
-      }
-    });
-
-    metroProcess.stderr?.on("data", (data: Buffer) => {
-      const lines = data
-        .toString()
-        .split("\n")
-        .filter((line) => line.trim());
-      for (const line of lines) {
-        metroLineBuffer.add(`[stderr] ${line}`);
-        console.error(`[metro] ${line}`);
-      }
-    });
-
-    const serverDir = path.resolve(__dirname, "../../..", "packages/server");
-    const tsxBin = execSync("which tsx").toString().trim();
-
-    let offerPayload: OfferPayload | null = null;
-    let offerResolve: (() => void) | null = null;
-    const offerPromise = new Promise<void>((resolve) => {
-      offerResolve = resolve;
-    });
-
-    daemonProcess = spawn(tsxBin, ["src/server/index.ts"], {
-      cwd: serverDir,
-      env: {
-        ...process.env,
-        PASEO_HOME: paseoHome,
-        PASEO_SERVER_ID: "srv_e2e_test_daemon",
-        PASEO_LISTEN: `0.0.0.0:${port}`,
-        PASEO_RELAY_ENDPOINT: `127.0.0.1:${relayPort}`,
-        PASEO_CORS_ORIGINS: `http://localhost:${metroPort}`,
-        PASEO_DICTATION_ENABLED: openAiUsable ? "1" : "0",
-        PASEO_VOICE_MODE_ENABLED: openAiUsable ? "1" : "0",
-        ...(openAiUsable
-          ? {
-              PASEO_DICTATION_STT_PROVIDER: "openai",
-              PASEO_VOICE_STT_PROVIDER: "openai",
-              PASEO_VOICE_TTS_PROVIDER: "openai",
-            }
-          : {}),
-        ...(localModelsDir ? { PASEO_LOCAL_MODELS_DIR: localModelsDir } : {}),
-        NODE_ENV: "development",
-      },
-      stdio: ["ignore", "pipe", "pipe"],
-      detached: false,
-    });
-
-    let stdoutBuffer = "";
-    daemonProcess.stdout?.on("data", (data: Buffer) => {
-      stdoutBuffer += data.toString("utf8");
-      const lines = stdoutBuffer.split("\n");
-      stdoutBuffer = lines.pop() ?? "";
-      for (const line of lines) {
-        const trimmed = line.trim();
-        if (!trimmed) continue;
-        daemonLineBuffer.add(`[stdout] ${trimmed}`);
-        if (!offerPayload) {
-          const clean = stripAnsi(trimmed);
-          try {
-            const obj = JSON.parse(clean) as { msg?: string; url?: string };
-            if (obj.msg === "pairing_offer" && typeof obj.url === "string") {
-              offerPayload = decodeOfferFromFragmentUrl(obj.url);
-              offerResolve?.();
-            }
-          } catch {
-            const match = clean.match(/https?:\/\/[^\s"]+#offer=[A-Za-z0-9_-]+/);
-            if (match && clean.includes("pairing_offer")) {
-              try {
-                offerPayload = decodeOfferFromFragmentUrl(match[0]);
-                offerResolve?.();
-              } catch {
-                // ignore parsing failures
-              }
-            }
-          }
-        }
-        console.log(`[daemon] ${trimmed}`);
-      }
-    });
-
-    daemonProcess.stderr?.on("data", (data: Buffer) => {
-      const lines = data
-        .toString()
-        .split("\n")
-        .filter((line) => line.trim());
-      for (const line of lines) {
-        daemonLineBuffer.add(`[stderr] ${line}`);
-        console.error(`[daemon] ${line}`);
-      }
-    });
-
-    // Wait for both daemon and Metro to be ready
     await Promise.all([
       waitForServer(port, {
         label: "Paseo daemon",
@@ -517,29 +721,23 @@ export default async function globalSetup() {
       }),
       waitForServer(metroPort, {
         label: "Metro web server",
-        timeoutMs: 120000, // Metro can take longer to start
+        timeoutMs: 120000,
         childProcess: metroProcess,
         getRecentOutput: metroLineBuffer.dump,
       }),
     ]);
 
-    // Wait for daemon to emit a pairing offer (includes relay session ID).
-    await Promise.race([
-      offerPromise,
-      new Promise((_, reject) =>
-        setTimeout(() => reject(new Error("Timed out waiting for pairing_offer log")), 15000),
-      ),
-    ]);
-    if (!offerPayload) {
-      throw new Error("pairing_offer was not parsed from daemon logs");
-    }
-    const offer = offerPayload as OfferPayload;
+    const offer = await waitForPairingOfferFromCli({
+      repoRoot,
+      paseoHome,
+    });
 
     process.env.E2E_DAEMON_PORT = String(port);
     process.env.E2E_RELAY_PORT = String(relayPort);
     process.env.E2E_SERVER_ID = offer.serverId;
     process.env.E2E_RELAY_DAEMON_PUBLIC_KEY = offer.daemonPublicKeyB64;
     process.env.E2E_METRO_PORT = String(metroPort);
+    process.env.E2E_PASEO_HOME = paseoHome;
     console.log(
       `[e2e] Test daemon started on port ${port}, Metro on port ${metroPort}, home: ${paseoHome}`,
     );

packages/app/e2e/helpers/agent-bottom-anchor.ts

@@ -2,26 +2,27 @@ import { expect, type Page } from "@playwright/test";
 import path from "node:path";
 import { pathToFileURL } from "node:url";
 import { randomUUID } from "node:crypto";
+import { createNodeWebSocketFactory, type NodeWebSocketFactory } from "./node-ws-factory";
 import { buildHostWorkspaceRoute } from "../../src/utils/host-routes";
 
 const NEAR_BOTTOM_THRESHOLD_PX = 72;
 
-export type ScrollMetrics = {
+export interface ScrollMetrics {
   offsetY: number;
   contentHeight: number;
   viewportHeight: number;
   distanceFromBottom: number;
-};
+}
 
-export type SeededAgent = {
+export interface SeededAgent {
   id: string;
   title: string;
   expectedTailText: string;
   url: string;
   workspaceUrl: string;
-};
+}
 
-export type DaemonClientInstance = {
+export interface DaemonClientInstance {
   connect(): Promise<void>;
   close(): Promise<void>;
   createAgent(options: {
@@ -35,7 +36,7 @@ export type DaemonClientInstance = {
   }): Promise<{ id: string }>;
   sendAgentMessage(agentId: string, text: string): Promise<void>;
   waitForFinish(agentId: string, timeout?: number): Promise<{ status: string }>;
-};
+}
 
 function getDaemonWsUrl(): string {
   const daemonPort = process.env.E2E_DAEMON_PORT;
@@ -83,33 +84,34 @@ export function createReplyTurn(label: string): {
   };
 }
 
+interface DaemonClientConfig {
+  url: string;
+  clientId: string;
+  clientType: "cli";
+  webSocketFactory?: NodeWebSocketFactory;
+}
+
 async function loadDaemonClientConstructor(): Promise<
-  new (config: {
-    url: string;
-    clientId: string;
-    clientType: "cli";
-  }) => DaemonClientInstance
+  new (config: DaemonClientConfig) => DaemonClientInstance
 > {
-  const repoRoot = path.resolve(process.cwd(), "../..");
+  const repoRoot = path.resolve(__dirname, "../../../../");
   const moduleUrl = pathToFileURL(
     path.join(repoRoot, "packages/server/dist/server/server/exports.js"),
   ).href;
   const mod = (await import(moduleUrl)) as {
-    DaemonClient: new (config: {
-      url: string;
-      clientId: string;
-      clientType: "cli";
-    }) => DaemonClientInstance;
+    DaemonClient: new (config: DaemonClientConfig) => DaemonClientInstance;
   };
   return mod.DaemonClient;
 }
 
 export async function connectDaemonClient(): Promise<DaemonClientInstance> {
   const DaemonClient = await loadDaemonClientConstructor();
+  const webSocketFactory = createNodeWebSocketFactory();
   const client = new DaemonClient({
     url: getDaemonWsUrl(),
     clientId: `app-e2e-${randomUUID()}`,
     clientType: "cli",
+    webSocketFactory,
   });
   await client.connect();
   return client;
@@ -127,7 +129,7 @@ export async function seedBottomAnchorAgent(input: {
   const lineCount = Math.max(14, input.lineCount ?? 14);
   const created = await input.client.createAgent({
     provider: "codex",
-    model: "gpt-5.1-codex-mini",
+    model: "gpt-5.4-mini",
     thinkingOptionId: "low",
     modeId: "full-access",
     cwd: input.cwd,

packages/app/e2e/helpers/agent-stream.ts

@@ -0,0 +1,35 @@
+import { expect, type Page } from "@playwright/test";
+import { readScrollMetrics, waitForContentGrowth, expectNearBottom } from "./agent-bottom-anchor";
+
+export async function awaitAssistantMessage(page: Page, hasText?: string | RegExp): Promise<void> {
+  const messages = page.getByTestId("assistant-message");
+  const target = hasText === undefined ? messages.first() : messages.filter({ hasText }).first();
+  await expect(target).toBeVisible({ timeout: 30_000 });
+}
+
+export async function awaitToolCall(page: Page, toolName: string | RegExp): Promise<void> {
+  await expect(
+    page.getByTestId("tool-call-badge").filter({ hasText: toolName }).first(),
+  ).toBeVisible({ timeout: 30_000 });
+}
+
+export async function expectAgentIdle(page: Page, timeout = 30_000): Promise<void> {
+  await expect(page.getByRole("button", { name: /stop|cancel/i })).toHaveCount(0, { timeout });
+}
+
+// The working indicator is an animated spinner View — no semantic ARIA role, testId is correct.
+export async function expectInlineWorkingIndicator(page: Page): Promise<void> {
+  await expect(page.getByTestId("turn-working-indicator")).toBeVisible({ timeout: 30_000 });
+}
+
+export async function expectTurnCopyButton(page: Page): Promise<void> {
+  await expect(page.getByRole("button", { name: "Copy turn" }).first()).toBeVisible({
+    timeout: 30_000,
+  });
+}
+
+export async function expectScrollFollowsNewContent(page: Page): Promise<void> {
+  const { contentHeight } = await readScrollMetrics(page);
+  await waitForContentGrowth(page, contentHeight);
+  await expectNearBottom(page);
+}

packages/app/e2e/helpers/app.ts

@@ -1,215 +1,44 @@
 import { expect, type Page } from "@playwright/test";
-import { buildCreateAgentPreferences, buildSeededHost } from "./daemon-registry";
 
 function escapeRegex(value: string): string {
   return value.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
 }
 
-function getE2EDaemonPort(): string {
-  const port = process.env.E2E_DAEMON_PORT;
-  if (!port) {
-    throw new Error("E2E_DAEMON_PORT is not set (expected from Playwright globalSetup).");
-  }
-  if (port === "6767") {
-    throw new Error(
-      "E2E_DAEMON_PORT is 6767. Refusing to run e2e against the default local daemon.",
-    );
-  }
-  return port;
-}
-
-async function ensureE2EStorageSeeded(page: Page): Promise<void> {
-  const port = getE2EDaemonPort();
-  const expectedEndpoint = `127.0.0.1:${port}`;
-  const expectedServerId = process.env.E2E_SERVER_ID;
-  if (!expectedServerId) {
-    throw new Error("E2E_SERVER_ID is not set (expected from Playwright globalSetup).");
-  }
-
-  const needsReset = await page.evaluate(
-    ({ expectedEndpoint, expectedServerId }) => {
-      const raw = localStorage.getItem("@paseo:daemon-registry");
-      if (!raw) return true;
-      try {
-        const parsed = JSON.parse(raw);
-        if (!Array.isArray(parsed) || parsed.length !== 1) return true;
-        const entry = parsed[0] as any;
-        if (entry?.serverId !== expectedServerId) return true;
-        const connections = entry?.connections;
-        if (!Array.isArray(connections)) return true;
-        if (
-          connections.some(
-            (c: any) =>
-              c?.type === "directTcp" &&
-              typeof c?.endpoint === "string" &&
-              /:6767\b/.test(c.endpoint),
-          )
-        )
-          return true;
-        return !connections.some(
-          (c: any) => c?.type === "directTcp" && c?.endpoint === expectedEndpoint,
-        );
-      } catch {
-        return true;
-      }
-    },
-    { expectedEndpoint, expectedServerId },
-  );
-
-  if (!needsReset) {
-    return;
-  }
-
-  const nowIso = new Date().toISOString();
-  const daemon = buildSeededHost({
-    serverId: expectedServerId,
-    endpoint: expectedEndpoint,
-    nowIso,
-  });
-  const preferences = buildCreateAgentPreferences(expectedServerId);
-  await page.evaluate(
-    ({ daemon, preferences }) => {
-      localStorage.setItem("@paseo:e2e", "1");
-      localStorage.setItem("@paseo:daemon-registry", JSON.stringify([daemon]));
-      localStorage.setItem("@paseo:create-agent-preferences", JSON.stringify(preferences));
-      localStorage.removeItem("@paseo:settings");
-    },
-    { daemon, preferences },
-  );
-
-  await page.reload();
-}
-
-async function assertE2EUsesSeededTestDaemon(page: Page): Promise<void> {
-  const port = getE2EDaemonPort();
-  const expectedEndpoint = `127.0.0.1:${port}`;
-  const expectedServerId = process.env.E2E_SERVER_ID;
-  if (!expectedServerId) {
-    throw new Error("E2E_SERVER_ID is not set (expected from Playwright globalSetup).");
-  }
-
-  const snapshot = await page.evaluate(() => {
-    const registryRaw = localStorage.getItem("@paseo:daemon-registry");
-    const prefsRaw = localStorage.getItem("@paseo:create-agent-preferences");
-    return { registryRaw, prefsRaw };
-  });
-
-  if (!snapshot.registryRaw) {
-    throw new Error("E2E expected @paseo:daemon-registry to be set before app load.");
-  }
-
-  let registry: any;
-  try {
-    registry = JSON.parse(snapshot.registryRaw);
-  } catch {
-    throw new Error("E2E expected @paseo:daemon-registry to be valid JSON.");
-  }
-
-  if (!Array.isArray(registry) || registry.length !== 1) {
-    throw new Error(
-      `E2E expected @paseo:daemon-registry to contain exactly 1 daemon (got ${Array.isArray(registry) ? registry.length : "non-array"}).`,
-    );
-  }
-
-  const daemon = registry[0];
-  if (typeof daemon?.serverId !== "string" || daemon.serverId.length === 0) {
-    throw new Error(
-      `E2E expected seeded daemon to have a string serverId (got ${String(daemon?.serverId)}).`,
-    );
-  }
-  if (daemon.serverId !== expectedServerId) {
-    throw new Error(
-      `E2E expected seeded daemon serverId to be ${expectedServerId} (got ${daemon.serverId}).`,
-    );
-  }
-
-  const connections: unknown = daemon?.connections;
-  if (
-    !Array.isArray(connections) ||
-    !connections.some((c: any) => c?.type === "directTcp" && c?.endpoint === expectedEndpoint)
-  ) {
-    throw new Error(
-      `E2E expected seeded daemon connections to include directTcp ${expectedEndpoint} (got ${JSON.stringify(connections)}).`,
-    );
-  }
-  if (
-    Array.isArray(connections) &&
-    connections.some(
-      (c: any) =>
-        c?.type === "directTcp" && typeof c?.endpoint === "string" && /:6767\b/.test(c.endpoint),
-    )
-  ) {
-    throw new Error(
-      `E2E detected a daemon endpoint pointing at :6767 (${JSON.stringify(connections)}).`,
-    );
-  }
-
-  if (!snapshot.prefsRaw) {
-    throw new Error("E2E expected @paseo:create-agent-preferences to be set before app load.");
-  }
-  try {
-    const prefs = JSON.parse(snapshot.prefsRaw) as any;
-    if (prefs?.serverId !== daemon.serverId) {
-      throw new Error(
-        `E2E expected create-agent-preferences.serverId to match seeded daemon serverId (${daemon.serverId}) (got ${String(prefs?.serverId)}).`,
-      );
-    }
-  } catch (error) {
-    if (error instanceof Error) throw error;
-    throw new Error("E2E expected @paseo:create-agent-preferences to be valid JSON.");
-  }
-}
-
 export const gotoAppShell = async (page: Page) => {
   await page.goto("/");
-  await ensureE2EStorageSeeded(page);
 };
 
 export const gotoHome = async (page: Page) => {
   await gotoAppShell(page);
-  const composer = page.getByRole("textbox", { name: "Message agent..." });
-  if (
-    !(await composer
-      .first()
-      .isVisible()
-      .catch(() => false))
-  ) {
-    const addProjectCta = page.getByText("Add a project", { exact: true }).first();
-    const addProjectSidebar = page.getByText("Add project", { exact: true }).first();
-    const newAgentButton = page.getByText("New agent", { exact: true }).first();
+  const composer = page.getByRole("textbox", { name: "Message agent..." }).first();
+  const entryButton = page
+    .getByText("Add a project", { exact: true })
+    .or(page.getByText("Add project", { exact: true }))
+    .or(page.getByText("New agent", { exact: true }))
+    .first();
 
-    await expect
-      .poll(
-        async () =>
-          (await addProjectCta.isVisible().catch(() => false)) ||
-          (await addProjectSidebar.isVisible().catch(() => false)) ||
-          (await newAgentButton.isVisible().catch(() => false)),
-        { timeout: 10000 },
-      )
-      .toBe(true);
+  await expect
+    .poll(
+      async () =>
+        (await composer.isVisible().catch(() => false)) ||
+        (await entryButton.isVisible().catch(() => false)),
+      { timeout: 10_000 },
+    )
+    .toBe(true);
 
-    if (await addProjectCta.isVisible().catch(() => false)) {
-      await addProjectCta.click();
-    } else if (await addProjectSidebar.isVisible().catch(() => false)) {
-      await addProjectSidebar.click();
-    } else {
-      await newAgentButton.click();
-    }
+  if (!(await composer.isVisible().catch(() => false))) {
+    await entryButton.click();
   }
-  await expect(composer.first()).toBeVisible({ timeout: 30000 });
+
+  await expect(composer).toBeVisible({ timeout: 30_000 });
 };
 
 export const openSettings = async (page: Page) => {
-  const serverId = process.env.E2E_SERVER_ID;
-  if (!serverId) {
-    throw new Error("E2E_SERVER_ID is not set (expected from Playwright globalSetup).");
-  }
-
   // Navigate through the real app control so route changes stay aligned with UI behavior.
   const settingsButton = page.locator('[data-testid="sidebar-settings"]:visible').first();
   await expect(settingsButton).toBeVisible();
   await settingsButton.click();
-  await expect(page).toHaveURL(new RegExp(`/h/${escapeRegex(serverId)}/settings$`));
+  await expect(page).toHaveURL(/\/settings\/general$/);
 };
 
 export const setWorkingDirectory = async (page: Page, directory: string) => {
@@ -313,7 +142,7 @@ export const setWorkingDirectory = async (page: Page, directory: string) => {
   if (trimmedDirectory.startsWith("/private/var/")) {
     directoryCandidates.add(trimmedDirectory.replace(/^\/private/, ""));
   }
-  const basename = trimmedDirectory.split("/").filter(Boolean).pop() ?? trimmedDirectory;
+  const basename = trimmedDirectory.split("/").findLast(Boolean) ?? trimmedDirectory;
 
   await expect
     .poll(
@@ -331,46 +160,6 @@ export const setWorkingDirectory = async (page: Page, directory: string) => {
 };
 
 export const ensureHostSelected = async (page: Page) => {
-  await ensureE2EStorageSeeded(page);
-
-  // Absolute verification that we're using the per-run e2e daemon (never :6767).
-  // Also self-heal a rare case where app code rewrites daemon IDs after boot, by
-  // realigning create-agent-preferences.serverId to the sole seeded daemon.
-  try {
-    await assertE2EUsesSeededTestDaemon(page);
-  } catch (error) {
-    const message = error instanceof Error ? error.message : String(error);
-    if (!/create-agent-preferences\.serverId/i.test(message)) {
-      throw error;
-    }
-
-    const fix = await page.evaluate(() => {
-      const registryRaw = localStorage.getItem("@paseo:daemon-registry");
-      const prefsRaw = localStorage.getItem("@paseo:create-agent-preferences");
-      if (!registryRaw || !prefsRaw) return { ok: false, reason: "missing storage" } as const;
-      const registry = JSON.parse(registryRaw) as any[];
-      const prefs = JSON.parse(prefsRaw) as any;
-      if (!Array.isArray(registry) || registry.length !== 1)
-        return { ok: false, reason: "registry shape" } as const;
-      const serverId = registry[0]?.serverId;
-      if (typeof serverId !== "string" || serverId.length === 0)
-        return { ok: false, reason: "missing serverId" } as const;
-      prefs.serverId = serverId;
-      localStorage.setItem("@paseo:create-agent-preferences", JSON.stringify(prefs));
-      // Prevent the fixture's init-script from overwriting the corrected prefs on reload.
-      const nonce = localStorage.getItem("@paseo:e2e-seed-nonce") ?? "1";
-      localStorage.setItem("@paseo:e2e-disable-default-seed-once", nonce);
-      return { ok: true } as const;
-    });
-
-    if (!fix.ok) {
-      throw error;
-    }
-
-    await page.reload();
-    await assertE2EUsesSeededTestDaemon(page);
-  }
-
   const input = page.getByRole("textbox", { name: "Message agent..." });
   await expect(input).toBeVisible();
 
@@ -382,7 +171,7 @@ export const ensureHostSelected = async (page: Page) => {
   if (await selectHostLabel.isVisible()) {
     await selectHostLabel.click();
 
-    // E2E safety: we enforce a single seeded daemon, so the option should be unambiguous.
+    // We enforce a single seeded daemon, so the option should be unambiguous.
     const localhostOption = page.getByText("localhost", { exact: true }).first();
     const daemonIdOption = page
       .getByText(process.env.E2E_SERVER_ID ?? "srv_e2e_test_daemon", { exact: true })
@@ -647,57 +436,3 @@ export const createAgentInRepo = async (
   await setWorkingDirectory(page, config.directory);
   await createAgent(page, config.prompt);
 };
-
-export const waitForPermissionPrompt = async (page: Page, timeout = 30000) => {
-  const promptText = page.getByTestId("permission-request-question").first();
-  await expect(promptText).toBeVisible({ timeout });
-};
-
-export const allowPermission = async (page: Page) => {
-  const acceptButton = page.getByTestId("permission-request-accept").first();
-  await expect(acceptButton).toBeVisible({ timeout: 5000 });
-  await acceptButton.click();
-};
-
-export const denyPermission = async (page: Page) => {
-  const denyButton = page.getByTestId("permission-request-deny").first();
-  await expect(denyButton).toBeVisible({ timeout: 5000 });
-  await denyButton.click();
-};
-
-export async function waitForAgentFinishUI(page: Page, timeout = 30000) {
-  // Wait for the stop button to disappear
-  const stopButton = page.getByRole("button", { name: /stop|cancel/i });
-
-  // First, let's debug what's happening - wait a bit to see the state
-  await page.waitForTimeout(2000);
-
-  // Check if stop button is visible
-  const isVisible = await stopButton.isVisible().catch(() => false);
-
-  if (isVisible) {
-    // If stop button is still visible after permission denial,
-    // it might be that the agent is waiting for something.
-    // Let's check if there's a tool call result or other UI indication
-
-    // Look for any indication that the agent has processed the permission denial
-    const toolCallResult = page.getByText(/permission.*denied|denied|blocked/i);
-
-    // Wait for the tool call result to appear
-    await expect(toolCallResult)
-      .toBeVisible({ timeout: 10000 })
-      .catch(() => {
-        // If no specific message, just wait for the button to disappear
-      });
-
-    // Now wait for the stop button to disappear
-    await expect(stopButton).not.toBeVisible({ timeout });
-  }
-}
-
-export async function getToolCallCount(page: Page): Promise<number> {
-  // Tool calls are rendered as ExpandableBadge components with tool names like Bash, Write, Read, etc.
-  // They appear as pressable badges in the agent stream
-  const toolCallBadges = page.locator('[data-testid="tool-call-badge"]');
-  return toolCallBadges.count();
-}

packages/app/e2e/helpers/archive-tab.ts

@@ -0,0 +1,323 @@
+import { randomUUID } from "node:crypto";
+import path from "node:path";
+import { pathToFileURL } from "node:url";
+import { expect, type Page } from "@playwright/test";
+import { buildCreateAgentPreferences, buildSeededHost } from "./daemon-registry";
+import { createNodeWebSocketFactory, type NodeWebSocketFactory } from "./node-ws-factory";
+import { waitForWorkspaceTabsVisible } from "./workspace-tabs";
+import {
+  buildHostAgentDetailRoute,
+  buildHostSessionsRoute,
+  buildHostWorkspaceRoute,
+} from "@/utils/host-routes";
+
+export interface ArchiveTabAgent {
+  id: string;
+  title: string;
+  cwd: string;
+}
+
+interface ArchiveTabDaemonClient {
+  connect(): Promise<void>;
+  close(): Promise<void>;
+  createAgent(options: {
+    provider: string;
+    model: string;
+    thinkingOptionId?: string;
+    modeId: string;
+    cwd: string;
+    title: string;
+    initialPrompt?: string;
+  }): Promise<{ id: string }>;
+  archiveAgent(agentId: string): Promise<{ archivedAt: string }>;
+  waitForFinish(agentId: string, timeout?: number): Promise<{ status: string }>;
+  waitForAgentUpsert(
+    agentId: string,
+    predicate: (snapshot: { status: string }) => boolean,
+    timeout?: number,
+  ): Promise<{ status: string }>;
+  fetchAgentHistory(options?: {
+    page?: { limit: number };
+  }): Promise<{ entries: Array<{ id: string }> }>;
+}
+
+function getDaemonPort(): string {
+  const daemonPort = process.env.E2E_DAEMON_PORT;
+  if (!daemonPort) {
+    throw new Error("E2E_DAEMON_PORT is not set.");
+  }
+  if (daemonPort === "6767") {
+    throw new Error("E2E_DAEMON_PORT must not point at the developer daemon.");
+  }
+  return daemonPort;
+}
+
+function getServerId(): string {
+  const serverId = process.env.E2E_SERVER_ID;
+  if (!serverId) {
+    throw new Error("E2E_SERVER_ID is not set.");
+  }
+  return serverId;
+}
+
+function getDaemonWsUrl(): string {
+  return `ws://127.0.0.1:${getDaemonPort()}/ws`;
+}
+
+function buildSeededStoragePayload() {
+  const nowIso = new Date().toISOString();
+  return {
+    daemon: buildSeededHost({
+      serverId: getServerId(),
+      endpoint: `127.0.0.1:${getDaemonPort()}`,
+      nowIso,
+    }),
+    preferences: buildCreateAgentPreferences(getServerId()),
+  };
+}
+
+interface ArchiveTabDaemonClientConfig {
+  url: string;
+  clientId: string;
+  clientType: "cli";
+  webSocketFactory?: NodeWebSocketFactory;
+}
+
+async function loadDaemonClientConstructor(): Promise<
+  new (config: ArchiveTabDaemonClientConfig) => ArchiveTabDaemonClient
+> {
+  const repoRoot = path.resolve(__dirname, "../../../../");
+  const moduleUrl = pathToFileURL(
+    path.join(repoRoot, "packages/server/dist/server/server/exports.js"),
+  ).href;
+  const mod = (await import(moduleUrl)) as {
+    DaemonClient: new (config: ArchiveTabDaemonClientConfig) => ArchiveTabDaemonClient;
+  };
+  return mod.DaemonClient;
+}
+
+export async function connectArchiveTabDaemonClient(): Promise<ArchiveTabDaemonClient> {
+  const DaemonClient = await loadDaemonClientConstructor();
+  const webSocketFactory = createNodeWebSocketFactory();
+  const client = new DaemonClient({
+    url: getDaemonWsUrl(),
+    clientId: `app-e2e-archive-tab-${randomUUID()}`,
+    clientType: "cli",
+    webSocketFactory,
+  });
+  await client.connect();
+  return client;
+}
+
+export async function createIdleAgent(
+  client: ArchiveTabDaemonClient,
+  input: { cwd: string; title: string },
+): Promise<ArchiveTabAgent> {
+  const created = await client.createAgent({
+    provider: "opencode",
+    model: "opencode/gpt-5-nano",
+    modeId: "bypassPermissions",
+    cwd: input.cwd,
+    title: input.title,
+  });
+  const snapshot = await client.waitForAgentUpsert(
+    created.id,
+    (agent) => agent.status === "idle",
+    30_000,
+  );
+  if (snapshot.status !== "idle") {
+    throw new Error(`Expected agent ${created.id} to become idle, got ${snapshot.status}.`);
+  }
+  return {
+    id: created.id,
+    title: input.title,
+    cwd: input.cwd,
+  };
+}
+
+export async function archiveAgentFromDaemon(
+  client: ArchiveTabDaemonClient,
+  agentId: string,
+): Promise<void> {
+  await client.archiveAgent(agentId);
+}
+
+export async function primeAdditionalPage(page: Page): Promise<void> {
+  const seedNonce = randomUUID();
+  const { daemon, preferences } = buildSeededStoragePayload();
+
+  await page.route(/:(6767)\b/, (route) => route.abort());
+  await page.routeWebSocket(/:(6767)\b/, async (ws) => {
+    await ws.close({ code: 1008, reason: "Blocked connection to localhost:6767 during e2e." });
+  });
+  await page.addInitScript(
+    ({ daemon: seededDaemon, preferences: seededPreferences, seedNonce: nonce }) => {
+      const disableOnceKey = "@paseo:e2e-disable-default-seed-once";
+      const disableValue = localStorage.getItem(disableOnceKey);
+      if (disableValue) {
+        localStorage.removeItem(disableOnceKey);
+        if (disableValue === nonce) {
+          return;
+        }
+      }
+
+      localStorage.setItem("@paseo:e2e", "1");
+      localStorage.setItem("@paseo:e2e-seed-nonce", nonce);
+      localStorage.setItem("@paseo:daemon-registry", JSON.stringify([seededDaemon]));
+      localStorage.removeItem("@paseo:settings");
+      localStorage.setItem("@paseo:create-agent-preferences", JSON.stringify(seededPreferences));
+    },
+    { daemon, preferences, seedNonce },
+  );
+  await page.goto("/");
+}
+
+export async function resetSeededPageState(page: Page): Promise<void> {
+  const { daemon, preferences } = buildSeededStoragePayload();
+  await page.goto("/");
+  await page.evaluate(
+    ({ daemon: seededDaemon, preferences: seededPreferences }) => {
+      localStorage.clear();
+      localStorage.setItem("@paseo:e2e", "1");
+      localStorage.setItem("@paseo:daemon-registry", JSON.stringify([seededDaemon]));
+      localStorage.setItem("@paseo:create-agent-preferences", JSON.stringify(seededPreferences));
+      localStorage.removeItem("@paseo:settings");
+    },
+    { daemon, preferences },
+  );
+  await page.goto("/");
+}
+
+export async function openWorkspaceWithAgents(
+  page: Page,
+  agents: [ArchiveTabAgent, ArchiveTabAgent],
+): Promise<void> {
+  const serverId = getServerId();
+  for (const agent of agents) {
+    await page.goto(buildHostAgentDetailRoute(serverId, agent.id, agent.cwd));
+
+    // The workspace layout consumes `?open=agent:xxx`, returns null during the effect,
+    // then replaces the URL with the clean workspace route after preparing the tab.
+    // On CI, Expo Router's rootNavigationState may take time to initialize,
+    // so we allow a generous timeout here (matching terminal-perf pattern).
+    await page.waitForURL(
+      (url) => url.pathname.includes("/workspace/") && !url.searchParams.has("open"),
+      { timeout: 60_000 },
+    );
+
+    await waitForWorkspaceTabsVisible(page);
+    await expectWorkspaceTabVisible(page, agent.id);
+  }
+}
+
+export async function expectWorkspaceTabVisible(page: Page, agentId: string): Promise<void> {
+  await expect(
+    page.getByTestId(`workspace-tab-agent_${agentId}`).filter({ visible: true }).first(),
+  ).toBeVisible({ timeout: 30_000 });
+}
+
+export async function expectWorkspaceTabHidden(page: Page, agentId: string): Promise<void> {
+  await expect(
+    page.getByTestId(`workspace-tab-agent_${agentId}`).filter({ visible: true }),
+  ).toHaveCount(0, {
+    timeout: 30_000,
+  });
+}
+
+export async function expectWorkspaceArchiveOutcome(
+  page: Page,
+  input: { archivedAgentId: string; survivingAgentId: string },
+): Promise<void> {
+  await expectWorkspaceTabHidden(page, input.archivedAgentId);
+  await expectWorkspaceTabVisible(page, input.survivingAgentId);
+}
+
+export async function closeWorkspaceAgentTab(page: Page, agentId: string): Promise<void> {
+  const closeButton = page.getByTestId(`workspace-agent-close-${agentId}`).filter({
+    visible: true,
+  });
+  await expect(closeButton.first()).toBeVisible({ timeout: 30_000 });
+  await closeButton.first().click();
+  await expectWorkspaceTabHidden(page, agentId);
+}
+
+export async function expectArchivedAgentFocused(page: Page, agentId: string): Promise<void> {
+  await expectWorkspaceTabVisible(page, agentId);
+  await expect(
+    page.getByText("This agent is archived").filter({ visible: true }).first(),
+  ).toBeVisible({
+    timeout: 30_000,
+  });
+}
+
+export async function reloadWorkspace(page: Page, workspaceId: string): Promise<void> {
+  const serverId = getServerId();
+  await page.goto(buildHostWorkspaceRoute(serverId, workspaceId));
+  await waitForWorkspaceTabsVisible(page);
+}
+
+export async function openSessions(page: Page): Promise<void> {
+  const sessionsButton = page.getByTestId("sidebar-sessions");
+  await expect(sessionsButton).toBeVisible({ timeout: 30_000 });
+  await sessionsButton.click();
+  await expect(page).toHaveURL(new RegExp(`${buildHostSessionsRoute(getServerId())}$`), {
+    timeout: 30_000,
+  });
+  await expect(page.getByText("Sessions", { exact: true }).last()).toBeVisible({
+    timeout: 30_000,
+  });
+}
+
+const AGENT_ROW_SELECTOR = '[data-testid^="agent-row-"]';
+
+function getSessionRowByTitle(page: Page, title: string) {
+  return page.locator(AGENT_ROW_SELECTOR).filter({ hasText: title }).first();
+}
+
+export async function expectSessionRowVisible(page: Page, title: string): Promise<void> {
+  await expect(getSessionRowByTitle(page, title)).toBeVisible({ timeout: 30_000 });
+}
+
+export async function expectSessionRowArchived(page: Page, title: string): Promise<void> {
+  await expect(getSessionRowByTitle(page, title)).toContainText("Archived", { timeout: 30_000 });
+}
+
+export async function clickSessionRow(page: Page, title: string): Promise<void> {
+  const row = getSessionRowByTitle(page, title);
+  await expect(row).toBeVisible({ timeout: 30_000 });
+  await row.click();
+}
+
+export async function expectSessionsEmptyState(page: Page): Promise<void> {
+  // Guard: if session rows appear, a prior spec polluted the shared daemon — see 00-sessions-empty.spec.ts.
+  await expect(page.locator(AGENT_ROW_SELECTOR)).toHaveCount(0, { timeout: 5_000 });
+  await expect(page.getByText("No sessions yet")).toBeVisible({ timeout: 30_000 });
+}
+
+export async function archiveAgentFromSessions(
+  page: Page,
+  input: { agentId: string; title: string },
+): Promise<void> {
+  const row = getSessionRowByTitle(page, input.title);
+  await expect(row).toBeVisible({ timeout: 30_000 });
+  const box = await row.boundingBox();
+  if (!box) {
+    throw new Error(`Could not read bounding box for session row ${input.agentId}.`);
+  }
+
+  // Long-press the row. Idle agents are archived immediately (no modal).
+  // Running/initializing agents show a confirmation modal instead.
+  await page.mouse.move(box.x + box.width / 2, box.y + box.height / 2);
+  await page.mouse.down();
+  await page.waitForTimeout(900);
+  await page.mouse.up();
+
+  // If a confirmation modal appears (running agent), click the archive button.
+  const archiveButton = page.getByTestId("agent-action-archive").first();
+  const modalVisible = await archiveButton.isVisible().catch(() => false);
+  if (modalVisible) {
+    await archiveButton.click();
+  }
+
+  await expectSessionRowArchived(page, input.title);
+}

packages/app/e2e/helpers/composer.ts

@@ -0,0 +1,195 @@
+import { expect, type Page } from "@playwright/test";
+import { buildHostWorkspaceRoute } from "@/utils/host-routes";
+import { createTempGitRepo } from "./workspace";
+import { connectTerminalClient, type TerminalPerfDaemonClient } from "./terminal-perf";
+import { connectWorkspaceSetupClient, openHomeWithProject } from "./workspace-setup";
+import { selectWorkspaceInSidebar } from "./sidebar";
+import { waitForTabBar } from "./launcher";
+
+function composerInput(page: Page) {
+  return page.getByRole("textbox", { name: "Message agent..." }).first();
+}
+
+export function composerLocator(page: Page) {
+  return composerInput(page);
+}
+
+export async function expectComposerVisible(
+  page: Page,
+  options?: { timeout?: number },
+): Promise<void> {
+  await expect(composerInput(page)).toBeVisible({ timeout: options?.timeout ?? 15_000 });
+}
+
+export async function expectComposerDisabled(page: Page): Promise<void> {
+  // React Native TextInput with editable={false} renders as <textarea readonly> on web,
+  // not <textarea disabled>. Use not.toBeEditable() to match either form.
+  await expect(composerInput(page)).not.toBeEditable({ timeout: 10_000 });
+}
+
+export async function expectComposerDraft(page: Page, text: string): Promise<void> {
+  await expect(composerInput(page)).toHaveValue(text, { timeout: 5_000 });
+}
+
+export async function expectComposerEditable(page: Page): Promise<void> {
+  await expect(composerInput(page)).toBeEditable({ timeout: 15_000 });
+}
+
+export async function submitMessage(page: Page, text: string): Promise<void> {
+  const input = composerInput(page);
+  await expect(input).toBeEditable({ timeout: 30_000 });
+  await input.fill(text);
+  await input.press("Enter");
+}
+
+export async function fillComposerDraft(page: Page, text: string): Promise<void> {
+  await composerInput(page).fill(text);
+}
+
+export async function sendDraftToQueue(page: Page): Promise<void> {
+  await composerInput(page).press("Control+Enter");
+}
+
+export async function expectQueuedMessageButton(page: Page): Promise<void> {
+  await expect(page.getByRole("button", { name: "Send queued message now" })).toBeVisible({
+    timeout: 10_000,
+  });
+}
+
+export async function cancelAgent(page: Page): Promise<void> {
+  const stopButton = page.getByRole("button", { name: /stop|cancel/i }).first();
+  await expect(stopButton).toBeVisible({ timeout: 10_000 });
+  await stopButton.click();
+}
+
+/** Escape is bound to the "agent.interrupt" keyboard shortcut. */
+export async function pressInterruptShortcut(page: Page): Promise<void> {
+  await page.keyboard.press("Escape");
+}
+
+export async function openAttachmentMenu(page: Page): Promise<void> {
+  await page.getByTestId("message-input-attach-button").filter({ visible: true }).first().click();
+  await expect(page.getByTestId("message-input-attachment-menu")).toBeVisible({ timeout: 5_000 });
+}
+
+export async function expectAttachButtonDisabled(page: Page): Promise<void> {
+  await expect(
+    page.getByTestId("message-input-attach-button").filter({ visible: true }).first(),
+  ).toBeDisabled({ timeout: 10_000 });
+}
+
+export async function attachImageFromMenu(
+  page: Page,
+  file: { name: string; mimeType: string; buffer: Buffer },
+): Promise<void> {
+  const chooserPromise = page.waitForEvent("filechooser", { timeout: 10_000 });
+  await openAttachmentMenu(page);
+  await page.getByTestId("message-input-attachment-menu-item-image").click();
+  const chooser = await chooserPromise;
+  await chooser.setFiles([file]);
+}
+
+export async function expectAttachmentPill(page: Page, testID: string): Promise<void> {
+  await expect(page.getByTestId(testID).first()).toBeVisible({ timeout: 10_000 });
+}
+
+/** Hover to reveal the X button (hidden until hover on desktop web), then click by accessible label. */
+export async function removeAttachmentPill(
+  page: Page,
+  pillTestId: string,
+  removeAccessibilityLabel: string,
+): Promise<void> {
+  await page.getByTestId(pillTestId).first().hover();
+  await page.getByRole("button", { name: removeAccessibilityLabel }).first().click();
+}
+
+export async function expectGithubAttachmentPill(
+  page: Page,
+  input: { number: number; title: string },
+): Promise<void> {
+  const pill = page.getByTestId("composer-github-attachment-pill").first();
+  await expect(pill).toBeVisible({ timeout: 10_000 });
+  await expect(pill).toContainText(`#${input.number}`);
+  await expect(pill).toContainText(input.title);
+}
+
+export async function openImageLightbox(page: Page): Promise<void> {
+  await page.getByRole("button", { name: "Open image attachment" }).first().click();
+  await expect(page.getByTestId("attachment-lightbox-close")).toBeVisible({ timeout: 5_000 });
+}
+
+export async function closeImageLightbox(page: Page): Promise<void> {
+  await page.keyboard.press("Escape");
+  await expect(page.getByTestId("attachment-lightbox-close")).not.toBeVisible({ timeout: 5_000 });
+}
+
+export async function openGithubPickerFromMenu(page: Page): Promise<void> {
+  await openAttachmentMenu(page);
+  await page.getByTestId("message-input-attachment-menu-item-github").click();
+  await expect(page.getByTestId("combobox-desktop-container")).toBeVisible({ timeout: 5_000 });
+}
+
+/** Open picker, type a query, wait for the matching option by id (e.g. "issue:3", "pr:1"), and click it. */
+export async function selectGithubOption(
+  page: Page,
+  searchTerm: string,
+  optionId: string,
+): Promise<void> {
+  await openGithubPickerFromMenu(page);
+  const searchInput = page.getByPlaceholder("Search issues and PRs...");
+  await expect(searchInput).toBeVisible({ timeout: 5_000 });
+  await searchInput.fill(searchTerm);
+  const option = page.getByTestId(`composer-github-option-${optionId}`);
+  await expect(option).toBeVisible({ timeout: 15_000 });
+  await option.click();
+}
+
+export interface MockAgentSetup {
+  client: TerminalPerfDaemonClient;
+  repo: Awaited<ReturnType<typeof createTempGitRepo>>;
+}
+
+/** Create a temp repo, start a mock agent, navigate to it, and wait for it to be running. */
+export async function startRunningMockAgent(
+  page: Page,
+  opts: { prefix: string; model: string; prompt: string },
+): Promise<MockAgentSetup> {
+  const serverId = process.env.E2E_SERVER_ID;
+  if (!serverId) throw new Error("E2E_SERVER_ID is not set.");
+
+  const repo = await createTempGitRepo(opts.prefix);
+  const client = await connectTerminalClient();
+  const opened = await client.openProject(repo.path);
+  if (!opened.workspace) throw new Error(opened.error ?? "Failed to open project");
+  const agent = await client.createAgent({
+    provider: "mock",
+    cwd: repo.path,
+    model: opts.model,
+    initialPrompt: opts.prompt,
+  });
+  const agentUrl = `${buildHostWorkspaceRoute(serverId, repo.path)}?open=${encodeURIComponent(`agent:${agent.id}`)}`;
+  await page.goto(agentUrl);
+  await expect(page.getByRole("button", { name: /stop|cancel/i }).first()).toBeVisible({
+    timeout: 30_000,
+  });
+  await expectComposerVisible(page);
+  return { client, repo };
+}
+
+export interface GithubWorkspaceHandle {
+  cleanup: () => Promise<void>;
+}
+
+/** Open a workspace backed by an existing repo path (e.g. a cloned GitHub repo). */
+export async function openGithubWorkspace(
+  page: Page,
+  repoPath: string,
+): Promise<GithubWorkspaceHandle> {
+  const client = await connectWorkspaceSetupClient();
+  const opened = await client.openProject(repoPath);
+  if (!opened.workspace) throw new Error(opened.error ?? `Failed to open project ${repoPath}`);
+  await openHomeWithProject(page, repoPath);
+  await selectWorkspaceInSidebar(page, opened.workspace.id);
+  await waitForTabBar(page);
+  return { cleanup: () => client.close().catch(() => undefined) };
+}

packages/app/e2e/helpers/daemon-registry.ts

@@ -2,7 +2,7 @@ export const TEST_HOST_LABEL = "localhost";
 
 export const TEST_PROVIDER_PREFERENCES = {
   claude: { model: "haiku" },
-  codex: { model: "gpt-5.1-codex-mini", thinkingOptionId: "low" },
+  codex: { model: "gpt-5.4-mini", thinkingOptionId: "low" },
 } as const;
 
 export function buildDirectTcpConnection(endpoint: string) {
@@ -30,10 +30,15 @@ export function buildSeededHost(input: {
   };
 }
 
+export const TEST_MOCK_PROVIDER_PREFERENCES = {
+  ...TEST_PROVIDER_PREFERENCES,
+  mock: { model: "ten-second-stream" },
+} as const;
+
 export function buildCreateAgentPreferences(serverId: string) {
   return {
     serverId,
-    provider: "codex" as const,
-    providerPreferences: TEST_PROVIDER_PREFERENCES,
+    provider: "mock" as const,
+    providerPreferences: TEST_MOCK_PROVIDER_PREFERENCES,
   };
 }

packages/app/e2e/helpers/desktop-updates.ts

@@ -0,0 +1,289 @@
+import { readFileSync } from "node:fs";
+import { expect, type Page } from "@playwright/test";
+import { openSettings } from "./app";
+import { openSettingsHost } from "./settings";
+
+interface DaemonApiStatus {
+  version: string;
+  serverId: string;
+  hostname: string;
+}
+
+interface PidFileContent {
+  pid: number;
+  desktopManaged: boolean;
+}
+
+export interface RealDaemonState {
+  version: string;
+  pid: number | null;
+  logPath: string;
+}
+
+/**
+ * Reads live state from the running E2E test daemon: version from the HTTP
+ * status endpoint, PID from the paseo.pid lock file, log path from the
+ * E2E_PASEO_HOME directory. Call this in Node test code (not in the browser).
+ */
+export async function loadRealDaemonState(): Promise<RealDaemonState> {
+  const port = process.env.E2E_DAEMON_PORT;
+  const paseoHome = process.env.E2E_PASEO_HOME;
+  if (!port) throw new Error("E2E_DAEMON_PORT not set — globalSetup must run first");
+  if (!paseoHome) throw new Error("E2E_PASEO_HOME not set — globalSetup must run first");
+
+  const resp = await fetch(`http://127.0.0.1:${port}/api/status`);
+  const data: DaemonApiStatus = await resp.json();
+
+  let pid: number | null = null;
+  try {
+    const raw = readFileSync(`${paseoHome}/paseo.pid`, "utf8");
+    const pidContent: PidFileContent = JSON.parse(raw);
+    pid = pidContent.pid ?? null;
+  } catch (err) {
+    // PID file may not be present yet on a very fresh daemon start
+    console.warn("[desktop-updates] paseo.pid not found:", err);
+  }
+
+  return { version: data.version, pid, logPath: `${paseoHome}/daemon.log` };
+}
+
+export interface DesktopBridgeConfig {
+  serverId: string;
+  updateAvailable?: boolean;
+  latestVersion?: string;
+  slowInstall?: boolean;
+  /** Initial PID reported by desktop_daemon_status. Defaults to null. */
+  daemonPid?: number | null;
+  daemonVersion?: string | null;
+  daemonLogPath?: string;
+  /** Initial manageBuiltInDaemon setting. Defaults to false. */
+  manageBuiltInDaemon?: boolean;
+  /**
+   * Controls what dialog.ask returns when the daemon management confirm dialog
+   * fires. True = confirm (proceed with the action), false = cancel. Defaults to
+   * false so tests that only assert copy don't inadvertently trigger state changes.
+   */
+  confirmShouldAccept?: boolean;
+}
+
+export interface ConfirmDialogCall {
+  message: string;
+  title: string | undefined;
+}
+
+declare global {
+  interface Window {
+    __capturedDialogCall: ConfirmDialogCall | undefined;
+  }
+}
+
+/**
+ * Injects window.paseoDesktop before app load so all Electron-gated code
+ * activates. The update-check IPC is mocked at the boundary so the real
+ * auto-updater never fires. Daemon start/stop commands are stateful: the mock
+ * tracks running state and assigns a fresh PID on each start, letting tests
+ * observe PID changes without touching the real E2E daemon process.
+ * dialog.ask captures call arguments on window.__capturedDialogCall so tests
+ * can assert dialog copy without depending on window.confirm concatenation.
+ */
+export async function injectDesktopBridge(page: Page, config: DesktopBridgeConfig): Promise<void> {
+  await page.addInitScript((cfg) => {
+    // Mutable state shared across IPC calls within this page
+    let manageDaemon = cfg.manageBuiltInDaemon ?? false;
+    let daemonRunning = true;
+    let currentPid: number | null = cfg.daemonPid ?? null;
+    let startCount = 0;
+
+    function buildDaemonStatus() {
+      return {
+        serverId: cfg.serverId,
+        status: daemonRunning ? "running" : "stopped",
+        listen: null,
+        hostname: null,
+        pid: currentPid,
+        home: "",
+        version: cfg.daemonVersion ?? null,
+        desktopManaged: manageDaemon,
+        error: null,
+      };
+    }
+
+    (window as unknown as { paseoDesktop: unknown }).paseoDesktop = {
+      platform: "darwin",
+      invoke: async (command: string, args?: Record<string, unknown>) => {
+        if (command === "check_app_update") {
+          return cfg.updateAvailable
+            ? {
+                hasUpdate: true,
+                readyToInstall: true,
+                currentVersion: "1.0.0",
+                latestVersion: cfg.latestVersion ?? "1.2.3",
+                body: null,
+                date: null,
+              }
+            : {
+                hasUpdate: false,
+                readyToInstall: false,
+                currentVersion: "1.0.0",
+                latestVersion: null,
+                body: null,
+                date: null,
+              };
+        }
+
+        if (command === "install_app_update") {
+          if (cfg.slowInstall) {
+            await new Promise<void>((resolve) => setTimeout(resolve, 3000));
+          }
+          return {
+            installed: true,
+            version: cfg.latestVersion ?? "1.2.3",
+            message: "App update installed. Restart required.",
+          };
+        }
+
+        if (command === "desktop_daemon_status") {
+          return buildDaemonStatus();
+        }
+
+        if (command === "desktop_daemon_logs") {
+          return { logPath: cfg.daemonLogPath ?? "", contents: "" };
+        }
+
+        if (command === "get_desktop_settings") {
+          return {
+            releaseChannel: "stable",
+            daemon: { manageBuiltInDaemon: manageDaemon, keepRunningAfterQuit: true },
+          };
+        }
+
+        if (command === "patch_desktop_settings") {
+          const daemon = args?.daemon;
+          if (
+            daemon !== null &&
+            typeof daemon === "object" &&
+            "manageBuiltInDaemon" in daemon &&
+            typeof daemon.manageBuiltInDaemon === "boolean"
+          ) {
+            manageDaemon = daemon.manageBuiltInDaemon;
+          }
+          return {
+            releaseChannel: "stable",
+            daemon: { manageBuiltInDaemon: manageDaemon, keepRunningAfterQuit: true },
+          };
+        }
+
+        if (command === "stop_desktop_daemon") {
+          daemonRunning = false;
+          currentPid = null;
+          return buildDaemonStatus();
+        }
+
+        if (command === "start_desktop_daemon") {
+          startCount += 1;
+          daemonRunning = true;
+          // First start (bootstrap) returns the configured PID; subsequent starts
+          // (after a stop) get a fresh PID so tests can observe the change.
+          currentPid = (cfg.daemonPid ?? 10000) + (startCount - 1) * 1000;
+          return buildDaemonStatus();
+        }
+
+        return null;
+      },
+      dialog: {
+        ask: async (message: string, options?: Record<string, unknown>) => {
+          window.__capturedDialogCall = {
+            message,
+            title: typeof options?.title === "string" ? options.title : undefined,
+          };
+          return cfg.confirmShouldAccept ?? false;
+        },
+      },
+      getPendingOpenProject: async () => null,
+      events: { on: async () => () => undefined },
+    };
+  }, config);
+}
+
+export async function openDesktopSettings(page: Page, serverId: string): Promise<void> {
+  await openSettings(page);
+  await openSettingsHost(page, serverId);
+  await expect(page.getByTestId("host-page-daemon-lifecycle-card")).toBeVisible({
+    timeout: 15_000,
+  });
+}
+
+export async function expectUpdateBanner(page: Page, version: string): Promise<void> {
+  const callout = page.getByTestId("update-callout");
+  await expect(callout).toBeVisible({ timeout: 15_000 });
+  await expect(callout).toContainText(`v${version.replace(/^v/i, "")}`);
+}
+
+export async function clickInstallUpdate(page: Page): Promise<void> {
+  await page.getByRole("button", { name: "Install & restart" }).click();
+}
+
+export async function expectInstallInProgress(page: Page): Promise<void> {
+  await expect(page.getByRole("button", { name: "Installing..." })).toBeVisible();
+}
+
+/**
+ * Clicks the daemon management switch and waits for dialog.ask to fire in the
+ * mock, then returns the captured call args (message + title). The mock auto-
+ * dismisses via confirmShouldAccept=false so callers can assert copy without
+ * worrying about state changes.
+ */
+export async function interceptDaemonManagementConfirmDialog(
+  page: Page,
+): Promise<ConfirmDialogCall> {
+  await page.getByRole("switch", { name: "Manage built-in daemon" }).click();
+  await page.waitForFunction(() => !!window.__capturedDialogCall, { timeout: 5_000 });
+  return page.evaluate(() => window.__capturedDialogCall!);
+}
+
+export async function toggleDaemonManagement(
+  page: Page,
+  _action: "enable" | "disable",
+): Promise<void> {
+  await page.getByRole("switch", { name: "Manage built-in daemon" }).click();
+}
+
+export function expectDaemonManagementConfirmDialog(args: ConfirmDialogCall): void {
+  expect(args.title).toBe("Pause built-in daemon");
+  expect(args.message).toContain("stop the built-in daemon immediately");
+}
+
+export async function expectDaemonManagementEnabled(page: Page): Promise<void> {
+  await expect(page.getByRole("switch", { name: "Manage built-in daemon" })).toBeChecked();
+}
+
+export async function expectDaemonManagementDisabled(page: Page): Promise<void> {
+  await expect(page.getByRole("switch", { name: "Manage built-in daemon" })).not.toBeChecked();
+}
+
+/**
+ * Asserts the daemon status card shows the given PID. Pass null to assert
+ * the cleared state (shown as "PID —" when the daemon is stopped).
+ */
+export async function expectDaemonStatusPid(page: Page, pid: number | null): Promise<void> {
+  const expected = pid !== null ? `PID ${pid}` : "PID —";
+  await expect(
+    page.getByTestId("host-page-daemon-lifecycle-card").getByText(expected),
+  ).toBeVisible();
+}
+
+export async function expectDaemonStatusLogPath(page: Page, logPath: string): Promise<void> {
+  await expect(
+    page.getByTestId("host-page-daemon-lifecycle-card").getByText(logPath),
+  ).toBeVisible();
+}
+
+/**
+ * Asserts the host page identity badge shows the given version string.
+ * The badge is populated from the live WebSocket session's serverInfo.version.
+ */
+export async function expectDaemonStatusVersion(page: Page, version: string): Promise<void> {
+  await expect(
+    page.getByTestId("host-page-identity").getByText(version, { exact: false }),
+  ).toBeVisible({ timeout: 15_000 });
+}

packages/app/e2e/helpers/file-explorer.ts

@@ -0,0 +1,41 @@
+import { expect, type Page } from "@playwright/test";
+
+function fileExplorerTree(page: Page) {
+  return page.getByTestId("file-explorer-tree-scroll");
+}
+
+function fileExplorerEntry(page: Page, name: string) {
+  return fileExplorerTree(page).getByText(name, { exact: true }).first();
+}
+
+export async function openFileExplorer(page: Page): Promise<void> {
+  await page.getByRole("button", { name: "Open explorer" }).first().click();
+  await page.getByTestId("explorer-tab-files").click();
+  await expect(fileExplorerTree(page)).toBeVisible({ timeout: 30_000 });
+}
+
+export async function expandFolder(page: Page, folderName: string): Promise<void> {
+  await fileExplorerEntry(page, folderName).click();
+}
+
+export async function collapseFolder(page: Page, folderName: string): Promise<void> {
+  await fileExplorerEntry(page, folderName).click();
+}
+
+export async function openFileFromExplorer(page: Page, fileName: string): Promise<void> {
+  await fileExplorerEntry(page, fileName).click();
+}
+
+export async function expectExplorerEntryVisible(page: Page, name: string): Promise<void> {
+  await expect(fileExplorerEntry(page, name)).toBeVisible({ timeout: 30_000 });
+}
+
+export async function expectExplorerEntryHidden(page: Page, name: string): Promise<void> {
+  await expect(fileExplorerEntry(page, name)).toBeHidden({ timeout: 30_000 });
+}
+
+export async function expectFileTabOpen(page: Page, filePath: string): Promise<void> {
+  await expect(page.getByTestId(`workspace-tab-file_${filePath}`).first()).toBeVisible({
+    timeout: 30_000,
+  });
+}

packages/app/e2e/helpers/github-fixtures.ts

@@ -0,0 +1,249 @@
+import { execFileSync, execSync } from "node:child_process";
+import { mkdtemp, rm, writeFile } from "node:fs/promises";
+import path from "node:path";
+
+export function hasGithubAuth(): boolean {
+  try {
+    execSync("gh auth status", { stdio: "ignore" });
+    return true;
+  } catch {
+    return false;
+  }
+}
+
+export interface CheckSpec {
+  context: string;
+  state: "success" | "failure" | "pending";
+}
+
+export interface PrSpec {
+  title: string;
+  state: "open" | "merged" | "closed" | "draft";
+  checks?: CheckSpec[];
+  commentCount?: number;
+}
+
+export interface IssueSpec {
+  title: string;
+  body?: string;
+  labels?: string[];
+  state?: "open" | "closed";
+}
+
+export interface GhPrFixture {
+  number: number;
+  title: string;
+  url: string;
+  branch: string;
+  localPath: string;
+}
+
+export interface GhIssueFixture {
+  number: number;
+  title: string;
+  url: string;
+}
+
+export interface GhRepoFixture {
+  owner: string;
+  name: string;
+  fullName: string;
+  prs: GhPrFixture[];
+  issues: GhIssueFixture[];
+  cleanup(): Promise<void>;
+}
+
+function gh(args: string[], opts?: { cwd?: string }): string {
+  return execFileSync("gh", args, {
+    cwd: opts?.cwd,
+    encoding: "utf8",
+    stdio: ["ignore", "pipe", "pipe"],
+  }).trim();
+}
+
+function git(args: string[], cwd: string): string {
+  return execFileSync("git", args, {
+    cwd,
+    encoding: "utf8",
+    stdio: ["ignore", "pipe", "pipe"],
+  }).trim();
+}
+
+async function seedPr(args: {
+  spec: PrSpec;
+  branch: string;
+  index: number;
+  basePath: string;
+  authedUrl: string;
+  fullName: string;
+  repoName: string;
+}): Promise<{ fixture: GhPrFixture; localPath: string }> {
+  const { spec, branch, index, basePath, authedUrl, fullName, repoName } = args;
+
+  const createArgs = [
+    "pr",
+    "create",
+    "--title",
+    spec.title,
+    "--base",
+    "main",
+    "--head",
+    branch,
+    "--body",
+    "",
+  ];
+  if (spec.state === "draft") createArgs.push("--draft");
+
+  const prUrl = gh(createArgs, { cwd: basePath });
+  const prNumber = parseInt(prUrl.split("/").pop() ?? "0", 10);
+
+  if (spec.checks && spec.checks.length > 0) {
+    const sha = git(["rev-parse", branch], basePath);
+    for (const check of spec.checks) {
+      gh([
+        "api",
+        `repos/${fullName}/statuses/${sha}`,
+        "--method",
+        "POST",
+        "-f",
+        `state=${check.state}`,
+        "-f",
+        `context=${check.context}`,
+        "-f",
+        `target_url=https://example.com/${encodeURIComponent(check.context)}`,
+      ]);
+    }
+  }
+
+  for (let j = 0; j < (spec.commentCount ?? 0); j++) {
+    gh(["pr", "comment", String(prNumber), "--body", `Test comment ${j + 1}`], { cwd: basePath });
+  }
+
+  if (spec.state === "merged") {
+    gh(["pr", "merge", String(prNumber), "--merge"], { cwd: basePath });
+  } else if (spec.state === "closed") {
+    gh(["pr", "close", String(prNumber)], { cwd: basePath });
+  }
+
+  const localPath = await mkdtemp(path.join("/tmp", `${repoName}-ws-${index}-`));
+  git(["clone", authedUrl, localPath, "--quiet", "-b", branch], basePath);
+  // Clean remote URL (no embedded token) so gh can parse owner/repo
+  git(["remote", "set-url", "origin", `https://github.com/${fullName}.git`], localPath);
+  git(["config", "user.email", "e2e@paseo.test"], localPath);
+  git(["config", "user.name", "Paseo E2E"], localPath);
+  git(["config", "commit.gpgsign", "false"], localPath);
+
+  return {
+    fixture: { number: prNumber, title: spec.title, url: prUrl, branch, localPath },
+    localPath,
+  };
+}
+
+function seedIssue(args: { spec: IssueSpec; basePath: string }): GhIssueFixture {
+  const { spec, basePath } = args;
+  const createArgs = ["issue", "create", "--title", spec.title, "--body", spec.body ?? ""];
+  for (const label of spec.labels ?? []) {
+    createArgs.push("--label", label);
+  }
+  const issueUrl = gh(createArgs, { cwd: basePath });
+  const issueNumber = parseInt(issueUrl.split("/").pop() ?? "0", 10);
+  if (spec.state === "closed") {
+    gh(["issue", "close", String(issueNumber)], { cwd: basePath });
+  }
+  return { number: issueNumber, title: spec.title, url: issueUrl };
+}
+
+// Single namespace for temporary GitHub repos created by Paseo tests.
+// Bulk cleanup relies on this prefix being unmistakable — never reuse `paseo-`
+// (collides with real repos like `paseo`, `paseo-website`).
+const TEMP_GITHUB_REPO_PREFIX = "paseotmp-";
+
+export async function createTempGithubRepo(options: {
+  category: string;
+  prs?: PrSpec[];
+  issues?: IssueSpec[];
+}): Promise<GhRepoFixture> {
+  const { category, prs = [], issues = [] } = options;
+  const uniqueSuffix = `${Date.now()}-${Math.random().toString(36).slice(2, 7)}`;
+  const repoName = `${TEMP_GITHUB_REPO_PREFIX}${category}-${uniqueSuffix}`;
+
+  // Bootstrap local git repo
+  const basePath = await mkdtemp(path.join("/tmp", `${repoName}-base-`));
+  git(["init", "-b", "main"], basePath);
+  git(["config", "user.email", "e2e@paseo.test"], basePath);
+  git(["config", "user.name", "Paseo E2E"], basePath);
+  git(["config", "commit.gpgsign", "false"], basePath);
+  await writeFile(path.join(basePath, "README.md"), "# E2E Test Repo\n");
+  git(["add", "README.md"], basePath);
+  git(["commit", "-m", "Initial commit"], basePath);
+
+  // Create GitHub repo and push initial commit
+  gh(["repo", "create", repoName, "--private", `--source=${basePath}`, "--push"]);
+
+  const owner = gh(["api", "user", "--jq", ".login"]);
+  const fullName = `${owner}/${repoName}`;
+  const token = gh(["auth", "token"]);
+  const authedUrl = `https://x-access-token:${token}@github.com/${fullName}.git`;
+
+  // Switch remote to authed URL for subsequent pushes
+  git(["remote", "set-url", "origin", authedUrl], basePath);
+
+  // Create a branch + commit for each PR spec
+  const branches: string[] = [];
+  for (let i = 0; i < prs.length; i++) {
+    const branch = `pr-branch-${i + 1}`;
+    branches.push(branch);
+    git(["checkout", "-b", branch], basePath);
+    await writeFile(path.join(basePath, `pr-${i + 1}.txt`), `PR ${i + 1}\n`);
+    git(["add", `pr-${i + 1}.txt`], basePath);
+    git(["commit", "-m", `Add PR ${i + 1}`], basePath);
+    git(["checkout", "main"], basePath);
+  }
+
+  if (branches.length > 0) {
+    git(["push", "origin", ...branches], basePath);
+  }
+
+  // Create PRs, seed checks/comments, apply state changes, clone workspaces
+  const prFixtures: GhPrFixture[] = [];
+  const localPaths: string[] = [];
+
+  for (let i = 0; i < prs.length; i++) {
+    const { fixture, localPath } = await seedPr({
+      spec: prs[i],
+      branch: branches[i],
+      index: i,
+      basePath,
+      authedUrl,
+      fullName,
+      repoName,
+    });
+    localPaths.push(localPath);
+    prFixtures.push(fixture);
+  }
+
+  // Create issues
+  const issueFixtures: GhIssueFixture[] = [];
+  for (const spec of issues) {
+    issueFixtures.push(seedIssue({ spec, basePath }));
+  }
+
+  return {
+    owner,
+    name: repoName,
+    fullName,
+    prs: prFixtures,
+    issues: issueFixtures,
+    cleanup: async () => {
+      try {
+        gh(["repo", "delete", fullName, "--yes"]);
+      } catch {
+        // Best-effort cleanup
+      }
+      await Promise.all([
+        rm(basePath, { recursive: true, force: true }),
+        ...localPaths.map((p) => rm(p, { recursive: true, force: true })),
+      ]);
+    },
+  };
+}

packages/app/e2e/helpers/launcher.ts

@@ -0,0 +1,197 @@
+import { expect, type Page } from "@playwright/test";
+import { buildHostWorkspaceRoute } from "../../src/utils/host-routes";
+import { createTempGitRepo } from "./workspace";
+
+// ─── Navigation ────────────────────────────────────────────────────────────
+
+function getServerId(): string {
+  const serverId = process.env.E2E_SERVER_ID;
+  if (!serverId) {
+    throw new Error("E2E_SERVER_ID is not set (expected from Playwright globalSetup).");
+  }
+  return serverId;
+}
+
+/** Navigate to a workspace and wait for the tab bar to appear. */
+export async function gotoWorkspace(page: Page, cwd: string): Promise<void> {
+  const route = buildHostWorkspaceRoute(getServerId(), cwd);
+  await page.goto(route);
+  await waitForTabBar(page);
+}
+
+// ─── Tab bar queries ───────────────────────────────────────────────────────
+
+/** Wait for the workspace tab bar to be visible. */
+export async function waitForTabBar(page: Page): Promise<void> {
+  await expect(
+    page.getByTestId("workspace-tabs-row").filter({ visible: true }).first(),
+  ).toBeVisible({
+    timeout: 30_000,
+  });
+}
+
+/** Return all tab test IDs currently in the tab bar. */
+export async function getTabTestIds(page: Page): Promise<string[]> {
+  const tabs = page
+    .locator('[data-testid^="workspace-tab-"]:not([data-testid^="workspace-tab-context-"])')
+    .filter({ visible: true });
+  const count = await tabs.count();
+  const ids: string[] = [];
+  for (let i = 0; i < count; i++) {
+    const testId = await tabs.nth(i).getAttribute("data-testid");
+    if (testId) ids.push(testId);
+  }
+  return ids;
+}
+
+/** Return the number of tabs matching a kind prefix (e.g. "launcher", "draft", "terminal", "agent"). */
+export async function countTabsOfKind(page: Page, kind: string): Promise<number> {
+  const ids = await getTabTestIds(page);
+  return ids.filter((id) => id.includes(kind)).length;
+}
+
+/** Return the currently active tab's test ID (the one with aria-selected or focus styling). */
+export async function getActiveTabTestId(page: Page): Promise<string | null> {
+  // Active tab has the focused highlight — check for the aria-selected or data-active attribute
+  const activeTab = page
+    .locator(
+      '[data-testid^="workspace-tab-"]:not([data-testid^="workspace-tab-context-"])[aria-selected="true"]',
+    )
+    .filter({ visible: true })
+    .first();
+  if (await activeTab.isVisible().catch(() => false)) {
+    return activeTab.getAttribute("data-testid");
+  }
+  // Fallback: the tab with focused styling
+  return null;
+}
+
+// ─── Tab actions ───────────────────────────────────────────────────────────
+
+/** Press Cmd+T (macOS) or Ctrl+T (Linux/Windows) to open a new tab. */
+export async function pressNewTabShortcut(page: Page): Promise<void> {
+  const modifier = process.platform === "darwin" ? "Meta" : "Control";
+  await page.keyboard.press(`${modifier}+t`);
+}
+
+// ─── Tab bar assertions ───────────────────────────────────────────────────
+
+/** Assert the new agent tab button is visible in the tab bar. */
+export async function assertNewChatTileVisible(page: Page): Promise<void> {
+  await expect(
+    page.getByTestId("workspace-new-agent-tab").filter({ visible: true }).first(),
+  ).toBeVisible();
+}
+
+/** Assert the new terminal button is visible in the tab bar. */
+export async function assertTerminalTileVisible(page: Page): Promise<void> {
+  await expect(
+    page.getByTestId("workspace-new-terminal").filter({ visible: true }).first(),
+  ).toBeVisible();
+}
+
+// ─── Tab creation actions ─────────────────────────────────────────────────
+
+/** Click the new agent tab button to create a draft/chat tab. */
+export async function clickNewChat(page: Page): Promise<void> {
+  const button = page.getByTestId("workspace-new-agent-tab").filter({ visible: true }).first();
+  await expect(button).toBeVisible({ timeout: 10_000 });
+  await button.click();
+}
+
+/** Click the new terminal button to create a terminal tab. */
+export async function clickNewTerminal(page: Page): Promise<void> {
+  const button = page.getByTestId("workspace-new-terminal").filter({ visible: true }).first();
+  await expect(button).toBeVisible({ timeout: 10_000 });
+  await button.click();
+}
+
+// ─── Tab title assertions ──────────────────────────────────────────────────
+
+/** Wait for any tab in the bar to display the given title text. */
+export async function waitForTabWithTitle(
+  page: Page,
+  title: string | RegExp,
+  timeout = 30_000,
+): Promise<void> {
+  const matcher = typeof title === "string" ? new RegExp(title, "i") : title;
+  await expect(
+    page
+      .locator('[data-testid^="workspace-tab-"]:not([data-testid^="workspace-tab-context-"])')
+      .filter({ hasText: matcher })
+      .filter({ visible: true })
+      .first(),
+  ).toBeVisible({ timeout });
+}
+
+/** Assert the new agent tab button is visible in the tab bar. */
+export async function assertSingleNewTabButton(page: Page): Promise<void> {
+  const buttons = page.getByTestId("workspace-new-agent-tab").filter({ visible: true });
+  const count = await buttons.count();
+  expect(count).toBeGreaterThanOrEqual(1);
+}
+
+// ─── No-flash measurement ──────────────────────────────────────────────────
+
+/**
+ * Measure the time between clicking a launcher tile and the replacement panel becoming visible.
+ * Returns elapsed milliseconds.
+ */
+export async function measureTileTransition(
+  page: Page,
+  clickAction: () => Promise<void>,
+  successLocator: ReturnType<Page["locator"]>,
+  timeout = 5_000,
+): Promise<number> {
+  const start = Date.now();
+  await clickAction();
+  await expect(successLocator).toBeVisible({ timeout });
+  return Date.now() - start;
+}
+
+/**
+ * Sample tab IDs at high frequency across a transition to detect blank/intermediate states.
+ * Returns all unique snapshots observed.
+ */
+export async function sampleTabsDuringTransition(
+  page: Page,
+  action: () => Promise<void>,
+  durationMs = 2_000,
+  intervalMs = 30,
+): Promise<string[][]> {
+  const snapshots: string[][] = [];
+  const startSampling = async () => {
+    const start = Date.now();
+    while (Date.now() - start < durationMs) {
+      snapshots.push(await getTabTestIds(page));
+      await page.waitForTimeout(intervalMs);
+    }
+  };
+
+  const samplingPromise = startSampling();
+  await action();
+  await samplingPromise;
+  return snapshots;
+}
+
+export function terminalSurfaceLocator(page: Page) {
+  return page.locator('[data-testid="terminal-surface"]').first();
+}
+
+export async function expectAgentTabActive(page: Page, agentId: string): Promise<void> {
+  const tabTestId = `workspace-tab-agent_${agentId}`;
+  await expect(page.getByTestId(tabTestId).filter({ visible: true })).toHaveAttribute(
+    "aria-selected",
+    "true",
+  );
+  await expect(getActiveTabTestId(page)).resolves.toBe(tabTestId);
+}
+
+// ─── Workspace setup ───────────────────────────────────────────────────────
+
+/** Create a temp git repo and return its path with a cleanup function. */
+export async function createWorkspace(
+  prefix = "launcher-e2e-",
+): ReturnType<typeof createTempGitRepo> {
+  return createTempGitRepo(prefix);
+}

packages/app/e2e/helpers/new-workspace.ts

@@ -0,0 +1,395 @@
+import { randomUUID } from "node:crypto";
+import path from "node:path";
+import { pathToFileURL } from "node:url";
+import { expect, type Page } from "@playwright/test";
+import type { DaemonClient as ServerDaemonClient } from "@server/client/daemon-client";
+import { decodeWorkspaceIdFromPathSegment } from "@/utils/host-routes";
+import { expectWorkspaceHeader, workspaceLabelFromPath } from "./workspace-ui";
+import { createNodeWebSocketFactory, type NodeWebSocketFactory } from "./node-ws-factory";
+
+type NewWorkspaceDaemonClient = Pick<
+  ServerDaemonClient,
+  | "archivePaseoWorktree"
+  | "archiveWorkspace"
+  | "close"
+  | "connect"
+  | "createPaseoWorktree"
+  | "openProject"
+>;
+
+interface NewWorkspaceDaemonClientConfig {
+  url: string;
+  clientId: string;
+  clientType: "cli";
+  webSocketFactory?: NodeWebSocketFactory;
+}
+
+type OpenProjectPayload = Awaited<ReturnType<NewWorkspaceDaemonClient["openProject"]>>;
+
+export interface OpenedProject {
+  workspaceId: string;
+  projectKey: string;
+  projectDisplayName: string;
+  workspaceName: string;
+}
+
+function getDaemonPort(): string {
+  const daemonPort = process.env.E2E_DAEMON_PORT;
+  if (!daemonPort) {
+    throw new Error("E2E_DAEMON_PORT is not set.");
+  }
+  if (daemonPort === "6767") {
+    throw new Error("E2E_DAEMON_PORT must not point at the developer daemon.");
+  }
+  return daemonPort;
+}
+
+function getDaemonWsUrl(): string {
+  return `ws://127.0.0.1:${getDaemonPort()}/ws`;
+}
+
+async function loadDaemonClientConstructor(): Promise<
+  new (config: NewWorkspaceDaemonClientConfig) => NewWorkspaceDaemonClient
+> {
+  const repoRoot = path.resolve(__dirname, "../../../../");
+  const moduleUrl = pathToFileURL(
+    path.join(repoRoot, "packages/server/dist/server/server/exports.js"),
+  ).href;
+  const mod = (await import(moduleUrl)) as {
+    DaemonClient: new (config: NewWorkspaceDaemonClientConfig) => NewWorkspaceDaemonClient;
+  };
+  return mod.DaemonClient;
+}
+
+function requireWorkspace(payload: OpenProjectPayload) {
+  if (payload.error) {
+    throw new Error(payload.error);
+  }
+  if (!payload.workspace) {
+    throw new Error("openProject returned no workspace.");
+  }
+  return payload.workspace;
+}
+
+function parseWorkspaceIdFromPageUrl(page: Page, serverId: string): string | null {
+  const pathname = new URL(page.url()).pathname;
+  const match = pathname.match(
+    new RegExp(`^/h/${serverId.replace(/[.*+?^${}()|[\]\\]/g, "\\$&")}/workspace/([^/?#]+)`),
+  );
+  if (!match?.[1]) {
+    return null;
+  }
+  return decodeWorkspaceIdFromPathSegment(match[1]);
+}
+
+export async function connectNewWorkspaceDaemonClient(): Promise<NewWorkspaceDaemonClient> {
+  const DaemonClient = await loadDaemonClientConstructor();
+  const webSocketFactory = createNodeWebSocketFactory();
+  const client = new DaemonClient({
+    url: getDaemonWsUrl(),
+    clientId: `app-e2e-new-workspace-${randomUUID()}`,
+    clientType: "cli",
+    webSocketFactory,
+  });
+  await client.connect();
+  return client;
+}
+
+export async function openProjectViaDaemon(
+  client: NewWorkspaceDaemonClient,
+  repoPath: string,
+): Promise<OpenedProject> {
+  const workspace = requireWorkspace(await client.openProject(repoPath));
+  return {
+    workspaceId: workspace.id,
+    projectKey: workspace.projectId,
+    projectDisplayName: workspace.projectDisplayName,
+    workspaceName: workspace.name,
+  };
+}
+
+export async function archiveWorkspaceFromDaemon(
+  client: NewWorkspaceDaemonClient,
+  workspaceId: string,
+): Promise<void> {
+  const payload = await client.archivePaseoWorktree({ worktreePath: workspaceId });
+  if (payload.error) {
+    throw new Error(payload.error.message);
+  }
+  if (!payload.success) {
+    throw new Error(`Failed to archive workspace: ${workspaceId}`);
+  }
+}
+
+export async function archiveLocalWorkspaceFromDaemon(
+  client: NewWorkspaceDaemonClient,
+  workspaceId: string,
+): Promise<void> {
+  const payload = await client.archiveWorkspace(workspaceId);
+  if (payload.error) {
+    throw new Error(payload.error);
+  }
+  if (!payload.archivedAt) {
+    throw new Error(`Failed to archive workspace: ${workspaceId}`);
+  }
+}
+
+export async function createWorktreeViaDaemon(
+  client: NewWorkspaceDaemonClient,
+  input: { cwd: string; slug: string },
+): Promise<OpenedProject> {
+  const payload = await client.createPaseoWorktree({
+    cwd: input.cwd,
+    worktreeSlug: input.slug,
+  });
+  const workspace = requireWorkspace(payload);
+  return {
+    workspaceId: workspace.id,
+    projectKey: workspace.projectId,
+    projectDisplayName: workspace.projectDisplayName,
+    workspaceName: workspace.name,
+  };
+}
+
+export async function openNewWorkspaceComposer(
+  page: Page,
+  input: { projectKey: string; projectDisplayName: string },
+): Promise<void> {
+  const projectRow = page.getByTestId(`sidebar-project-row-${input.projectKey}`).first();
+  await expect(projectRow).toBeVisible({ timeout: 30_000 });
+  await projectRow.hover();
+
+  const button = page.getByTestId(`sidebar-project-new-worktree-${input.projectKey}`).first();
+  await expect(button).toBeVisible({ timeout: 30_000 });
+  await button.click();
+
+  await expect(page).toHaveURL(/\/h\/[^/]+\/new(?:\?.*)?$/, {
+    timeout: 30_000,
+  });
+}
+
+export async function clickNewWorkspaceButton(
+  page: Page,
+  input: { projectKey: string; projectDisplayName: string; prompt?: string },
+): Promise<void> {
+  await openNewWorkspaceComposer(page, input);
+  const composer = page.getByRole("textbox", { name: "Message agent..." });
+  await expect(composer).toBeVisible({ timeout: 30_000 });
+  await composer.fill(input.prompt ?? "Hello from e2e");
+  const createButton = page
+    .getByTestId("message-input-root")
+    .getByRole("button", { name: "Create" });
+  await expect(createButton).toBeVisible({ timeout: 30_000 });
+  await createButton.click();
+}
+
+export async function openStartingRefPicker(page: Page): Promise<void> {
+  const trigger = page.getByTestId("new-workspace-ref-picker-trigger");
+  await expect(trigger).toBeVisible({ timeout: 30_000 });
+  await trigger.click();
+}
+
+export async function selectBranchInPicker(page: Page, name: string): Promise<void> {
+  const branchRow = page.getByTestId(`new-workspace-ref-picker-branch-${name}`);
+  await expect(branchRow).toBeVisible({ timeout: 30_000 });
+  await branchRow.click();
+}
+
+export async function selectGitHubPrInPicker(page: Page, number: number): Promise<void> {
+  const prRow = page.getByTestId(`new-workspace-ref-picker-pr-${number}`);
+  await expect(prRow).toBeVisible({ timeout: 30_000 });
+  await prRow.click();
+}
+
+export async function expectStartingRefPickerTriggerPr(
+  page: Page,
+  input: { number: number; title: string; headRef: string },
+): Promise<void> {
+  const trigger = page.getByRole("button", { name: "Starting ref" });
+  await expect(trigger).toContainText(`#${input.number}`);
+  await expect(trigger).toContainText(input.title);
+  await expect(trigger).not.toContainText(input.headRef);
+}
+
+export async function openBranchPicker(page: Page): Promise<void> {
+  const trigger = page.getByRole("button", { name: "Starting ref" });
+  await expect(trigger).toBeVisible({ timeout: 30_000 });
+  await trigger.click();
+}
+
+export async function selectPickerOptionByKeyboard(page: Page, label: string): Promise<void> {
+  const searchInput = page.getByPlaceholder("Search branches and PRs");
+  await expect(searchInput).toBeVisible({ timeout: 30_000 });
+  await page.keyboard.type(label);
+  await page.keyboard.press("ArrowDown");
+  await page.keyboard.press("Enter");
+}
+
+export async function closeBranchPicker(page: Page): Promise<void> {
+  await page.keyboard.press("Escape");
+}
+
+export async function expectPickerOpen(page: Page): Promise<void> {
+  await expect(page.getByTestId("combobox-desktop-container")).toBeVisible({ timeout: 30_000 });
+}
+
+export async function expectPickerClosed(page: Page): Promise<void> {
+  await expect(page.getByTestId("combobox-desktop-container")).not.toBeVisible({
+    timeout: 30_000,
+  });
+}
+
+export async function expectPickerSelected(page: Page, label: string): Promise<void> {
+  const trigger = page.getByRole("button", { name: "Starting ref" });
+  await expect(trigger).toContainText(label);
+}
+
+export async function expectComposerGithubAttachmentPill(
+  page: Page,
+  input: { number: number; title: string },
+): Promise<void> {
+  const pills = page.getByTestId("composer-github-attachment-pill");
+  await expect(pills).toHaveCount(1);
+  await expect(pills.first()).toContainText(`#${input.number}`);
+  await expect(pills.first()).toContainText(input.title);
+}
+
+export async function assertNewWorkspaceSidebarAndHeader(
+  page: Page,
+  input: { serverId: string; previousWorkspaceId: string; projectDisplayName: string },
+): Promise<{ workspaceId: string }> {
+  // Wait for URL to redirect to the newly created workspace.
+  // Uses URL as source of truth to avoid picking up sidebar rows from concurrent tests.
+  let workspaceId: string | null = null;
+  const deadline = Date.now() + 60_000;
+  while (Date.now() < deadline) {
+    workspaceId = parseWorkspaceIdFromPageUrl(page, input.serverId);
+    if (workspaceId && workspaceId !== input.previousWorkspaceId) {
+      break;
+    }
+    await page.waitForTimeout(250);
+  }
+
+  if (!workspaceId || workspaceId === input.previousWorkspaceId) {
+    throw new Error(`Expected URL to redirect to a new workspace.\nCurrent URL: ${page.url()}`);
+  }
+
+  const createdWorkspaceRow = page.getByTestId(
+    `sidebar-workspace-row-${input.serverId}:${workspaceId}`,
+  );
+  await expect(createdWorkspaceRow.first()).toBeVisible({ timeout: 30_000 });
+
+  await expectWorkspaceHeader(page, {
+    title: workspaceLabelFromPath(workspaceId),
+    subtitle: input.projectDisplayName,
+  });
+
+  return { workspaceId };
+}
+
+type WebSocketMessage = string | Buffer;
+
+function parseWebSocketJson(message: WebSocketMessage): unknown {
+  const rawMessage = typeof message === "string" ? message : message.toString("utf8");
+  try {
+    return JSON.parse(rawMessage);
+  } catch {
+    return null;
+  }
+}
+
+function getSessionMessage(message: WebSocketMessage): Record<string, unknown> | null {
+  const envelope = parseWebSocketJson(message);
+  if (!envelope || typeof envelope !== "object") {
+    return null;
+  }
+  const maybeEnvelope = envelope as { type?: unknown; message?: unknown };
+  if (maybeEnvelope.type !== "session" || !maybeEnvelope.message) {
+    return null;
+  }
+  if (typeof maybeEnvelope.message !== "object") {
+    return null;
+  }
+  return maybeEnvelope.message as Record<string, unknown>;
+}
+
+function getStringField(input: Record<string, unknown>, key: string): string | null {
+  const value = input[key];
+  return typeof value === "string" ? value : null;
+}
+
+export interface AgentCreatedDelayControl {
+  release(): void;
+  waitForCreateRequest(): Promise<void>;
+  waitForDelayedCreatedStatus(): Promise<void>;
+}
+
+export async function delayBrowserAgentCreatedStatus(
+  page: Page,
+): Promise<AgentCreatedDelayControl> {
+  const daemonPort = process.env.E2E_DAEMON_PORT;
+  if (!daemonPort) {
+    throw new Error("E2E_DAEMON_PORT is not set.");
+  }
+
+  const daemonPortPattern = new RegExp(`:${daemonPort.replace(/[.*+?^${}()|[\]\\]/g, "\\$&")}\\b`);
+  const createRequestIds = new Set<string>();
+  const delayedForwards: Array<() => void> = [];
+  let releaseRequested = false;
+  let resolveCreateRequest: (() => void) | null = null;
+  let resolveDelayedCreatedStatus: (() => void) | null = null;
+  const createRequestSeen = new Promise<void>((resolve) => {
+    resolveCreateRequest = resolve;
+  });
+  const delayedCreatedStatusSeen = new Promise<void>((resolve) => {
+    resolveDelayedCreatedStatus = resolve;
+  });
+
+  await page.routeWebSocket(daemonPortPattern, (ws) => {
+    const server = ws.connectToServer();
+
+    ws.onMessage((message) => {
+      const sessionMessage = getSessionMessage(message);
+      if (sessionMessage?.type === "create_agent_request") {
+        const requestId = getStringField(sessionMessage, "requestId");
+        if (requestId) {
+          createRequestIds.add(requestId);
+          resolveCreateRequest?.();
+        }
+      }
+      server.send(message);
+    });
+
+    server.onMessage((message) => {
+      const sessionMessage = getSessionMessage(message);
+      const payload =
+        sessionMessage?.type === "status" && typeof sessionMessage.payload === "object"
+          ? (sessionMessage.payload as Record<string, unknown>)
+          : null;
+      const requestId = payload ? getStringField(payload, "requestId") : null;
+
+      if (payload?.status === "agent_created" && requestId && createRequestIds.has(requestId)) {
+        resolveDelayedCreatedStatus?.();
+        if (releaseRequested) {
+          ws.send(message);
+          return;
+        }
+        delayedForwards.push(() => ws.send(message));
+        return;
+      }
+
+      ws.send(message);
+    });
+  });
+
+  return {
+    release() {
+      releaseRequested = true;
+      for (const forward of delayedForwards.splice(0)) {
+        forward();
+      }
+    },
+    waitForCreateRequest: () => createRequestSeen,
+    waitForDelayedCreatedStatus: () => delayedCreatedStatusSeen,
+  };
+}

packages/app/e2e/helpers/node-ws-factory.ts

@@ -0,0 +1,27 @@
+import { WebSocket } from "ws";
+
+interface WebSocketLike {
+  readyState: number;
+  send: (data: string | Uint8Array | ArrayBuffer) => void;
+  close: (code?: number, reason?: string) => void;
+  binaryType?: string;
+  on?: (event: string, listener: (...args: unknown[]) => void) => void;
+  off?: (event: string, listener: (...args: unknown[]) => void) => void;
+  removeListener?: (event: string, listener: (...args: unknown[]) => void) => void;
+  addEventListener?: (event: string, listener: (event: unknown) => void) => void;
+  removeEventListener?: (event: string, listener: (event: unknown) => void) => void;
+  onopen?: ((event: unknown) => void) | null;
+  onclose?: ((event: unknown) => void) | null;
+  onerror?: ((event: unknown) => void) | null;
+  onmessage?: ((event: unknown) => void) | null;
+}
+
+export type NodeWebSocketFactory = (
+  url: string,
+  options?: { headers?: Record<string, string> },
+) => WebSocketLike;
+
+export function createNodeWebSocketFactory(): NodeWebSocketFactory {
+  return (url: string, options?: { headers?: Record<string, string> }) =>
+    new WebSocket(url, { headers: options?.headers }) as unknown as WebSocketLike;
+}

packages/app/e2e/helpers/paseo-home-fork.ts

@@ -0,0 +1,131 @@
+import { existsSync } from "node:fs";
+import { copyFile, mkdir, readdir, rm, stat } from "node:fs/promises";
+import { homedir } from "node:os";
+import path from "node:path";
+
+export interface PaseoHomeMetadataForkResult {
+  sourceHome: string;
+  targetHome: string;
+  agentFiles: number;
+  agentBytes: number;
+  projectFiles: number;
+  projectBytes: number;
+  copiedFiles: number;
+  copiedBytes: number;
+  skippedMissing: string[];
+}
+
+interface CopyStats {
+  files: number;
+  bytes: number;
+  skippedMissing: string[];
+}
+
+export function resolvePaseoHomePath(value: string): string {
+  if (value === "~") {
+    return homedir();
+  }
+  if (value.startsWith("~/")) {
+    return path.join(homedir(), value.slice(2));
+  }
+  return path.resolve(value);
+}
+
+async function copyJsonTree(sourceDir: string, targetDir: string): Promise<CopyStats> {
+  if (!existsSync(sourceDir)) {
+    return { files: 0, bytes: 0, skippedMissing: [sourceDir] };
+  }
+
+  const stats: CopyStats = { files: 0, bytes: 0, skippedMissing: [] };
+  const entries = await readdir(sourceDir, { withFileTypes: true });
+  await mkdir(targetDir, { recursive: true });
+
+  for (const entry of entries) {
+    const sourcePath = path.join(sourceDir, entry.name);
+    const targetPath = path.join(targetDir, entry.name);
+
+    if (entry.isDirectory()) {
+      const nested = await copyJsonTree(sourcePath, targetPath);
+      stats.files += nested.files;
+      stats.bytes += nested.bytes;
+      stats.skippedMissing.push(...nested.skippedMissing);
+      continue;
+    }
+
+    if (!entry.isFile() || !entry.name.endsWith(".json")) {
+      continue;
+    }
+
+    await mkdir(path.dirname(targetPath), { recursive: true });
+    await copyFile(sourcePath, targetPath);
+    const fileStat = await stat(sourcePath);
+    stats.files += 1;
+    stats.bytes += fileStat.size;
+  }
+
+  return stats;
+}
+
+async function copyProjectRegistryFiles(
+  sourceHome: string,
+  targetHome: string,
+): Promise<CopyStats> {
+  const stats: CopyStats = { files: 0, bytes: 0, skippedMissing: [] };
+  const sourceProjectsDir = path.join(sourceHome, "projects");
+  const targetProjectsDir = path.join(targetHome, "projects");
+  await mkdir(targetProjectsDir, { recursive: true });
+
+  for (const fileName of ["projects.json", "workspaces.json"]) {
+    const sourcePath = path.join(sourceProjectsDir, fileName);
+    const targetPath = path.join(targetProjectsDir, fileName);
+    if (!existsSync(sourcePath)) {
+      stats.skippedMissing.push(sourcePath);
+      continue;
+    }
+    await copyFile(sourcePath, targetPath);
+    const fileStat = await stat(sourcePath);
+    stats.files += 1;
+    stats.bytes += fileStat.size;
+  }
+
+  return stats;
+}
+
+export async function forkPaseoHomeMetadata(input: {
+  sourceHome: string;
+  targetHome: string;
+}): Promise<PaseoHomeMetadataForkResult> {
+  const sourceHome = resolvePaseoHomePath(input.sourceHome);
+  const targetHome = resolvePaseoHomePath(input.targetHome);
+
+  if (sourceHome === targetHome) {
+    throw new Error("Refusing to fork Paseo metadata onto the same PASEO_HOME.");
+  }
+
+  await mkdir(targetHome, { recursive: true });
+
+  // Reset only the copied metadata surface. In particular, do not copy or remove
+  // worktrees here: forked workspace records should continue to point at the
+  // original checkout/worktree paths from the source home.
+  await rm(path.join(targetHome, "agents"), { recursive: true, force: true });
+  await rm(path.join(targetHome, "projects", "projects.json"), { force: true });
+  await rm(path.join(targetHome, "projects", "workspaces.json"), { force: true });
+
+  const agents = await copyJsonTree(
+    path.join(sourceHome, "agents"),
+    path.join(targetHome, "agents"),
+  );
+  const projects = await copyProjectRegistryFiles(sourceHome, targetHome);
+
+  return {
+    sourceHome,
+    targetHome,
+    agentFiles: agents.files,
+    agentBytes: agents.bytes,
+    projectFiles: projects.files,
+    projectBytes: projects.bytes,
+    copiedFiles: agents.files + projects.files,
+    copiedBytes: agents.bytes + projects.bytes,
+    skippedMissing: [...agents.skippedMissing, ...projects.skippedMissing],
+  };
+}

packages/app/e2e/helpers/permissions.ts

@@ -0,0 +1,17 @@
+import { expect, type Page } from "@playwright/test";
+
+export async function waitForPermissionPrompt(page: Page, timeout = 30_000): Promise<void> {
+  await expect(page.getByTestId("permission-request-question").first()).toBeVisible({ timeout });
+}
+
+export async function allowPermission(page: Page): Promise<void> {
+  const acceptButton = page.getByTestId("permission-request-accept").first();
+  await expect(acceptButton).toBeVisible({ timeout: 5_000 });
+  await acceptButton.click();
+}
+
+export async function denyPermission(page: Page): Promise<void> {
+  const denyButton = page.getByTestId("permission-request-deny").first();
+  await expect(denyButton).toBeVisible({ timeout: 5_000 });
+  await denyButton.click();
+}

packages/app/e2e/helpers/pr-pane.ts

@@ -0,0 +1,42 @@
+import { expect, type Page } from "@playwright/test";
+import { getStateLabel } from "@/git/pr-pane-data";
+
+export async function openPrPane(page: Page): Promise<void> {
+  await page.getByRole("button", { name: "Open explorer" }).click();
+  await page.getByTestId("explorer-tab-pr").click();
+  await expect(page.getByTestId("pr-pane")).toBeVisible({ timeout: 15_000 });
+}
+
+export async function expectPrPaneTitle(page: Page, title: string): Promise<void> {
+  await expect(page.getByTestId("pr-pane-title")).toContainText(title, { timeout: 15_000 });
+}
+
+export async function expectPrPaneState(
+  page: Page,
+  state: "open" | "merged" | "closed" | "draft",
+): Promise<void> {
+  await expect(page.getByTestId("pr-pane-state")).toHaveText(getStateLabel(state), {
+    timeout: 15_000,
+  });
+}
+
+async function assertCheckPill(page: Page, testId: string, count: number): Promise<void> {
+  const locator = page.getByTestId(testId);
+  await expect(locator).toHaveCount(count > 0 ? 1 : 0, { timeout: 15_000 });
+  if (count > 0) {
+    await expect(locator).toContainText(String(count));
+  }
+}
+
+export async function expectPrPaneCheckSummary(
+  page: Page,
+  counts: { passed: number; failed: number; pending: number },
+): Promise<void> {
+  await assertCheckPill(page, "pr-pane-check-passed", counts.passed);
+  await assertCheckPill(page, "pr-pane-check-failed", counts.failed);
+  await assertCheckPill(page, "pr-pane-check-pending", counts.pending);
+}
+
+export async function expectPrPaneActivityCount(page: Page, count: number): Promise<void> {
+  await expect(page.getByTestId("pr-pane-activity-row")).toHaveCount(count, { timeout: 15_000 });
+}

packages/app/e2e/helpers/project-settings.ts

@@ -0,0 +1,269 @@
+import { chmod, readFile, writeFile } from "node:fs/promises";
+import path from "node:path";
+import { expect, type Page } from "@playwright/test";
+import type { WebSocketRoute } from "@playwright/test";
+import { gotoAppShell, openSettings } from "./app";
+
+// --- Navigation ---
+
+export async function openProjects(page: Page): Promise<void> {
+  await gotoAppShell(page);
+  await openSettings(page);
+  await page.getByTestId("settings-projects").click();
+  await expect(page).toHaveURL(/\/settings\/projects$/);
+}
+
+export async function openProjectSettings(page: Page, projectName: string): Promise<void> {
+  await page.getByRole("button", { name: `Edit ${projectName}`, exact: true }).click();
+  await expect(page.getByRole("textbox", { name: "Worktree setup commands" })).toBeVisible({
+    timeout: 30_000,
+  });
+}
+
+export async function navigateToProjectSettings(page: Page, projectName: string): Promise<void> {
+  await page.getByRole("button", { name: `Edit ${projectName}`, exact: true }).click();
+}
+
+// --- Form interactions ---
+
+export async function editWorktreeSetup(page: Page, setupCommands: string[]): Promise<void> {
+  await page
+    .getByRole("textbox", { name: "Worktree setup commands" })
+    .fill(setupCommands.join("\n"));
+}
+
+export async function clickSaveProjectSettings(page: Page): Promise<void> {
+  await page.getByRole("button", { name: "Save project config" }).click();
+}
+
+export async function clickRetryProjectSettingsSave(page: Page): Promise<void> {
+  // action-0 is always "Try again"; action-1 is always "Reload".
+  // The write-failed callout renders these two buttons in a fixed order.
+  await page.getByTestId("write-failed-callout-action-0").click();
+}
+
+export async function clickReloadProjectSettings(page: Page): Promise<void> {
+  // Scope to the active error callout so the locator is unambiguous.
+  // At most one error callout renders at a time.
+  await page.locator('[data-testid$="-callout"]').getByRole("button", { name: "Reload" }).click();
+}
+
+// --- Error-state assertions ---
+
+type ErrorKind = "stale" | "invalid" | "write_failed" | "transport" | "read_failed";
+
+const errorCalloutTestId: Record<ErrorKind, string> = {
+  stale: "stale-callout",
+  invalid: "invalid-callout",
+  write_failed: "write-failed-callout",
+  transport: "read-transport-callout",
+  read_failed: "read-failed-callout",
+};
+
+export async function expectProjectSettingsError(page: Page, kind: ErrorKind): Promise<void> {
+  await expect(page.getByTestId(errorCalloutTestId[kind])).toBeVisible({ timeout: 15_000 });
+}
+
+export async function expectNoProjectSettingsError(
+  page: Page,
+  kind: ErrorKind,
+  timeout = 15_000,
+): Promise<void> {
+  await expect(page.getByTestId(errorCalloutTestId[kind])).not.toBeVisible({ timeout });
+}
+
+export async function expectWriteFailedCalloutActions(page: Page): Promise<void> {
+  await expect(page.getByTestId("write-failed-callout-action-0")).toHaveText("Try again");
+  await expect(page.getByTestId("write-failed-callout-action-1")).toHaveText("Reload");
+}
+
+export async function expectSaveButtonDisabled(page: Page): Promise<void> {
+  await expect(page.getByRole("button", { name: "Save project config" })).toBeDisabled();
+}
+
+// --- Form-state assertions ---
+
+export async function expectProjectSettingsFormVisible(page: Page): Promise<void> {
+  await expect(page.getByRole("textbox", { name: "Worktree setup commands" })).toBeVisible({
+    timeout: 15_000,
+  });
+}
+
+export async function expectProjectSettingsFormHidden(page: Page): Promise<void> {
+  await expect(page.getByRole("textbox", { name: "Worktree setup commands" })).not.toBeVisible();
+}
+
+export async function expectNoEditableTarget(page: Page): Promise<void> {
+  await expect(page.getByTestId("project-settings-back-button")).toBeVisible({ timeout: 30_000 });
+}
+
+// --- Host-section assertions ---
+
+export async function expectHostIndicatorVisible(page: Page): Promise<void> {
+  await expect(page.getByTestId("host-indicator")).toBeVisible();
+}
+
+export async function expectHostPickerHidden(page: Page): Promise<void> {
+  await expect(page.getByTestId("host-picker")).not.toBeVisible();
+}
+
+// --- Script-list assertions and interactions ---
+
+// Counts only row Views, not kebab-trigger elements (which share the "script-row-"
+// prefix but contain "-menu-").
+export async function expectScriptRowCount(page: Page, count: number): Promise<void> {
+  await expect(
+    page
+      .getByTestId("scripts-list")
+      .locator('[data-testid^="script-row-"]:not([data-testid*="-menu-"])'),
+  ).toHaveCount(count);
+}
+
+export async function expectEmptyScriptList(page: Page): Promise<void> {
+  await expect(page.getByText("No scripts yet.")).toBeVisible();
+}
+
+export async function removeProjectScript(page: Page, scriptName: string): Promise<void> {
+  const row = page
+    .getByTestId("scripts-list")
+    .locator('[data-testid^="script-row-"]:not([data-testid*="-menu-"])')
+    .filter({ hasText: scriptName })
+    .first();
+  // DropdownMenuTrigger renders as a Pressable (no role="button"); derive its testID
+  // from the row's testID to avoid scoped locator unreliability.
+  const id = (await row.getAttribute("data-testid"))!.replace("script-row-", "");
+  await page.getByTestId(`script-row-menu-${id}`).click();
+  page.once("dialog", (dialog) => void dialog.accept());
+  await page.getByRole("button", { name: "Remove" }).click();
+}
+
+// --- File manipulation ---
+
+export async function corruptPaseoConfig(repoPath: string): Promise<void> {
+  await writeFile(path.join(repoPath, "paseo.json"), "{not valid json}");
+}
+
+export async function bumpPaseoConfigOnDisk(repoPath: string): Promise<void> {
+  const configPath = path.join(repoPath, "paseo.json");
+  const raw = await readFile(configPath, "utf8");
+  const config = JSON.parse(raw) as Record<string, unknown>;
+  config._bump = Date.now();
+  await writeFile(configPath, JSON.stringify(config, null, 2) + "\n");
+}
+
+export async function restorePaseoConfig(
+  repoPath: string,
+  config: Record<string, unknown>,
+): Promise<void> {
+  await writeFile(path.join(repoPath, "paseo.json"), JSON.stringify(config, null, 2) + "\n");
+}
+
+// The daemon writes atomically via a temp file + rename, so blocking writes requires
+// removing write permission from the *directory*, not just the file.
+export async function blockPaseoConfigWrites(repoPath: string): Promise<void> {
+  await chmod(repoPath, 0o555);
+}
+
+export async function unblockPaseoConfigWrites(repoPath: string): Promise<void> {
+  await chmod(repoPath, 0o755);
+}
+
+// --- WebSocket helpers ---
+
+function buildDaemonPortPattern(): RegExp {
+  const port = process.env.E2E_DAEMON_PORT;
+  if (!port) throw new Error("E2E_DAEMON_PORT not set — globalSetup must run first");
+  return new RegExp(`:${port.replace(/[.*+?^${}()|[\]\\]/g, "\\$&")}\\b`);
+}
+
+// Proxies all daemon WS traffic transparently until a read_project_config_request
+// is seen, then closes that connection (triggering readQuery.isError). Subsequent
+// connections pass through so the Reload action can succeed.
+export async function installReadTransportFailure(page: Page): Promise<void> {
+  let armed = true;
+
+  await page.routeWebSocket(buildDaemonPortPattern(), (ws) => {
+    const server = ws.connectToServer();
+
+    ws.onMessage((message) => {
+      if (armed && typeof message === "string") {
+        try {
+          const envelope = JSON.parse(message) as {
+            type?: string;
+            message?: { type?: string };
+          };
+          if (
+            envelope.type === "session" &&
+            envelope.message?.type === "read_project_config_request"
+          ) {
+            armed = false;
+            void ws.close({ code: 1001 });
+            return;
+          }
+        } catch {
+          // binary or malformed frame — pass through
+        }
+      }
+      try {
+        server.send(message);
+      } catch {
+        // server socket already closed
+      }
+    });
+
+    server.onMessage((message) => {
+      try {
+        ws.send(message);
+      } catch {
+        // client socket already closed
+      }
+    });
+  });
+}
+
+// Installs a transparent WS proxy that can later drop all active daemon connections
+// and block new ones. Code 1001 (Going Away) without reason triggers "error" state
+// in DaemonClient due to describeTransportClose returning a non-empty string.
+export async function installDaemonConnectionGate(
+  page: Page,
+): Promise<{ drop: () => Promise<void> }> {
+  let acceptingConnections = true;
+  const activeSockets = new Set<WebSocketRoute>();
+
+  await page.routeWebSocket(buildDaemonPortPattern(), (ws) => {
+    if (!acceptingConnections) {
+      void ws.close({ code: 1001 });
+      return;
+    }
+
+    activeSockets.add(ws);
+    const server = ws.connectToServer();
+
+    ws.onMessage((message) => {
+      if (!acceptingConnections) return;
+      try {
+        server.send(message);
+      } catch {
+        activeSockets.delete(ws);
+      }
+    });
+
+    server.onMessage((message) => {
+      if (!acceptingConnections) return;
+      try {
+        ws.send(message);
+      } catch {
+        activeSockets.delete(ws);
+      }
+    });
+  });
+
+  return {
+    async drop(): Promise<void> {
+      acceptingConnections = false;
+      const sockets = Array.from(activeSockets);
+      activeSockets.clear();
+      await Promise.all(sockets.map((ws) => ws.close({ code: 1001 }).catch(() => undefined)));
+    },
+  };
+}

packages/app/e2e/helpers/settings.ts

@@ -0,0 +1,273 @@
+import { expect, type Page } from "@playwright/test";
+import { requireServerId } from "./sidebar";
+
+function escapeRegex(value: string): string {
+  return value.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
+}
+
+const SECTION_LABELS = {
+  general: "General",
+  shortcuts: "Shortcuts",
+  integrations: "Integrations",
+  permissions: "Permissions",
+  diagnostics: "Diagnostics",
+  about: "About",
+} as const;
+
+export type SettingsSection = keyof typeof SECTION_LABELS | "projects";
+
+export async function openSettingsSection(page: Page, section: SettingsSection): Promise<void> {
+  const sidebar = page.getByTestId("settings-sidebar");
+  await expect(sidebar).toBeVisible();
+
+  if (section === "projects") {
+    await page.getByTestId("settings-projects").click();
+    await expect(page).toHaveURL(/\/settings\/projects$/);
+    return;
+  }
+
+  await sidebar.getByRole("button", { name: SECTION_LABELS[section], exact: true }).click();
+  await expect(page).toHaveURL(new RegExp(`/settings/${section}$`));
+}
+
+export async function openSettingsHost(page: Page, serverId: string): Promise<void> {
+  await page.getByTestId(`settings-host-entry-${serverId}`).click();
+  await expect(page.getByTestId(`settings-host-page-${serverId}`)).toBeVisible();
+}
+
+export async function expectSettingsHeader(page: Page, title: string): Promise<void> {
+  await expect(page.getByTestId("settings-detail-header-title")).toHaveText(title);
+}
+
+export async function openAddHostFlow(page: Page): Promise<void> {
+  await page.getByTestId("settings-add-host").click();
+  await expect(page.getByText("Add connection", { exact: true })).toBeVisible();
+}
+
+export async function selectHostConnectionType(
+  page: Page,
+  type: "direct" | "relay",
+): Promise<void> {
+  const label = type === "direct" ? "Direct connection" : "Paste pairing link";
+  await page.getByRole("button", { name: label }).click();
+}
+
+export async function toggleHostAdvanced(page: Page): Promise<void> {
+  await page.getByTestId("direct-host-advanced-toggle").click();
+}
+
+export async function openCompactSettings(page: Page): Promise<void> {
+  await expect(page).toHaveURL(/\/h\/|\/welcome/, { timeout: 15000 });
+  await page.getByRole("button", { name: "Open menu", exact: true }).first().click();
+  const settingsButton = page.locator('[data-testid="sidebar-settings"]:visible').first();
+  await expect(settingsButton).toBeVisible();
+  await settingsButton.click();
+  await expect(page).toHaveURL(/\/settings$/);
+  await expect(page.getByTestId("settings-sidebar")).toBeVisible();
+}
+
+export async function expectCompactSettingsList(page: Page): Promise<void> {
+  await expect(page).toHaveURL(/\/settings$/);
+  await expect(page.getByTestId("settings-sidebar")).toBeVisible();
+  await expect(page.getByText("Theme", { exact: true })).toHaveCount(0);
+  await expect(page.getByRole("button", { name: "Play test" })).toHaveCount(0);
+  await expect(page.locator('[data-testid^="settings-host-page-"]')).toHaveCount(0);
+}
+
+export async function expectSettingsSidebarVisible(page: Page): Promise<void> {
+  await expect(page.getByTestId("settings-sidebar")).toBeVisible();
+}
+
+export async function expectSettingsSidebarHidden(page: Page): Promise<void> {
+  await expect(page.locator('[data-testid="settings-sidebar"]:visible')).toHaveCount(0);
+}
+
+export async function expectSettingsSidebarSections(
+  page: Page,
+  sections: Array<Exclude<SettingsSection, "projects">>,
+): Promise<void> {
+  const sidebar = page.getByTestId("settings-sidebar");
+  for (const section of sections) {
+    await expect(
+      sidebar.getByRole("button", { name: SECTION_LABELS[section], exact: true }),
+    ).toBeVisible();
+  }
+}
+
+export async function goBackInSettings(page: Page): Promise<void> {
+  await page.getByRole("button", { name: "Back", exact: true }).click();
+}
+
+export async function expectSettingsBackButton(page: Page): Promise<void> {
+  await expect(page.getByRole("button", { name: "Back", exact: true })).toBeVisible();
+}
+
+export async function clickSettingsBackToWorkspace(page: Page): Promise<void> {
+  await page.getByTestId("settings-back-to-workspace").click();
+}
+
+export async function expectHostSettingsUrl(page: Page, serverId: string): Promise<void> {
+  await expect(page).toHaveURL(
+    new RegExp(`/settings/hosts/${escapeRegex(encodeURIComponent(serverId))}$`),
+  );
+}
+
+export async function verifyLegacyHostSettingsRedirect(page: Page): Promise<void> {
+  const serverId = requireServerId();
+  await page.goto(`/h/${encodeURIComponent(serverId)}/settings`);
+  await expectHostSettingsUrl(page, serverId);
+}
+
+export async function openCompactSettingsHost(page: Page): Promise<void> {
+  const serverId = requireServerId();
+  await openSettingsHost(page, serverId);
+  await expectHostSettingsUrl(page, serverId);
+}
+
+export async function expectAddHostMethodOptions(page: Page): Promise<void> {
+  await expect(page.getByRole("button", { name: "Direct connection" })).toBeVisible();
+  await expect(page.getByRole("button", { name: "Paste pairing link" })).toBeVisible();
+}
+
+export async function fillDirectHostUri(page: Page, uri: string): Promise<void> {
+  await page.getByTestId("direct-host-uri-input").fill(uri);
+}
+
+export async function expectDirectHostFormValues(
+  page: Page,
+  fields: { host: string; port: string; password: string },
+): Promise<void> {
+  await expect(page.getByTestId("direct-host-input")).toHaveValue(fields.host);
+  await expect(page.getByTestId("direct-port-input")).toHaveValue(fields.port);
+  await expect(page.getByTestId("direct-password-input")).toHaveValue(fields.password);
+}
+
+export async function expectDirectHostSslEnabled(page: Page): Promise<void> {
+  await expect(page.getByTestId("direct-ssl-toggle-checked")).toBeVisible();
+}
+
+export async function expectDirectHostUriValue(page: Page, uri: string): Promise<void> {
+  await expect(page.getByTestId("direct-host-uri-input")).toHaveValue(uri);
+}
+
+export async function expectDirectHostUriHidden(page: Page): Promise<void> {
+  await expect(page.getByTestId("direct-host-uri-input")).toHaveCount(0);
+}
+
+export async function expectDiagnosticsContent(page: Page): Promise<void> {
+  await expect(page.getByRole("button", { name: "Play test" })).toBeVisible();
+}
+
+export async function expectAboutContent(page: Page): Promise<void> {
+  await expect(page.getByText("Version", { exact: true }).first()).toBeVisible();
+}
+
+export async function expectGeneralContent(page: Page): Promise<void> {
+  await expect(page.getByText("Theme", { exact: true }).first()).toBeVisible();
+}
+
+export async function expectHostLabelDisplayed(page: Page): Promise<void> {
+  await expect(page.getByTestId("host-page-label-edit-button")).toBeVisible();
+  await expect(page.getByTestId("host-page-label-input")).toHaveCount(0);
+}
+
+export async function clickEditHostLabel(page: Page): Promise<void> {
+  await page.getByTestId("host-page-label-edit-button").click();
+}
+
+export async function expectHostLabelEditMode(page: Page, expectedLabel: string): Promise<void> {
+  await expect(page.getByTestId("host-page-label-input")).toBeVisible();
+  await expect(page.getByTestId("host-page-label-input")).toHaveValue(expectedLabel);
+  await expect(page.getByTestId("host-page-label-save")).toBeVisible();
+}
+
+export async function expectHostConnectionsCard(page: Page, port: string): Promise<void> {
+  const card = page.getByTestId("host-page-connections-card");
+  await expect(card).toBeVisible();
+  await expect(page.getByText("Connections", { exact: true })).toBeVisible();
+  await expect(
+    card.getByText(new RegExp(`TCP \\((localhost|127\\.0\\.0\\.1):${port}\\)`)),
+  ).toBeVisible();
+}
+
+export async function expectHostInjectMcpCard(page: Page): Promise<void> {
+  const card = page.getByTestId("host-page-inject-mcp-card");
+  await expect(card).toBeVisible();
+  await expect(card.getByRole("switch", { name: "Inject Paseo tools" })).toBeVisible();
+}
+
+export async function expectHostActionCards(page: Page): Promise<void> {
+  await expect(page.getByTestId("host-page-restart-card")).toBeVisible();
+  await expect(page.getByTestId("host-page-restart-button")).toBeVisible();
+  await expect(page.getByTestId("host-page-providers-card")).toBeVisible();
+  await expect(page.getByTestId("host-page-remove-host-card")).toBeVisible();
+  await expect(page.getByTestId("host-page-remove-host-button")).toBeVisible();
+}
+
+export async function serveJson(page: Page, url: string, body: unknown): Promise<void> {
+  await page.route(url, async (route) => {
+    await route.fulfill({
+      status: 200,
+      contentType: "application/json",
+      body: JSON.stringify(body),
+    });
+  });
+}
+
+export async function openAddProviderModal(page: Page): Promise<void> {
+  await page.getByRole("button", { name: "Add provider", exact: true }).click();
+  await expect(page.getByRole("textbox", { name: "Search providers" })).toBeVisible();
+}
+
+export async function findAcpCatalogProvider(page: Page, providerName: string): Promise<void> {
+  await page.getByRole("textbox", { name: "Search providers" }).fill(providerName);
+  await expect(page.getByText(providerName, { exact: true })).toBeVisible();
+}
+
+export async function installAcpCatalogProvider(page: Page, providerName: string): Promise<void> {
+  await findAcpCatalogProvider(page, providerName);
+  await page.getByRole("button", { name: "Add", exact: true }).click();
+  await expect(page.getByRole("textbox", { name: "Search providers" })).toHaveCount(0);
+}
+
+export async function expectProviderInstalledInSettings(
+  page: Page,
+  providerName: string,
+): Promise<void> {
+  await expect(
+    page.getByRole("button", { name: `${providerName} provider details`, exact: true }),
+  ).toBeVisible();
+}
+
+export async function expectHostNoLocalOnlyRows(page: Page): Promise<void> {
+  await expect(page.getByTestId("host-page-pair-device-row")).toHaveCount(0);
+  await expect(page.getByTestId("host-page-daemon-lifecycle-card")).toHaveCount(0);
+}
+
+export async function expectRetiredSidebarSectionsAbsent(page: Page): Promise<void> {
+  const sidebar = page.getByTestId("settings-sidebar");
+  await expect(sidebar).toBeVisible();
+  await expect(sidebar.getByRole("button", { name: "Hosts", exact: true })).toHaveCount(0);
+  await expect(sidebar.getByRole("button", { name: "Providers", exact: true })).toHaveCount(0);
+  await expect(sidebar.getByRole("button", { name: "Pair device", exact: true })).toHaveCount(0);
+  await expect(sidebar.getByRole("button", { name: "Daemon", exact: true })).toHaveCount(0);
+  await expect(sidebar.getByRole("button", { name: "General", exact: true })).toBeVisible();
+  await expect(sidebar.getByRole("button", { name: "Diagnostics", exact: true })).toBeVisible();
+  await expect(sidebar.getByRole("button", { name: "About", exact: true })).toBeVisible();
+}
+
+export async function expectHostPageVisible(page: Page, serverId: string): Promise<void> {
+  await expect(page.getByTestId(`settings-host-page-${serverId}`)).toBeVisible();
+}
+
+export async function expectLocalHostEntryFirst(page: Page, serverId: string): Promise<void> {
+  const sidebar = page.getByTestId("settings-sidebar");
+  await expect(sidebar).toBeVisible({ timeout: 15_000 });
+  await expect(sidebar.locator('[data-testid^="settings-host-entry-"]').first()).toHaveAttribute(
+    "data-testid",
+    `settings-host-entry-${serverId}`,
+  );
+  const localHostEntry = page.getByTestId(`settings-host-entry-${serverId}`);
+  await expect(localHostEntry.getByTestId("settings-host-local-marker")).toBeVisible();
+  await expect(localHostEntry.getByText("Local", { exact: true })).toBeVisible();
+}

packages/app/e2e/helpers/sidebar.ts

@@ -0,0 +1,39 @@
+import { expect, type Page } from "@playwright/test";
+
+export function requireServerId(): string {
+  const serverId = process.env.E2E_SERVER_ID;
+  if (!serverId) {
+    throw new Error("E2E_SERVER_ID is not set (expected from Playwright globalSetup).");
+  }
+  return serverId;
+}
+
+export async function selectWorkspaceInSidebar(page: Page, workspaceId: string): Promise<void> {
+  const row = page.getByTestId(`sidebar-workspace-row-${requireServerId()}:${workspaceId}`);
+  await expect(row).toBeVisible({ timeout: 30_000 });
+  await row.click();
+}
+
+export async function expectWorkspaceListed(page: Page, name: string): Promise<void> {
+  await expect(
+    page.locator('[data-testid^="sidebar-workspace-row-"]').filter({ hasText: name }).first(),
+  ).toBeVisible({ timeout: 30_000 });
+}
+
+export async function openMobileAgentSidebar(page: Page): Promise<void> {
+  await page.getByTestId("menu-button").click();
+}
+
+// force=true: the overlay covers the button when the mobile sidebar is open.
+export async function closeMobileAgentSidebar(page: Page): Promise<void> {
+  await page.getByTestId("menu-button").click({ force: true });
+}
+
+// The mobile sidebar panel animates via translateX; toBeInViewport reflects the rendered position.
+export async function expectMobileAgentSidebarVisible(page: Page): Promise<void> {
+  await expect(page.getByTestId("sidebar-sessions")).toBeInViewport({ timeout: 5_000 });
+}
+
+export async function expectMobileAgentSidebarHidden(page: Page): Promise<void> {
+  await expect(page.getByTestId("sidebar-sessions")).not.toBeInViewport({ timeout: 5_000 });
+}

packages/app/e2e/helpers/startup-dsl.ts

@@ -0,0 +1,233 @@
+import { expect, type Page } from "../fixtures";
+import { buildCreateAgentPreferences, buildSeededHost } from "./daemon-registry";
+
+const DISABLE_DEFAULT_SEED_ONCE_KEY = "@paseo:e2e-disable-default-seed-once";
+const SEED_NONCE_KEY = "@paseo:e2e-seed-nonce";
+const REGISTRY_KEY = "@paseo:daemon-registry";
+const E2E_KEY = "@paseo:e2e";
+
+interface SavedHostInput {
+  serverId: string;
+  label: string;
+  endpoint: string;
+}
+
+export function startupScenario(page: Page) {
+  return new StartupScenario(page);
+}
+
+class StartupScenario {
+  private readonly page: Page;
+  private savedHosts: SavedHostInput[] = [];
+  private desktopBridge = false;
+  private blockedEndpointPorts = new Set<string>();
+  private viewport: { width: number; height: number } | null = null;
+
+  constructor(page: Page) {
+    this.page = page;
+  }
+
+  withMobileViewport(): this {
+    this.viewport = { width: 390, height: 844 };
+    return this;
+  }
+
+  withSavedHost(input: SavedHostInput): this {
+    this.savedHosts.push(input);
+    const port = input.endpoint.match(/:(\d+)$/)?.[1];
+    if (port) {
+      this.blockedEndpointPorts.add(port);
+    }
+    return this;
+  }
+
+  withPendingDesktopDaemon(): this {
+    this.desktopBridge = true;
+    return this;
+  }
+
+  withBlockedPort(port: string): this {
+    this.blockedEndpointPorts.add(port);
+    return this;
+  }
+
+  async openRoot(): Promise<StartupAssertions> {
+    await this.prepare();
+    await this.page.goto("/");
+    return new StartupAssertions(this.page);
+  }
+
+  async openHostWorkspace(input: {
+    serverId: string;
+    workspaceId: string;
+  }): Promise<StartupAssertions> {
+    await this.prepare();
+    await this.page.goto(
+      `/h/${encodeURIComponent(input.serverId)}/workspace/${encodeURIComponent(input.workspaceId)}`,
+    );
+    return new StartupAssertions(this.page);
+  }
+
+  private async prepare(): Promise<void> {
+    if (this.viewport) {
+      await this.page.setViewportSize(this.viewport);
+    }
+
+    if (this.desktopBridge) {
+      await installPendingDesktopBridge(this.page);
+    }
+
+    for (const port of this.blockedEndpointPorts) {
+      await this.page.routeWebSocket(new RegExp(`:${escapeRegex(port)}\\b`), async (ws) => {
+        await ws.close({ code: 1008, reason: "Blocked unreachable startup test host." });
+      });
+    }
+
+    if (this.savedHosts.length === 0) {
+      return;
+    }
+
+    // Let the shared fixture create its seed nonce, then opt out of that seed for
+    // the next navigation so this scenario owns the stored host registry.
+    await this.page.goto("/");
+    const nowIso = new Date().toISOString();
+    const registry = this.savedHosts.map((host) =>
+      buildStoredHost({
+        serverId: host.serverId,
+        endpoint: host.endpoint,
+        label: host.label,
+        nowIso,
+      }),
+    );
+    const firstHost = registry[0];
+    if (!firstHost) {
+      throw new Error("Expected at least one startup test host.");
+    }
+    const createAgentPreferences = buildStoredCreateAgentPreferences(firstHost.serverId);
+
+    await this.page.evaluate(
+      ({ keys, registry: storedRegistry, createAgentPreferences: storedPreferences }) => {
+        const nonce = localStorage.getItem(keys.seedNonce);
+        if (!nonce) {
+          throw new Error("Expected e2e seed nonce before overriding startup registry.");
+        }
+
+        localStorage.setItem(keys.e2e, "1");
+        localStorage.setItem(keys.registry, JSON.stringify(storedRegistry));
+        localStorage.setItem("@paseo:create-agent-preferences", JSON.stringify(storedPreferences));
+        localStorage.setItem(keys.disableDefaultSeedOnce, nonce);
+      },
+      {
+        keys: {
+          disableDefaultSeedOnce: DISABLE_DEFAULT_SEED_ONCE_KEY,
+          e2e: E2E_KEY,
+          registry: REGISTRY_KEY,
+          seedNonce: SEED_NONCE_KEY,
+        },
+        registry,
+        createAgentPreferences,
+      },
+    );
+  }
+}
+
+class StartupAssertions {
+  private readonly page: Page;
+
+  constructor(page: Page) {
+    this.page = page;
+  }
+
+  async expectsReconnectWelcome(): Promise<this> {
+    await expect(this.page.getByTestId("welcome-screen")).toBeVisible({ timeout: 15_000 });
+    await expect(this.page.getByTestId("welcome-open-settings")).toBeVisible();
+    await expect(this.page.getByTestId("welcome-direct-connection")).toBeVisible();
+    await expect(this.page.getByTestId("welcome-paste-pairing-link")).toBeVisible();
+    await expect(this.page.getByTestId("welcome-scan-qr")).toHaveCount(0);
+    return this;
+  }
+
+  async expectsNoSavedHostStatus(input: { label: string }): Promise<this> {
+    await expect(this.page.getByText(input.label, { exact: true })).toHaveCount(0);
+    await expect(this.page.getByText("Connection error", { exact: true })).toHaveCount(0);
+    await expect(this.page.getByText("Offline", { exact: true })).toHaveCount(0);
+    return this;
+  }
+
+  async expectsNoLocalServerStartupCopy(): Promise<this> {
+    await expect(this.page.getByText("Starting local server...", { exact: true })).toHaveCount(0);
+    await expect(this.page.getByText("Connecting to local server...", { exact: true })).toHaveCount(
+      0,
+    );
+    return this;
+  }
+
+  async expectsDesktopDaemonStartup(): Promise<this> {
+    await expect(this.page.getByTestId("startup-splash")).toBeVisible({
+      timeout: 15_000,
+    });
+    return this;
+  }
+
+  async expectsSidebarHidden(): Promise<this> {
+    await expect(this.page.locator('[data-testid="sidebar-settings"]:visible')).toHaveCount(0);
+    await expect(this.page.locator('[data-testid="sidebar-project-list"]:visible')).toHaveCount(0);
+    return this;
+  }
+
+  async expectsNoUndefinedRoute(): Promise<this> {
+    await expect(this.page).not.toHaveURL(/\/h\/undefined\/workspace\/undefined/);
+    return this;
+  }
+}
+
+async function installPendingDesktopBridge(page: Page): Promise<void> {
+  await page.addInitScript(() => {
+    (window as unknown as { paseoDesktop: unknown }).paseoDesktop = {
+      platform: "darwin",
+      invoke: async (command: string) => {
+        if (command === "start_desktop_daemon") {
+          await new Promise(() => {
+            // Keep the daemon in the startup phase until the test ends.
+          });
+        }
+        if (command === "desktop_daemon_status") {
+          return {
+            serverId: "srv_desktop_pending",
+            status: "starting",
+            listen: null,
+            hostname: null,
+            pid: null,
+            home: "",
+            version: null,
+            desktopManaged: true,
+            error: null,
+          };
+        }
+        if (command === "desktop_daemon_logs") {
+          return { logPath: "", contents: "" };
+        }
+        return null;
+      },
+      getPendingOpenProject: async () => null,
+      events: { on: async () => () => undefined },
+    };
+  });
+}
+
+function buildStoredHost(input: {
+  serverId: string;
+  endpoint: string;
+  label: string;
+  nowIso: string;
+}) {
+  return buildSeededHost(input);
+}
+
+function buildStoredCreateAgentPreferences(serverId: string) {
+  return buildCreateAgentPreferences(serverId);
+}
+
+function escapeRegex(value: string): string {
+  return value.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
+}

packages/app/e2e/helpers/terminal-dsl.ts

@@ -0,0 +1,98 @@
+import type { Page } from "@playwright/test";
+import { createTempGitRepo } from "./workspace";
+import {
+  connectTerminalClient,
+  navigateToTerminal,
+  setupDeterministicPrompt,
+  type TerminalPerfDaemonClient,
+} from "./terminal-perf";
+
+interface TempRepo {
+  path: string;
+  cleanup: () => Promise<void>;
+}
+
+export interface TerminalInstance {
+  id: string;
+  name: string;
+  cwd: string;
+}
+
+export class TerminalE2EHarness {
+  readonly client: TerminalPerfDaemonClient;
+  readonly tempRepo: TempRepo;
+  readonly workspaceId: string;
+
+  private constructor(input: {
+    client: TerminalPerfDaemonClient;
+    tempRepo: TempRepo;
+    workspaceId: string;
+  }) {
+    this.client = input.client;
+    this.tempRepo = input.tempRepo;
+    this.workspaceId = input.workspaceId;
+  }
+
+  static async create(input: { tempPrefix: string }): Promise<TerminalE2EHarness> {
+    const tempRepo = await createTempGitRepo(input.tempPrefix);
+    const client = await connectTerminalClient();
+    const seedResult = await client.openProject(tempRepo.path);
+    if (!seedResult.workspace) {
+      await client.close().catch(() => {});
+      await tempRepo.cleanup().catch(() => {});
+      throw new Error(seedResult.error ?? "Failed to seed workspace");
+    }
+    return new TerminalE2EHarness({
+      client,
+      tempRepo,
+      workspaceId: seedResult.workspace.id,
+    });
+  }
+
+  async cleanup(): Promise<void> {
+    await this.client.close().catch(() => {});
+    await this.tempRepo.cleanup().catch(() => {});
+  }
+
+  async createTerminal(input: { name: string }): Promise<TerminalInstance> {
+    const result = await this.client.createTerminal(this.tempRepo.path, input.name);
+    if (!result.terminal) {
+      throw new Error(`Failed to create terminal: ${result.error}`);
+    }
+    return result.terminal;
+  }
+
+  async killTerminal(terminalId: string): Promise<void> {
+    await this.client.killTerminal(terminalId).catch(() => {});
+  }
+
+  async openTerminal(page: Page, input: { terminalId: string }): Promise<void> {
+    await navigateToTerminal(page, {
+      workspaceId: this.workspaceId,
+      terminalId: input.terminalId,
+    });
+  }
+
+  terminalSurface(page: Page) {
+    return page.locator('[data-testid="terminal-surface"]');
+  }
+
+  async setupPrompt(page: Page, sentinel?: string): Promise<void> {
+    await setupDeterministicPrompt(page, sentinel);
+  }
+}
+
+export async function withTerminalInApp<T>(
+  page: Page,
+  harness: TerminalE2EHarness,
+  input: { name: string },
+  fn: (terminal: TerminalInstance) => Promise<T>,
+): Promise<T> {
+  const terminal = await harness.createTerminal({ name: input.name });
+  try {
+    await harness.openTerminal(page, { terminalId: terminal.id });
+    return await fn(terminal);
+  } finally {
+    await harness.killTerminal(terminal.id);
+  }
+}

packages/app/e2e/helpers/terminal-perf.ts

@@ -1,12 +1,17 @@
-import type { Page } from "@playwright/test";
+import { expect, type Page } from "@playwright/test";
 import path from "node:path";
 import { pathToFileURL } from "node:url";
 import { randomUUID } from "node:crypto";
+import { createNodeWebSocketFactory, type NodeWebSocketFactory } from "./node-ws-factory";
 import { buildHostWorkspaceRoute } from "../../src/utils/host-routes";
 
-export type TerminalPerfDaemonClient = {
+export interface TerminalPerfDaemonClient {
   connect(): Promise<void>;
   close(): Promise<void>;
+  openProject(cwd: string): Promise<{
+    workspace: { id: string; name: string; projectRootPath: string } | null;
+    error: string | null;
+  }>;
   createTerminal(
     cwd: string,
     name?: string,
@@ -14,6 +19,17 @@ export type TerminalPerfDaemonClient = {
     terminal: { id: string; name: string; cwd: string } | null;
     error: string | null;
   }>;
+  createAgent(options: {
+    provider: string;
+    cwd: string;
+    title?: string;
+    modeId?: string;
+    model?: string;
+    thinkingOptionId?: string;
+    featureValues?: Record<string, unknown>;
+    initialPrompt?: string;
+  }): Promise<{ id: string; status: string }>;
+  sendAgentMessage(agentId: string, text: string): Promise<void>;
   subscribeTerminal(
     terminalId: string,
   ): Promise<{ terminalId: string; slot: number; error: null } | { error: string }>;
@@ -25,7 +41,7 @@ export type TerminalPerfDaemonClient = {
     handler: (event: { terminalId: string; type: string; data?: Uint8Array }) => void,
   ): () => void;
   killTerminal(terminalId: string): Promise<{ error: string | null }>;
-};
+}
 
 function getDaemonWsUrl(): string {
   const daemonPort = process.env.E2E_DAEMON_PORT;
@@ -43,43 +59,64 @@ function getServerId(): string {
   return serverId;
 }
 
+interface TerminalPerfDaemonClientConfig {
+  url: string;
+  clientId: string;
+  clientType: "cli";
+  webSocketFactory?: NodeWebSocketFactory;
+}
+
 async function loadDaemonClientConstructor(): Promise<
-  new (config: { url: string; clientId: string; clientType: "cli" }) => TerminalPerfDaemonClient
+  new (config: TerminalPerfDaemonClientConfig) => TerminalPerfDaemonClient
 > {
-  const repoRoot = path.resolve(process.cwd(), "../..");
+  const repoRoot = path.resolve(__dirname, "../../../../");
   const moduleUrl = pathToFileURL(
     path.join(repoRoot, "packages/server/dist/server/server/exports.js"),
   ).href;
   const mod = (await import(moduleUrl)) as {
-    DaemonClient: new (config: {
-      url: string;
-      clientId: string;
-      clientType: "cli";
-    }) => TerminalPerfDaemonClient;
+    DaemonClient: new (config: TerminalPerfDaemonClientConfig) => TerminalPerfDaemonClient;
   };
   return mod.DaemonClient;
 }
 
 export async function connectTerminalClient(): Promise<TerminalPerfDaemonClient> {
   const DaemonClient = await loadDaemonClientConstructor();
+  const webSocketFactory = createNodeWebSocketFactory();
   const client = new DaemonClient({
     url: getDaemonWsUrl(),
     clientId: `terminal-perf-${randomUUID()}`,
     clientType: "cli",
+    webSocketFactory,
   });
   await client.connect();
   return client;
 }
 
-export function buildTerminalWorkspaceUrl(cwd: string, terminalId: string): string {
+export function buildTerminalWorkspaceUrl(workspaceId: string, terminalId: string): string {
   const serverId = getServerId();
-  const route = buildHostWorkspaceRoute(serverId, cwd);
+  const route = buildHostWorkspaceRoute(serverId, workspaceId);
   return `${route}?open=${encodeURIComponent(`terminal:${terminalId}`)}`;
 }
 
+function buildWorkspaceUrl(workspaceId: string): string {
+  return buildHostWorkspaceRoute(getServerId(), workspaceId);
+}
+
 export async function getTerminalBufferText(page: Page): Promise<string> {
   return page.evaluate(() => {
-    const term = (window as any).__paseoTerminal;
+    const term = (
+      window as Window & {
+        __paseoTerminal?: {
+          buffer: {
+            active: {
+              length: number;
+              getLine: (i: number) => { translateToString: (trim: boolean) => string } | null;
+            };
+          };
+          onWriteParsed: (cb: () => void) => { dispose: () => void };
+        };
+      }
+    ).__paseoTerminal;
     if (!term) {
       return "";
     }
@@ -113,38 +150,38 @@ export async function waitForTerminalContent(
 
 export async function navigateToTerminal(
   page: Page,
-  input: { cwd: string; terminalId: string },
+  input: { workspaceId: string; terminalId: string },
 ): Promise<void> {
   // Boot the app at the workspace route directly.
   // The fixtures.ts beforeEach addInitScript seeds localStorage on every navigation,
   // so the daemon registry is already configured when the app starts.
-  const workspaceRoute = buildHostWorkspaceRoute(getServerId(), input.cwd);
+  const workspaceRoute = buildTerminalWorkspaceUrl(input.workspaceId, input.terminalId);
   await page.goto(workspaceRoute);
 
+  // The workspace layout consumes `?open=...`, returns null during the effect,
+  // then replaces the URL with the clean workspace route after preparing the tab.
+  // On CI, Expo Router's rootNavigationState may take time to initialize,
+  // so we allow a generous timeout here.
+  const cleanWorkspaceRoute = buildWorkspaceUrl(input.workspaceId);
+  await page.waitForURL(
+    (url) => url.pathname === cleanWorkspaceRoute && !url.searchParams.has("open"),
+    { timeout: 30_000 },
+  );
+
   // Wait for daemon connection (sidebar shows host label)
-  await page.getByText("localhost", { exact: true }).first().waitFor({ state: "visible", timeout: 15_000 });
+  await page
+    .getByText("localhost", { exact: true })
+    .first()
+    .waitFor({ state: "visible", timeout: 30_000 });
+
+  // The open intent should have prepared and focused the exact pre-created terminal tab.
+  // The tab reconciliation effect also auto-creates terminal tabs once hydration completes,
+  // so we give it enough time for the full workspace hydration + tab creation cycle.
+  const terminalTab = page.locator(`[data-testid="workspace-tab-terminal_${input.terminalId}"]`);
+  await terminalTab.waitFor({ state: "visible", timeout: 30_000 });
+  await terminalTab.click();
 
-  // The workspace should now query listTerminals and discover our terminal.
-  // Click the terminal tab if it auto-appeared, or wait for it.
   const terminalSurface = page.locator('[data-testid="terminal-surface"]');
-  const surfaceVisible = await terminalSurface.isVisible().catch(() => false);
-
-  if (!surfaceVisible) {
-    // Terminal tab might not be focused — look for it in the tab row and click it
-    const terminalTab = page.locator(`[data-testid="workspace-tab-terminal:${input.terminalId}"]`);
-    const tabExists = await terminalTab.isVisible({ timeout: 5_000 }).catch(() => false);
-
-    if (tabExists) {
-      await terminalTab.click();
-    } else {
-      // Terminal tab not yet created — click "New terminal tab" to create one through the UI
-      const newTerminalBtn = page.getByRole("button", { name: "New terminal tab" });
-      await newTerminalBtn.waitFor({ state: "visible", timeout: 10_000 });
-      await newTerminalBtn.click();
-    }
-  }
-
-  // Wait for terminal surface to be visible
   await terminalSurface.waitFor({ state: "visible", timeout: 15_000 });
 
   // Wait for loading overlay to disappear (terminal attached)
@@ -155,6 +192,7 @@ export async function navigateToTerminal(
       // overlay may never appear if attachment is instant
     });
 
+  await terminalSurface.scrollIntoViewIfNeeded();
   await terminalSurface.click();
 }
 
@@ -169,10 +207,10 @@ export async function setupDeterministicPrompt(page: Page, sentinel?: string): P
   await page.waitForTimeout(300);
 }
 
-export type LatencySample = {
+export interface LatencySample {
   char: string;
   latencyMs: number;
-};
+}
 
 /**
  * Measures keystroke echo round-trip latency.
@@ -183,12 +221,16 @@ export type LatencySample = {
  */
 export async function measureKeystrokeLatency(page: Page, char: string): Promise<number> {
   await page.evaluate(() => {
-    const term = (window as any).__paseoTerminal;
-    if (!term) {
+    const win = window as Window & {
+      __paseoTerminal?: { onWriteParsed: (cb: () => void) => { dispose: () => void } };
+      __perfKeystroke?: { promise: Promise<number> | null };
+    };
+    if (!win.__paseoTerminal) {
       throw new Error("__paseoTerminal not available");
     }
+    const term = win.__paseoTerminal;
 
-    const state = ((window as any).__perfKeystroke = {
+    const state = (win.__perfKeystroke = {
       promise: null as Promise<number> | null,
     });
 
@@ -214,7 +256,39 @@ export async function measureKeystrokeLatency(page: Page, char: string): Promise
 
   await page.keyboard.press(char);
 
-  return page.evaluate(() => (window as any).__perfKeystroke.promise);
+  return page.evaluate(
+    () =>
+      (window as unknown as { __perfKeystroke: { promise: Promise<number> } }).__perfKeystroke
+        .promise,
+  );
+}
+
+export async function expectTerminalSurfaceVisible(
+  page: Page,
+  options?: { timeout?: number },
+): Promise<void> {
+  await expect(page.locator('[data-testid="terminal-surface"]').first()).toBeVisible({
+    timeout: options?.timeout ?? 20_000,
+  });
+}
+
+export async function focusTerminalSurface(page: Page): Promise<void> {
+  await expectTerminalSurfaceVisible(page);
+  await page.locator('[data-testid="terminal-surface"]').first().click();
+}
+
+export async function typeInTerminal(page: Page, text: string): Promise<void> {
+  await page
+    .locator('[data-testid="terminal-surface"]')
+    .first()
+    .pressSequentially(text, { delay: 0 });
+}
+
+export async function waitForTerminalAttached(page: Page): Promise<void> {
+  await page
+    .locator('[data-testid="terminal-attach-loading"]')
+    .waitFor({ state: "hidden", timeout: 10_000 })
+    .catch(() => undefined);
 }
 
 export function computePercentile(samples: number[], p: number): number {

packages/app/e2e/helpers/terminal-probes.ts

@@ -0,0 +1,711 @@
+import type { Page } from "@playwright/test";
+
+export interface TerminalRenderProbeSnapshot {
+  setCount: number;
+  unsetCount: number;
+  writeCount: number;
+  resetWrites: number;
+  clearScreenWrites: number;
+  altEnterWrites: number;
+  altExitWrites: number;
+  events: TerminalRenderProbeEvent[];
+  frames: TerminalFrame[];
+}
+
+export interface TerminalRenderProbeEvent {
+  at: number;
+  type: "set" | "unset" | "reset-write" | "clear-write" | "alt-enter-write" | "alt-exit-write";
+  preview?: string;
+}
+
+export interface TerminalFrame {
+  at: number;
+  rowCount: number;
+  nonEmptyRows: number;
+  firstNonEmptyRow: number | null;
+  text: string;
+  topText: string;
+}
+
+export type TerminalRenderProbeSummary = Omit<TerminalRenderProbeSnapshot, "frames"> & {
+  frameCount: number;
+};
+
+export interface TerminalKeystrokeStressReport {
+  inputTextLength: number;
+  keydownCount: number;
+  inputFrameCount: number;
+  outputFrameCount: number;
+  textMessageFrameCount: number;
+  textMessagePayloadBytes: number;
+  largeTextMessageCount: number;
+  largestTextMessageBytes: number;
+  agentStreamTextMessageCount: number;
+  agentStreamTextMessagePayloadBytes: number;
+  largeAgentStreamTextMessageCount: number;
+  largestAgentStreamTextMessageBytes: number;
+  appEventCount: number;
+  appEventCounts: Record<string, number>;
+  runtimeMaxQueueDepth: number;
+  xtermWriteCount: number;
+  inputFramePayloadBytes: number;
+  outputFramePayloadBytes: number;
+  keydownToInputFrameMs: LatencyStats | null;
+  inputFrameToOutputFrameMs: LatencyStats | null;
+  appBinaryReceivedToFrameDecodedMs: LatencyStats | null;
+  appFrameDecodedToTerminalEmitMs: LatencyStats | null;
+  appTerminalEmitListenerDurationMs: LatencyStats | null;
+  appTerminalEmitToStreamControllerOutputMs: LatencyStats | null;
+  appStreamControllerDecodeToOnOutputMs: LatencyStats | null;
+  appStreamControllerToEmulatorWriteMs: LatencyStats | null;
+  appEmulatorWriteToRuntimeEnqueuedMs: LatencyStats | null;
+  appRuntimeEnqueuedToOperationStartMs: LatencyStats | null;
+  appRuntimeOperationStartToXtermWriteMs: LatencyStats | null;
+  appRuntimeXtermWriteToCommitMs: LatencyStats | null;
+  appBinaryReceivedToRuntimeEnqueuedMs: LatencyStats | null;
+  appBinaryReceivedToRuntimeOperationStartMs: LatencyStats | null;
+  appBinaryReceivedToXtermCommitMs: LatencyStats | null;
+  outputFrameToXtermWriteMs: LatencyStats | null;
+  xtermWriteDurationMs: LatencyStats | null;
+  keydownToXtermCommitMs: LatencyStats | null;
+  firstKeydownAt: number | null;
+  lastXtermCommitAt: number | null;
+}
+
+export interface LatencyStats {
+  count: number;
+  minMs: number;
+  p50Ms: number;
+  p95Ms: number;
+  maxMs: number;
+  avgMs: number;
+}
+
+export async function installTerminalRenderProbe(page: Page): Promise<void> {
+  await page.addInitScript(() => {
+    interface ProbeTerm {
+      write?: (data: string | Uint8Array, callback?: () => void) => void;
+      __paseoRenderProbeWriteWrapped?: boolean;
+    }
+    interface ProbeState {
+      term: ProbeTerm | undefined;
+      setCount: number;
+      unsetCount: number;
+      writeCount: number;
+      resetWrites: number;
+      clearScreenWrites: number;
+      altEnterWrites: number;
+      altExitWrites: number;
+      events: TerminalRenderProbeEvent[];
+      frames: TerminalFrame[];
+      rafId: number | null;
+      sampleUntil: number;
+      reset: () => void;
+      snapshot: () => TerminalRenderProbeSnapshot;
+      startSampling: (durationMs: number) => void;
+    }
+
+    const win = window as unknown as Record<string, unknown> & {
+      __terminalRenderProbe?: ProbeState;
+      __paseoTerminal?: ProbeTerm;
+    };
+    const existingDescriptor = Object.getOwnPropertyDescriptor(win, "__paseoTerminal");
+    const getExisting = () =>
+      existingDescriptor?.get ? existingDescriptor.get.call(win) : existingDescriptor?.value;
+
+    const probe: ProbeState = {
+      term: getExisting(),
+      setCount: 0,
+      unsetCount: 0,
+      writeCount: 0,
+      resetWrites: 0,
+      clearScreenWrites: 0,
+      altEnterWrites: 0,
+      altExitWrites: 0,
+      events: [],
+      frames: [],
+      rafId: null,
+      sampleUntil: 0,
+      reset() {
+        this.setCount = 0;
+        this.unsetCount = 0;
+        this.writeCount = 0;
+        this.resetWrites = 0;
+        this.clearScreenWrites = 0;
+        this.altEnterWrites = 0;
+        this.altExitWrites = 0;
+        this.events = [];
+        this.frames = [];
+      },
+      snapshot() {
+        return {
+          setCount: this.setCount,
+          unsetCount: this.unsetCount,
+          writeCount: this.writeCount,
+          resetWrites: this.resetWrites,
+          clearScreenWrites: this.clearScreenWrites,
+          altEnterWrites: this.altEnterWrites,
+          altExitWrites: this.altExitWrites,
+          events: this.events,
+          frames: this.frames,
+        };
+      },
+      startSampling(durationMs: number) {
+        this.sampleUntil = performance.now() + durationMs;
+        if (this.rafId !== null) {
+          cancelAnimationFrame(this.rafId);
+        }
+        const sample = () => {
+          const rows = Array.from(document.querySelectorAll(".xterm-rows > div")).map(
+            (row) => row.textContent ?? "",
+          );
+          const nonEmptyRows = rows.filter((row) => row.trim().length > 0);
+          const firstNonEmptyRow = rows.findIndex((row) => row.trim().length > 0);
+          this.frames.push({
+            at: performance.now(),
+            rowCount: rows.length,
+            nonEmptyRows: nonEmptyRows.length,
+            firstNonEmptyRow: firstNonEmptyRow === -1 ? null : firstNonEmptyRow,
+            text: rows.join("\n"),
+            topText: rows.slice(0, 3).join("\n"),
+          });
+          if (performance.now() < this.sampleUntil) {
+            this.rafId = requestAnimationFrame(sample);
+          } else {
+            this.rafId = null;
+          }
+        };
+        this.rafId = requestAnimationFrame(sample);
+      },
+    };
+
+    Object.defineProperty(win, "__terminalRenderProbe", {
+      configurable: true,
+      value: probe,
+    });
+
+    Object.defineProperty(win, "__paseoTerminal", {
+      configurable: true,
+      get() {
+        return probe.term;
+      },
+      set(next: ProbeTerm | undefined) {
+        if (next === undefined) {
+          probe.unsetCount += 1;
+          probe.events.push({ at: performance.now(), type: "unset" });
+          probe.term = next;
+          return;
+        }
+
+        probe.setCount += 1;
+        probe.events.push({ at: performance.now(), type: "set" });
+        probe.term = next;
+
+        if (next?.write && !next.__paseoRenderProbeWriteWrapped) {
+          const originalWrite = next.write.bind(next);
+          next.write = (data: string | Uint8Array, callback?: () => void) => {
+            const text = typeof data === "string" ? data : new TextDecoder().decode(data);
+            probe.writeCount += 1;
+            const preview = text
+              .replaceAll("\u001b", "\\x1b")
+              .replace(/\r/g, "\\r")
+              .replace(/\n/g, "\\n")
+              .slice(0, 160);
+            if (text.includes("\u001bc")) {
+              probe.resetWrites += 1;
+              probe.events.push({ at: performance.now(), type: "reset-write", preview });
+            }
+            if (text.includes("\u001b[2J")) {
+              probe.clearScreenWrites += 1;
+              probe.events.push({ at: performance.now(), type: "clear-write", preview });
+            }
+            if (text.includes("\u001b[?1049h")) {
+              probe.altEnterWrites += 1;
+              probe.events.push({ at: performance.now(), type: "alt-enter-write", preview });
+            }
+            if (text.includes("\u001b[?1049l")) {
+              probe.altExitWrites += 1;
+              probe.events.push({ at: performance.now(), type: "alt-exit-write", preview });
+            }
+            return originalWrite(data, callback);
+          };
+          next.__paseoRenderProbeWriteWrapped = true;
+        }
+      },
+    });
+  });
+}
+
+interface TerminalRenderProbeWindow {
+  __terminalRenderProbe: {
+    reset: () => void;
+    startSampling: (durationMs: number) => void;
+    snapshot: () => TerminalRenderProbeSnapshot;
+  };
+}
+
+export async function resetTerminalRenderProbe(page: Page): Promise<void> {
+  await page.evaluate(() => {
+    (window as unknown as TerminalRenderProbeWindow).__terminalRenderProbe.reset();
+  });
+}
+
+export async function startTerminalFrameSampling(page: Page, durationMs = 2500): Promise<void> {
+  await page.evaluate((ms) => {
+    (window as unknown as TerminalRenderProbeWindow).__terminalRenderProbe.startSampling(ms);
+  }, durationMs);
+}
+
+export async function readTerminalRenderProbe(page: Page): Promise<TerminalRenderProbeSnapshot> {
+  return page.evaluate(() =>
+    (window as unknown as TerminalRenderProbeWindow).__terminalRenderProbe.snapshot(),
+  );
+}
+
+export async function terminalVisibleText(page: Page): Promise<string> {
+  return page.evaluate(() =>
+    Array.from(document.querySelectorAll(".xterm-rows > div"))
+      .map((row) => row.textContent ?? "")
+      .join("\n"),
+  );
+}
+
+export function summarizeTerminalRenderProbe(
+  probe: TerminalRenderProbeSnapshot,
+): TerminalRenderProbeSummary {
+  return {
+    setCount: probe.setCount,
+    unsetCount: probe.unsetCount,
+    writeCount: probe.writeCount,
+    resetWrites: probe.resetWrites,
+    clearScreenWrites: probe.clearScreenWrites,
+    altEnterWrites: probe.altEnterWrites,
+    altExitWrites: probe.altExitWrites,
+    events: probe.events,
+    frameCount: probe.frames.length,
+  };
+}
+
+export async function installTerminalKeystrokeStressProbe(page: Page): Promise<void> {
+  await page.addInitScript(() => {
+    interface TimedTextEvent {
+      at: number;
+      text: string;
+      bytes: number;
+    }
+    interface TimedTextMessageEvent {
+      at: number;
+      bytes: number;
+      kind: string | null;
+    }
+    interface XtermWriteEvent {
+      at: number;
+      committedAt: number | null;
+      text: string;
+      bytes: number;
+    }
+    interface AppProbeEvent {
+      type: string;
+      at: number;
+      bytes?: number;
+      queueDepth?: number;
+    }
+    interface StressProbeState {
+      keydowns: Array<{ at: number; key: string }>;
+      inputFrames: TimedTextEvent[];
+      outputFrames: TimedTextEvent[];
+      textMessageFrames: TimedTextMessageEvent[];
+      xtermWrites: XtermWriteEvent[];
+      appEvents: AppProbeEvent[];
+      reset: () => void;
+      report: (inputText: string) => TerminalKeystrokeStressReport;
+    }
+
+    const INPUT_OPCODE = 0x02;
+    const OUTPUT_OPCODE = 0x01;
+    const decoder = new TextDecoder();
+
+    function bytesFrom(data: unknown): Uint8Array | null {
+      if (data instanceof Uint8Array) return data;
+      if (data instanceof ArrayBuffer) return new Uint8Array(data);
+      if (ArrayBuffer.isView(data)) {
+        return new Uint8Array(data.buffer, data.byteOffset, data.byteLength);
+      }
+      return null;
+    }
+
+    function eventDataBytes(data: unknown): Promise<Uint8Array | null> {
+      const bytes = bytesFrom(data);
+      if (bytes) return Promise.resolve(bytes);
+      if (typeof Blob !== "undefined" && data instanceof Blob) {
+        return data.arrayBuffer().then((buffer) => new Uint8Array(buffer));
+      }
+      return Promise.resolve(null);
+    }
+
+    function frameText(bytes: Uint8Array): string {
+      return decoder.decode(bytes.slice(2));
+    }
+
+    function eventDataText(data: unknown): Promise<string | null> {
+      if (typeof data === "string") {
+        return Promise.resolve(data);
+      }
+      if (typeof Blob !== "undefined" && data instanceof Blob) {
+        return data.text();
+      }
+      return Promise.resolve(null);
+    }
+
+    function textMessageKind(text: string): string | null {
+      try {
+        const parsed = JSON.parse(text) as {
+          type?: unknown;
+          message?: {
+            type?: unknown;
+          };
+        };
+        if (typeof parsed.type !== "string") {
+          return null;
+        }
+        if (typeof parsed.message?.type === "string") {
+          return `${parsed.type}:${parsed.message.type}`;
+        }
+        return parsed.type;
+      } catch {
+        return null;
+      }
+    }
+
+    function summarize(values: number[]): LatencyStats | null {
+      if (values.length === 0) return null;
+      const sorted = [...values].sort((a, b) => a - b);
+      const percentile = (p: number) => {
+        const index = Math.ceil((p / 100) * sorted.length) - 1;
+        return sorted[Math.max(0, Math.min(sorted.length - 1, index))];
+      };
+      const total = values.reduce((sum, value) => sum + value, 0);
+      const round2 = (value: number) => Math.round(value * 100) / 100;
+      return {
+        count: values.length,
+        minMs: round2(sorted[0] ?? 0),
+        p50Ms: round2(percentile(50)),
+        p95Ms: round2(percentile(95)),
+        maxMs: round2(sorted[sorted.length - 1] ?? 0),
+        avgMs: round2(total / values.length),
+      };
+    }
+
+    function firstAtOrAfter<T extends { at: number }>(events: T[], at: number): T | null {
+      return events.find((event) => event.at >= at) ?? null;
+    }
+
+    function firstCommitAtOrAfter(events: XtermWriteEvent[], at: number): XtermWriteEvent | null {
+      return (
+        events.find((event) => typeof event.committedAt === "number" && event.committedAt >= at) ??
+        null
+      );
+    }
+
+    function countByType(events: AppProbeEvent[]): Record<string, number> {
+      const counts: Record<string, number> = {};
+      for (const event of events) {
+        counts[event.type] = (counts[event.type] ?? 0) + 1;
+      }
+      return counts;
+    }
+
+    function appEventsOf(type: string, events: AppProbeEvent[]): AppProbeEvent[] {
+      return events.filter((event) => event.type === type);
+    }
+
+    function latencyByIndex(from: AppProbeEvent[], to: AppProbeEvent[]): number[] {
+      const count = Math.min(from.length, to.length);
+      const values: number[] = [];
+      for (let index = 0; index < count; index += 1) {
+        values.push(to[index].at - from[index].at);
+      }
+      return values;
+    }
+
+    const probe: StressProbeState = {
+      keydowns: [],
+      inputFrames: [],
+      outputFrames: [],
+      textMessageFrames: [],
+      xtermWrites: [],
+      appEvents: [],
+      reset() {
+        this.keydowns = [];
+        this.inputFrames = [];
+        this.outputFrames = [];
+        this.textMessageFrames = [];
+        this.xtermWrites = [];
+        this.appEvents = [];
+      },
+      report(inputText: string) {
+        const binaryReceived = appEventsOf("daemon-client-binary-received", this.appEvents);
+        const frameDecoded = appEventsOf("daemon-client-frame-decoded", this.appEvents);
+        const terminalEmit = appEventsOf("daemon-client-terminal-emit", this.appEvents);
+        const terminalEmitted = appEventsOf("daemon-client-terminal-emitted", this.appEvents);
+        const streamControllerOutput = appEventsOf("stream-controller-output", this.appEvents);
+        const streamControllerOnOutput = appEventsOf("stream-controller-on-output", this.appEvents);
+        const emulatorWriteOutput = appEventsOf("terminal-emulator-write-output", this.appEvents);
+        const runtimeWriteEnqueued = appEventsOf("runtime-write-enqueued", this.appEvents);
+        const runtimeOperationStart = appEventsOf("runtime-operation-start", this.appEvents);
+        const runtimeXtermWrite = appEventsOf("runtime-xterm-write", this.appEvents);
+        const runtimeXtermCommitted = appEventsOf("runtime-xterm-committed", this.appEvents);
+        const keydownToInputFrame = this.keydowns
+          .map((keydown) => firstAtOrAfter(this.inputFrames, keydown.at)?.at ?? null)
+          .filter((at): at is number => at !== null)
+          .map((at, index) => at - this.keydowns[index].at);
+        const inputFrameToOutputFrame = this.inputFrames
+          .map((input) => {
+            const output = firstAtOrAfter(this.outputFrames, input.at);
+            return output ? output.at - input.at : null;
+          })
+          .filter((value): value is number => value !== null);
+        const outputFrameToXtermWrite = this.outputFrames
+          .map((output) => {
+            const write = firstAtOrAfter(this.xtermWrites, output.at);
+            return write ? write.at - output.at : null;
+          })
+          .filter((value): value is number => value !== null);
+        const xtermWriteDurations = this.xtermWrites
+          .map((write) => (write.committedAt === null ? null : write.committedAt - write.at))
+          .filter((value): value is number => value !== null);
+        const keydownToXtermCommit = this.keydowns
+          .map((keydown) => {
+            const write = firstCommitAtOrAfter(this.xtermWrites, keydown.at);
+            return write?.committedAt ? write.committedAt - keydown.at : null;
+          })
+          .filter((value): value is number => value !== null);
+
+        return {
+          inputTextLength: inputText.length,
+          keydownCount: this.keydowns.length,
+          inputFrameCount: this.inputFrames.length,
+          outputFrameCount: this.outputFrames.length,
+          textMessageFrameCount: this.textMessageFrames.length,
+          textMessagePayloadBytes: this.textMessageFrames.reduce(
+            (sum, frame) => sum + frame.bytes,
+            0,
+          ),
+          largeTextMessageCount: this.textMessageFrames.filter((frame) => frame.bytes >= 50_000)
+            .length,
+          largestTextMessageBytes: Math.max(
+            0,
+            ...this.textMessageFrames.map((frame) => frame.bytes),
+          ),
+          agentStreamTextMessageCount: this.textMessageFrames.filter(
+            (frame) => frame.kind === "session:agent_stream",
+          ).length,
+          agentStreamTextMessagePayloadBytes: this.textMessageFrames
+            .filter((frame) => frame.kind === "session:agent_stream")
+            .reduce((sum, frame) => sum + frame.bytes, 0),
+          largeAgentStreamTextMessageCount: this.textMessageFrames.filter(
+            (frame) => frame.kind === "session:agent_stream" && frame.bytes >= 50_000,
+          ).length,
+          largestAgentStreamTextMessageBytes: Math.max(
+            0,
+            ...this.textMessageFrames
+              .filter((frame) => frame.kind === "session:agent_stream")
+              .map((frame) => frame.bytes),
+          ),
+          appEventCount: this.appEvents.length,
+          appEventCounts: countByType(this.appEvents),
+          runtimeMaxQueueDepth: Math.max(
+            0,
+            ...this.appEvents
+              .map((event) => event.queueDepth)
+              .filter((value): value is number => typeof value === "number"),
+          ),
+          xtermWriteCount: this.xtermWrites.length,
+          inputFramePayloadBytes: this.inputFrames.reduce((sum, frame) => sum + frame.bytes, 0),
+          outputFramePayloadBytes: this.outputFrames.reduce((sum, frame) => sum + frame.bytes, 0),
+          keydownToInputFrameMs: summarize(keydownToInputFrame),
+          inputFrameToOutputFrameMs: summarize(inputFrameToOutputFrame),
+          appBinaryReceivedToFrameDecodedMs: summarize(
+            latencyByIndex(binaryReceived, frameDecoded),
+          ),
+          appFrameDecodedToTerminalEmitMs: summarize(latencyByIndex(frameDecoded, terminalEmit)),
+          appTerminalEmitListenerDurationMs: summarize(
+            latencyByIndex(terminalEmit, terminalEmitted),
+          ),
+          appTerminalEmitToStreamControllerOutputMs: summarize(
+            latencyByIndex(terminalEmit, streamControllerOutput),
+          ),
+          appStreamControllerDecodeToOnOutputMs: summarize(
+            latencyByIndex(streamControllerOutput, streamControllerOnOutput),
+          ),
+          appStreamControllerToEmulatorWriteMs: summarize(
+            latencyByIndex(streamControllerOnOutput, emulatorWriteOutput),
+          ),
+          appEmulatorWriteToRuntimeEnqueuedMs: summarize(
+            latencyByIndex(emulatorWriteOutput, runtimeWriteEnqueued),
+          ),
+          appRuntimeEnqueuedToOperationStartMs: summarize(
+            latencyByIndex(runtimeWriteEnqueued, runtimeOperationStart),
+          ),
+          appRuntimeOperationStartToXtermWriteMs: summarize(
+            latencyByIndex(runtimeOperationStart, runtimeXtermWrite),
+          ),
+          appRuntimeXtermWriteToCommitMs: summarize(
+            latencyByIndex(runtimeXtermWrite, runtimeXtermCommitted),
+          ),
+          appBinaryReceivedToRuntimeEnqueuedMs: summarize(
+            latencyByIndex(binaryReceived, runtimeWriteEnqueued),
+          ),
+          appBinaryReceivedToRuntimeOperationStartMs: summarize(
+            latencyByIndex(binaryReceived, runtimeOperationStart),
+          ),
+          appBinaryReceivedToXtermCommitMs: summarize(
+            latencyByIndex(binaryReceived, runtimeXtermCommitted),
+          ),
+          outputFrameToXtermWriteMs: summarize(outputFrameToXtermWrite),
+          xtermWriteDurationMs: summarize(xtermWriteDurations),
+          keydownToXtermCommitMs: summarize(keydownToXtermCommit),
+          firstKeydownAt: this.keydowns[0]?.at ?? null,
+          lastXtermCommitAt:
+            this.xtermWrites
+              .map((write) => write.committedAt)
+              .findLast((at): at is number => typeof at === "number") ?? null,
+        };
+      },
+    };
+
+    Object.defineProperty(window, "__terminalKeystrokeStressProbe", {
+      configurable: true,
+      value: probe,
+    });
+
+    document.addEventListener(
+      "keydown",
+      (event) => {
+        if (event.key.length === 1) {
+          probe.keydowns.push({ at: performance.now(), key: event.key });
+        }
+      },
+      true,
+    );
+
+    const NativeWebSocket = window.WebSocket;
+    class InstrumentedWebSocket extends NativeWebSocket {
+      constructor(url: string | URL, protocols?: string | string[]) {
+        if (protocols === undefined) {
+          super(url);
+        } else {
+          super(url, protocols);
+        }
+        super.addEventListener("message", (event) => {
+          void eventDataBytes(event.data).then((bytes) => {
+            if (!bytes || bytes.byteLength < 2 || bytes[0] !== OUTPUT_OPCODE) {
+              return;
+            }
+            probe.outputFrames.push({
+              at: performance.now(),
+              text: frameText(bytes),
+              bytes: bytes.byteLength - 2,
+            });
+            return;
+          });
+          void eventDataText(event.data).then((text) => {
+            if (text === null) {
+              return;
+            }
+            probe.textMessageFrames.push({
+              at: performance.now(),
+              bytes: new TextEncoder().encode(text).byteLength,
+              kind: textMessageKind(text),
+            });
+            return;
+          });
+        });
+      }
+
+      send(data: Parameters<WebSocket["send"]>[0]): void {
+        const bytes = bytesFrom(data);
+        if (bytes && bytes.byteLength >= 2 && bytes[0] === INPUT_OPCODE) {
+          probe.inputFrames.push({
+            at: performance.now(),
+            text: frameText(bytes),
+            bytes: bytes.byteLength - 2,
+          });
+        }
+        super.send(data);
+      }
+    }
+
+    Object.defineProperty(InstrumentedWebSocket, "CONNECTING", {
+      value: NativeWebSocket.CONNECTING,
+    });
+    Object.defineProperty(InstrumentedWebSocket, "OPEN", { value: NativeWebSocket.OPEN });
+    Object.defineProperty(InstrumentedWebSocket, "CLOSING", { value: NativeWebSocket.CLOSING });
+    Object.defineProperty(InstrumentedWebSocket, "CLOSED", { value: NativeWebSocket.CLOSED });
+    window.WebSocket = InstrumentedWebSocket as typeof WebSocket;
+
+    const existingDescriptor = Object.getOwnPropertyDescriptor(window, "__paseoTerminal");
+    const getExisting = () =>
+      existingDescriptor?.get ? existingDescriptor.get.call(window) : existingDescriptor?.value;
+
+    let terminal = getExisting();
+    Object.defineProperty(window, "__paseoTerminal", {
+      configurable: true,
+      get() {
+        return terminal;
+      },
+      set(next: {
+        write?: (data: string | Uint8Array, callback?: () => void) => void;
+        __paseoKeystrokeProbeWriteWrapped?: boolean;
+      }) {
+        terminal = next;
+        if (next?.write && !next.__paseoKeystrokeProbeWriteWrapped) {
+          const originalWrite = next.write.bind(next);
+          next.write = (data: string | Uint8Array, callback?: () => void) => {
+            const text = typeof data === "string" ? data : new TextDecoder().decode(data);
+            const event: XtermWriteEvent = {
+              at: performance.now(),
+              committedAt: null,
+              text,
+              bytes: text.length,
+            };
+            probe.xtermWrites.push(event);
+            return originalWrite(data, () => {
+              event.committedAt = performance.now();
+              callback?.();
+            });
+          };
+          next.__paseoKeystrokeProbeWriteWrapped = true;
+        }
+      },
+    });
+  });
+}
+
+interface TerminalKeystrokeStressProbeWindow {
+  __terminalKeystrokeStressProbe: {
+    reset: () => void;
+    report: (text: string) => TerminalKeystrokeStressReport;
+  };
+}
+
+export async function resetTerminalKeystrokeStressProbe(page: Page): Promise<void> {
+  await page.evaluate(() => {
+    (
+      window as unknown as TerminalKeystrokeStressProbeWindow
+    ).__terminalKeystrokeStressProbe.reset();
+  });
+}
+
+export async function readTerminalKeystrokeStressReport(
+  page: Page,
+  inputText: string,
+): Promise<TerminalKeystrokeStressReport> {
+  return page.evaluate(
+    (text) =>
+      (
+        window as unknown as TerminalKeystrokeStressProbeWindow
+      ).__terminalKeystrokeStressProbe.report(text),
+    inputText,
+  );
+}

packages/app/e2e/helpers/with-workspace.ts

@@ -0,0 +1,105 @@
+import { execSync } from "node:child_process";
+import { realpath } from "node:fs/promises";
+import path from "node:path";
+import type { Page } from "@playwright/test";
+import { waitForTabBar } from "./launcher";
+import { selectWorkspaceInSidebar } from "./sidebar";
+import { createTempGitRepo } from "./workspace";
+import {
+  connectWorkspaceSetupClient,
+  openHomeWithProject,
+  type WorkspaceSetupDaemonClient,
+} from "./workspace-setup";
+
+export interface CreatedWorkspace {
+  workspaceId: string;
+  repoPath: string;
+  navigateTo(): Promise<void>;
+}
+
+export interface WithWorkspaceOptions {
+  worktree?: boolean;
+  prefix?: string;
+}
+
+export type WithWorkspace = (options?: WithWorkspaceOptions) => Promise<CreatedWorkspace>;
+
+interface WorktreeRecord {
+  repoPath: string;
+  worktreePath: string;
+}
+
+export interface WithWorkspaceHandle {
+  withWorkspace: WithWorkspace;
+  cleanup: () => Promise<void>;
+}
+
+export function createWithWorkspace(page: Page): WithWorkspaceHandle {
+  let client: WorkspaceSetupDaemonClient | null = null;
+  const repos: Array<{ cleanup: () => Promise<void> }> = [];
+  const worktrees: WorktreeRecord[] = [];
+
+  const withWorkspace: WithWorkspace = async (options) => {
+    if (!client) {
+      client = await connectWorkspaceSetupClient();
+    }
+    const prefix = options?.prefix ?? (options?.worktree ? "wt-" : "ws-");
+    const repo = await createTempGitRepo(prefix);
+    repos.push(repo);
+
+    let workspacePath = repo.path;
+    if (options?.worktree) {
+      const tempRoot = await realpath("/tmp");
+      workspacePath = path.join(
+        tempRoot,
+        `paseo-wt-${Date.now()}-${Math.random().toString(36).slice(2)}`,
+      );
+      const branchName = `paseo-wt-${Date.now()}`;
+      execSync(
+        `git worktree add ${JSON.stringify(workspacePath)} -b ${JSON.stringify(branchName)} main`,
+        { cwd: repo.path, stdio: "ignore" },
+      );
+      worktrees.push({ repoPath: repo.path, worktreePath: workspacePath });
+      // Register the parent project so the sidebar lists it before we navigate.
+      await client.openProject(repo.path);
+    }
+
+    const opened = await client.openProject(workspacePath);
+    if (!opened.workspace) {
+      throw new Error(opened.error ?? `Failed to open project ${workspacePath}`);
+    }
+    const workspaceId = opened.workspace.id;
+
+    return {
+      workspaceId,
+      repoPath: workspacePath,
+      navigateTo: async () => {
+        await openHomeWithProject(page, repo.path);
+        await selectWorkspaceInSidebar(page, workspaceId);
+        await waitForTabBar(page);
+      },
+    };
+  };
+
+  return {
+    withWorkspace,
+    cleanup: async () => {
+      for (const { repoPath, worktreePath } of worktrees) {
+        try {
+          execSync(`git worktree remove ${JSON.stringify(worktreePath)} --force`, {
+            cwd: repoPath,
+            stdio: "ignore",
+          });
+        } catch {
+          // Best-effort cleanup so the original test failure is preserved.
+        }
+      }
+      for (const repo of repos) {
+        await repo.cleanup();
+      }
+      if (client) {
+        await client.close().catch(() => undefined);
+      }
+    },
+  };
+}

packages/app/e2e/helpers/workspace-lifecycle.ts

@@ -0,0 +1,32 @@
+import { expect, type Page } from "@playwright/test";
+import { clickNewChat, clickNewTerminal } from "./launcher";
+import { setupDeterministicPrompt, waitForTerminalContent } from "./terminal-perf";
+
+function terminalSurface(page: Page) {
+  return page.locator('[data-testid="terminal-surface"]').first();
+}
+
+function composerInput(page: Page) {
+  return page.getByRole("textbox", { name: "Message agent..." }).first();
+}
+
+export async function expectTerminalCwd(page: Page, expectedPath: string): Promise<void> {
+  const terminal = terminalSurface(page);
+  await expect(terminal).toBeVisible({ timeout: 20_000 });
+  await terminal.click();
+  await setupDeterministicPrompt(page, `SENTINEL_${Date.now()}`);
+  await terminal.pressSequentially("pwd\n", { delay: 0 });
+  await waitForTerminalContent(page, (text) => text.includes(expectedPath), 10_000);
+}
+
+export async function createStandaloneTerminalFromLauncher(page: Page): Promise<void> {
+  await clickNewTerminal(page);
+  await expect(terminalSurface(page)).toBeVisible({ timeout: 20_000 });
+}
+
+export async function createAgentChatFromLauncher(page: Page): Promise<void> {
+  await clickNewChat(page);
+  await expect(composerInput(page)).toBeVisible({ timeout: 15_000 });
+  await expect(composerInput(page)).toBeEditable({ timeout: 15_000 });
+  await expect(page.getByTestId("agent-loading")).toHaveCount(0);
+}

packages/app/e2e/helpers/workspace-setup.ts

@@ -0,0 +1,366 @@
+import { realpathSync } from "node:fs";
+import path from "node:path";
+import { randomUUID } from "node:crypto";
+import { pathToFileURL } from "node:url";
+import { expect, type Page } from "@playwright/test";
+import { parseHostWorkspaceRouteFromPathname } from "../../src/utils/host-routes";
+import { gotoAppShell } from "./app";
+import { createNodeWebSocketFactory, type NodeWebSocketFactory } from "./node-ws-factory";
+import { switchWorkspaceViaSidebar } from "./workspace-ui";
+import type { SessionOutboundMessage } from "@server/shared/messages";
+
+interface WorkspaceSetupDaemonClient {
+  connect(): Promise<void>;
+  close(): Promise<void>;
+  openProject(cwd: string): Promise<{
+    workspace: {
+      id: string;
+      name: string;
+      workspaceDirectory: string;
+      projectRootPath: string;
+    } | null;
+    error: string | null;
+  }>;
+  createPaseoWorktree(input: { cwd: string; worktreeSlug?: string }): Promise<{
+    workspace: {
+      id: string;
+      name: string;
+      workspaceDirectory: string;
+      projectRootPath: string;
+    } | null;
+    error: string | null;
+  }>;
+  fetchWorkspaces(): Promise<{
+    entries: Array<{
+      id: string;
+      name: string;
+      workspaceDirectory: string;
+      projectRootPath: string;
+    }>;
+  }>;
+  fetchAgents(): Promise<{
+    entries: Array<{
+      agent: { id: string; cwd: string; workspaceId?: string | null };
+    }>;
+  }>;
+  fetchAgent(agentId: string): Promise<{
+    agent: { id: string; cwd: string } | null;
+    project: unknown;
+  } | null>;
+  listTerminals(cwd: string): Promise<{
+    cwd?: string;
+    terminals: Array<{ id: string; cwd: string; name: string }>;
+    error?: string | null;
+  }>;
+  subscribeRawMessages(handler: (message: SessionOutboundMessage) => void): () => void;
+}
+
+export type WorkspaceSetupProgressPayload = Extract<
+  SessionOutboundMessage,
+  { type: "workspace_setup_progress" }
+>["payload"];
+
+export type { WorkspaceSetupDaemonClient };
+
+function getDaemonWsUrl(): string {
+  const daemonPort = process.env.E2E_DAEMON_PORT;
+  if (!daemonPort) {
+    throw new Error("E2E_DAEMON_PORT is not set.");
+  }
+  return `ws://127.0.0.1:${daemonPort}/ws`;
+}
+
+async function loadDaemonClientConstructor(): Promise<
+  new (config: {
+    url: string;
+    clientId: string;
+    clientType: "cli";
+    webSocketFactory?: NodeWebSocketFactory;
+  }) => WorkspaceSetupDaemonClient
+> {
+  const repoRoot = path.resolve(process.cwd(), "../..");
+  const moduleUrl = pathToFileURL(
+    path.join(repoRoot, "packages/server/dist/server/server/exports.js"),
+  ).href;
+  const mod = (await import(moduleUrl)) as {
+    DaemonClient: new (config: {
+      url: string;
+      clientId: string;
+      clientType: "cli";
+      webSocketFactory?: NodeWebSocketFactory;
+    }) => WorkspaceSetupDaemonClient;
+  };
+  return mod.DaemonClient;
+}
+
+export async function connectWorkspaceSetupClient(): Promise<WorkspaceSetupDaemonClient> {
+  const DaemonClient = await loadDaemonClientConstructor();
+  const webSocketFactory = createNodeWebSocketFactory();
+  const client = new DaemonClient({
+    url: getDaemonWsUrl(),
+    clientId: `workspace-setup-${randomUUID()}`,
+    clientType: "cli",
+    webSocketFactory,
+  });
+  await client.connect();
+  return client;
+}
+
+export async function seedProjectForWorkspaceSetup(
+  client: WorkspaceSetupDaemonClient,
+  repoPath: string,
+): Promise<void> {
+  const result = await client.openProject(repoPath);
+  if (!result.workspace || result.error) {
+    throw new Error(result.error ?? `Failed to open project ${repoPath}`);
+  }
+}
+
+export function projectNameFromPath(repoPath: string): string {
+  return repoPath.replace(/\/+$/, "").split("/").findLast(Boolean) ?? repoPath;
+}
+
+export async function openHomeWithProject(page: Page, repoPath: string): Promise<void> {
+  await gotoAppShell(page);
+  await expect(
+    page
+      .locator('[data-testid^="sidebar-project-row-"]')
+      .filter({ hasText: projectNameFromPath(repoPath) })
+      .first(),
+  ).toBeVisible({ timeout: 30_000 });
+}
+
+function createWorkspaceButton(page: Page, repoPath: string) {
+  return page.getByRole("button", {
+    name: `Create a new workspace for ${projectNameFromPath(repoPath)}`,
+  });
+}
+
+async function revealWorkspaceButton(page: Page, repoPath: string): Promise<void> {
+  await page
+    .locator('[data-testid^="sidebar-project-row-"]')
+    .filter({ hasText: projectNameFromPath(repoPath) })
+    .first()
+    .hover();
+}
+
+export async function createWorkspaceFromSidebar(page: Page, repoPath: string): Promise<void> {
+  const button = createWorkspaceButton(page, repoPath);
+  await revealWorkspaceButton(page, repoPath);
+  await expect(button).toBeVisible({ timeout: 30_000 });
+  await expect(button).toBeEnabled({ timeout: 30_000 });
+  await button.click();
+  await expect(page).toHaveURL(/\/new\?/, { timeout: 30_000 });
+  await expect(page.getByRole("textbox", { name: "Message agent..." }).first()).toBeVisible({
+    timeout: 30_000,
+  });
+}
+
+export async function getCurrentWorkspaceIdFromRoute(page: Page): Promise<string> {
+  await expect
+    .poll(
+      () => parseHostWorkspaceRouteFromPathname(new URL(page.url()).pathname)?.workspaceId ?? null,
+      { timeout: 30_000 },
+    )
+    .not.toBeNull();
+
+  const workspaceId =
+    parseHostWorkspaceRouteFromPathname(new URL(page.url()).pathname)?.workspaceId ?? null;
+  if (!workspaceId) {
+    throw new Error(`Expected a workspace route but found ${page.url()}`);
+  }
+
+  return workspaceId;
+}
+
+function workspaceSetupDialog(page: Page) {
+  return page.getByTestId("workspace-setup-dialog");
+}
+
+export async function createChatAgentFromWorkspaceSetup(
+  page: Page,
+  input: { message: string },
+): Promise<void> {
+  const messageInput = page.getByRole("textbox", { name: "Message agent..." }).first();
+  await expect(messageInput).toBeVisible({ timeout: 15_000 });
+  await messageInput.fill(input.message);
+  await messageInput.press("Enter");
+}
+
+/**
+ * @deprecated The new workspace screen no longer has a standalone terminal button.
+ * Use the daemon API to create a workspace, then open a terminal from the launcher.
+ */
+export async function createStandaloneTerminalFromWorkspaceSetup(page: Page): Promise<void> {
+  await workspaceSetupDialog(page)
+    .getByRole("button", { name: /^Terminal Create the workspace/i })
+    .click();
+}
+
+export async function waitForWorkspaceSetupDialogToClose(
+  page: Page,
+  timeoutMs = 45_000,
+): Promise<void> {
+  const dialog = workspaceSetupDialog(page);
+
+  try {
+    await expect(dialog).toHaveCount(0, { timeout: timeoutMs });
+  } catch (error) {
+    const dialogText = (await dialog.textContent().catch(() => null))?.replace(/\s+/g, " ").trim();
+    throw new Error(
+      dialogText
+        ? `Workspace setup dialog stayed open. Visible text: ${dialogText}`
+        : `Workspace setup dialog did not close within ${timeoutMs}ms`,
+      { cause: error },
+    );
+  }
+}
+
+export async function expectSetupPanel(page: Page): Promise<void> {
+  // If the setup panel is already visible (auto-opened), we're done.
+  const panel = page.getByTestId("workspace-setup-panel");
+  if (await panel.isVisible().catch(() => false)) {
+    return;
+  }
+  // Otherwise open it manually via workspace header actions menu.
+  // Use the specific testID to avoid matching the sidebar kebab which shares
+  // the same "Workspace actions" accessibility label.
+  const actionsButton = page.getByTestId("workspace-header-menu-trigger");
+  await expect(actionsButton).toBeVisible({ timeout: 10_000 });
+  await actionsButton.click();
+  const showSetup = page.getByTestId("workspace-header-show-setup");
+  await expect(showSetup).toBeVisible({ timeout: 5_000 });
+  await showSetup.click();
+  await expect(panel).toBeVisible({ timeout: 30_000 });
+}
+
+export async function expectSetupStatus(
+  page: Page,
+  status: "Running" | "Completed" | "Failed",
+): Promise<void> {
+  await expect(page.getByTestId("workspace-setup-status")).toContainText(status, {
+    timeout: 30_000,
+  });
+}
+
+export async function expectSetupLogContains(page: Page, text: string): Promise<void> {
+  await expect(page.getByTestId("workspace-setup-log")).toContainText(text, {
+    timeout: 30_000,
+  });
+}
+
+export async function expectNoSetupMessage(page: Page): Promise<void> {
+  await expect(
+    page.getByText("No setup commands ran for this workspace.", { exact: true }),
+  ).toBeVisible({
+    timeout: 30_000,
+  });
+}
+
+export async function createWorkspaceThroughDaemon(
+  client: WorkspaceSetupDaemonClient,
+  input: { cwd: string; worktreeSlug: string },
+): Promise<{ id: string; name: string }> {
+  const result = await client.createPaseoWorktree(input);
+  if (!result.workspace || result.error) {
+    throw new Error(result.error ?? `Failed to create workspace for ${input.cwd}`);
+  }
+  return {
+    id: result.workspace.id,
+    name: result.workspace.name,
+  };
+}
+
+export async function findWorktreeWorkspaceForProject(
+  client: WorkspaceSetupDaemonClient,
+  repoPath: string,
+): Promise<{
+  id: string;
+  name: string;
+  projectRootPath: string;
+  workspaceDirectory: string;
+}> {
+  const payload = await client.fetchWorkspaces();
+  const normalizedRepoPath = realpathSync(repoPath);
+  const workspace =
+    payload.entries.find(
+      (entry) =>
+        entry.projectRootPath === normalizedRepoPath &&
+        entry.workspaceDirectory !== normalizedRepoPath,
+    ) ?? null;
+  if (!workspace) {
+    throw new Error(`Failed to find created worktree workspace for ${repoPath}`);
+  }
+  return {
+    id: workspace.id,
+    name: workspace.name,
+    projectRootPath: workspace.projectRootPath,
+    workspaceDirectory: workspace.workspaceDirectory,
+  };
+}
+
+export async function fetchWorkspaceById(
+  client: WorkspaceSetupDaemonClient,
+  workspaceId: string,
+): Promise<{
+  id: string;
+  name: string;
+  workspaceDirectory: string;
+  projectRootPath: string;
+}> {
+  const payload = await client.fetchWorkspaces();
+  const workspace = payload.entries.find((entry) => entry.id === workspaceId) ?? null;
+  if (!workspace) {
+    throw new Error(`Workspace not found: ${workspaceId}`);
+  }
+  return workspace;
+}
+
+export async function navigateToWorkspaceViaSidebar(
+  page: Page,
+  workspaceId: string,
+): Promise<void> {
+  const serverId = process.env.E2E_SERVER_ID;
+  if (!serverId) {
+    throw new Error("E2E_SERVER_ID is not set.");
+  }
+  await switchWorkspaceViaSidebar({ page, serverId, targetWorkspacePath: workspaceId });
+}
+
+export async function openWorkspaceScriptsMenu(page: Page): Promise<void> {
+  await page.getByTestId("workspace-scripts-button").click();
+  await expect(page.getByTestId("workspace-scripts-menu")).toBeVisible({ timeout: 10_000 });
+}
+
+export async function startWorkspaceScriptFromMenu(page: Page, scriptName: string): Promise<void> {
+  await page.getByTestId(`workspace-scripts-start-${scriptName}`).click();
+}
+
+export async function closeWorkspaceScriptsMenu(page: Page): Promise<void> {
+  await page.getByTestId("workspace-scripts-menu-backdrop").click();
+}
+
+export async function waitForWorkspaceSetupProgress(
+  client: WorkspaceSetupDaemonClient,
+  predicate: (payload: WorkspaceSetupProgressPayload) => boolean,
+  timeoutMs = 30_000,
+): Promise<WorkspaceSetupProgressPayload> {
+  return new Promise((resolve, reject) => {
+    const timeout = setTimeout(() => {
+      unsubscribe();
+      reject(new Error(`Timed out waiting for workspace_setup_progress after ${timeoutMs}ms`));
+    }, timeoutMs);
+
+    const unsubscribe = client.subscribeRawMessages((message) => {
+      if (message.type !== "workspace_setup_progress") {
+        return;
+      }
+      if (!predicate(message.payload)) {
+        return;
+      }
+      clearTimeout(timeout);
+      unsubscribe();
+      resolve(message.payload);
+    });
+  });
+}

packages/app/e2e/helpers/workspace-tabs.ts

@@ -13,15 +13,49 @@ export async function getWorkspaceTabTestIds(page: Page): Promise<string[]> {
   return ids;
 }
 
+function visibleTestId(page: Page, testId: string) {
+  return page.getByTestId(testId).filter({ visible: true });
+}
+
 export async function waitForWorkspaceTabsVisible(page: Page): Promise<void> {
-  await expect(page.getByTestId("workspace-tabs-row").first()).toBeVisible({
+  await expect(visibleTestId(page, "workspace-tabs-row").first()).toBeVisible({
     timeout: 30_000,
   });
-  await expect(page.getByTestId("workspace-new-agent-tab").first()).toBeVisible({
+  await expect(visibleTestId(page, "workspace-new-agent-tab").first()).toBeVisible({
     timeout: 30_000,
   });
 }
 
+export async function getVisibleWorkspaceAgentTabIds(page: Page): Promise<string[]> {
+  const tabs = page.locator('[data-testid^="workspace-tab-agent_"]').filter({ visible: true });
+  const count = await tabs.count();
+  const ids: string[] = [];
+  for (let index = 0; index < count; index += 1) {
+    const testId = await tabs.nth(index).getAttribute("data-testid");
+    if (testId && !ids.includes(testId)) {
+      ids.push(testId);
+    }
+  }
+  return ids;
+}
+
+export async function expectOnlyWorkspaceAgentTabsVisible(
+  page: Page,
+  expectedAgentIds: string[],
+): Promise<void> {
+  const expected = new Set(expectedAgentIds.map((id) => `workspace-tab-agent_${id}`));
+  const visible = await getVisibleWorkspaceAgentTabIds(page);
+  const unexpected = visible.filter((id) => !expected.has(id));
+
+  expect(unexpected).toEqual([]);
+  expect(visible.length).toBe(expected.size);
+  for (const expectedId of expectedAgentIds) {
+    await expect(visibleTestId(page, `workspace-tab-agent_${expectedId}`).first()).toBeVisible({
+      timeout: 30_000,
+    });
+  }
+}
+
 export async function ensureWorkspaceAgentPaneVisible(page: Page): Promise<void> {
   const toggle = page.getByTestId("workspace-explorer-toggle").first();
   if (!(await toggle.isVisible().catch(() => false))) {
@@ -36,6 +70,29 @@ export async function ensureWorkspaceAgentPaneVisible(page: Page): Promise<void>
   }
 }
 
+export async function expectWorkspaceTabsAbsent(page: Page): Promise<void> {
+  await expect(page.getByTestId("workspace-tabs-row")).toHaveCount(0);
+}
+
+export async function expectNoTerminalTabs(page: Page): Promise<void> {
+  await expect(page.locator('[data-testid^="workspace-tab-terminal_"]')).toHaveCount(0);
+}
+
+export async function clickFirstTerminalTab(
+  page: Page,
+  options?: { timeout?: number },
+): Promise<void> {
+  const tab = page.locator('[data-testid^="workspace-tab-terminal_"]').first();
+  await expect(tab).toBeVisible({ timeout: options?.timeout ?? 30_000 });
+  await tab.click();
+}
+
+export async function expectFirstTerminalTabContains(page: Page, text: string): Promise<void> {
+  await expect(page.locator('[data-testid^="workspace-tab-terminal_"]').first()).toContainText(
+    text,
+  );
+}
+
 export async function sampleWorkspaceTabIds(
   page: Page,
   options: { durationMs?: number; intervalMs?: number } = {},

packages/app/e2e/helpers/workspace-ui.ts

@@ -6,6 +6,17 @@ export async function openNewAgentComposer(page: Page): Promise<void> {
   await gotoHome(page);
 }
 
+/**
+ * Wait for the sidebar to show at least one project row, indicating that the
+ * WebSocket connection is up and workspace hydration has completed.
+ */
+export async function waitForSidebarHydration(page: Page, timeout = 60_000): Promise<void> {
+  await page
+    .locator('[data-testid^="sidebar-project-row-"]')
+    .first()
+    .waitFor({ state: "visible", timeout });
+}
+
 export function workspaceLabelFromPath(value: string): string {
   const normalized = value.replace(/\\/g, "/").replace(/\/+$/, "");
   const parts = normalized.split("/").filter(Boolean);
@@ -35,6 +46,29 @@ function workspaceRowLocator(page: Page, serverId: string, workspacePath: string
   return page.locator(ids.join(",")).first();
 }
 
+export async function expectSidebarWorkspaceSelected(input: {
+  page: Page;
+  serverId: string;
+  workspaceId: string;
+  selected?: boolean;
+}): Promise<void> {
+  const row = workspaceRowLocator(input.page, input.serverId, input.workspaceId);
+  await expect(row).toBeVisible({ timeout: 30_000 });
+  const expected = input.selected === false ? "false" : "true";
+
+  const hasDataSelected = await row.getAttribute("data-selected");
+  if (hasDataSelected !== null) {
+    await expect(row).toHaveAttribute("data-selected", expected, {
+      timeout: 30_000,
+    });
+    return;
+  }
+
+  await expect(row).toHaveAttribute("aria-selected", expected, {
+    timeout: 30_000,
+  });
+}
+
 export async function switchWorkspaceViaSidebar(input: {
   page: Page;
   serverId: string;
@@ -50,12 +84,27 @@ export async function switchWorkspaceViaSidebar(input: {
   });
 }
 
+/**
+ * Wait for a workspace's sidebar row to appear, confirming the workspace
+ * descriptor has been hydrated into the session store.
+ */
+export async function waitForWorkspaceInSidebar(
+  page: Page,
+  input: { serverId: string; workspaceId: string },
+): Promise<void> {
+  const candidates = candidateWorkspaceIds(input.workspaceId);
+  const selector = candidates
+    .map((id) => `[data-testid="sidebar-workspace-row-${input.serverId}:${id}"]`)
+    .join(",");
+  await page.locator(selector).first().waitFor({ state: "visible", timeout: 60_000 });
+}
+
 export async function expectWorkspaceHeader(
   page: Page,
   input: { title: string; subtitle: string },
 ): Promise<void> {
-  const titleLocator = page.getByTestId("workspace-header-title");
-  const subtitleLocator = page.getByTestId("workspace-header-subtitle");
+  const titleLocator = page.getByTestId("workspace-header-title").filter({ visible: true });
+  const subtitleLocator = page.getByTestId("workspace-header-subtitle").filter({ visible: true });
 
   await expect(titleLocator.first()).toHaveText(input.title, {
     timeout: 30_000,
@@ -65,6 +114,49 @@ export async function expectWorkspaceHeader(
   });
 }
 
+export async function expectReconnectingToastVisible(
+  page: Page,
+  options?: { timeout?: number },
+): Promise<void> {
+  await expect(page.getByTestId("agent-reconnecting-toast")).toBeVisible({
+    timeout: options?.timeout ?? 30_000,
+  });
+}
+
+export async function expectReconnectingToastGone(
+  page: Page,
+  options?: { timeout?: number },
+): Promise<void> {
+  await expect(page.getByTestId("agent-reconnecting-toast")).toHaveCount(0, {
+    timeout: options?.timeout ?? 30_000,
+  });
+}
+
+export async function expectHostConnectingOrOffline(
+  page: Page,
+  options?: { timeout?: number },
+): Promise<void> {
+  await expect(
+    page.getByText(/^Connecting$|localhost is offline|Cannot reach localhost/i),
+  ).toBeVisible({ timeout: options?.timeout ?? 30_000 });
+}
+
+export async function expectMenuButtonVisible(page: Page): Promise<void> {
+  await expect(page.getByTestId("menu-button")).toBeVisible();
+}
+
+export async function expectWorkspaceHeaderAbsent(page: Page): Promise<void> {
+  await expect(page.getByTestId("workspace-header-title")).toHaveCount(0);
+}
+
+export function workspaceDeckEntryLocator(page: Page, serverId: string, workspaceId: string) {
+  return page.getByTestId(`workspace-deck-entry-${serverId}:${workspaceId}`);
+}
+
+export async function expectWorkspaceDeckEntryCount(page: Page, count: number): Promise<void> {
+  await expect(page.locator('[data-testid^="workspace-deck-entry-"]')).toHaveCount(count);
+}
+
 export async function seedWorkspaceActivity(page: Page, marker: string): Promise<void> {
   const input = page.getByRole("textbox", { name: "Message agent..." });
   await expect(input).toBeEditable({ timeout: 30_000 });

packages/app/e2e/helpers/workspace.ts

@@ -1,19 +1,51 @@
 import { execSync } from "node:child_process";
-import { mkdtemp, writeFile, rm, mkdir } from "node:fs/promises";
+import { mkdtemp, writeFile, rm, mkdir, realpath } from "node:fs/promises";
 import { tmpdir } from "node:os";
 import path from "node:path";
 
-type TempRepo = {
+interface TempRepo {
   path: string;
+  branchHeads: Record<string, string>;
   cleanup: () => Promise<void>;
-};
+}
+
+async function configureRemote(input: {
+  repoPath: string;
+  withRemote: boolean;
+  originUrl: string | undefined;
+}): Promise<void> {
+  const { repoPath, withRemote, originUrl } = input;
+  if (withRemote) {
+    // Deterministic local remote to avoid relying on external auth/network in e2e.
+    const remoteDir = path.join(repoPath, "remote.git");
+    await mkdir(remoteDir, { recursive: true });
+    execSync(`git init --bare -b main ${remoteDir}`, { cwd: repoPath, stdio: "ignore" });
+    execSync(`git remote add origin ${remoteDir}`, { cwd: repoPath, stdio: "ignore" });
+    execSync("git push -u origin --all", { cwd: repoPath, stdio: "ignore" });
+    return;
+  }
+  if (originUrl) {
+    // Daemon reads origin for project grouping; no fetch occurs, so a synthetic URL is fine.
+    execSync(`git remote add origin ${JSON.stringify(originUrl)}`, {
+      cwd: repoPath,
+      stdio: "ignore",
+    });
+  }
+}
 
 export const createTempGitRepo = async (
   prefix = "paseo-e2e-",
-  options?: { withRemote?: boolean },
+  options?: {
+    withRemote?: boolean;
+    originUrl?: string;
+    paseoConfig?: Record<string, unknown>;
+    files?: Array<{ path: string; content: string }>;
+    branches?: string[];
+  },
 ): Promise<TempRepo> => {
   // Keep E2E repo paths short so terminal prompt + typed commands stay visible without zsh clipping.
-  const tempRoot = process.platform === "win32" ? tmpdir() : "/tmp";
+  // Resolve symlinks (macOS: /tmp → /private/tmp) so paths match the daemon's resolved paths.
+  const tempRoot = process.platform === "win32" ? tmpdir() : await realpath("/tmp");
   const repoPath = await mkdtemp(path.join(tempRoot, prefix));
   const withRemote = options?.withRemote ?? false;
 
@@ -22,22 +54,85 @@ export const createTempGitRepo = async (
   execSync('git config user.name "Paseo E2E"', { cwd: repoPath, stdio: "ignore" });
   execSync("git config commit.gpgsign false", { cwd: repoPath, stdio: "ignore" });
   await writeFile(path.join(repoPath, "README.md"), "# Temp Repo\n");
+  if (options?.paseoConfig) {
+    await writeFile(
+      path.join(repoPath, "paseo.json"),
+      JSON.stringify(options.paseoConfig, null, 2),
+    );
+  }
+  for (const file of options?.files ?? []) {
+    const filePath = path.join(repoPath, file.path);
+    await mkdir(path.dirname(filePath), { recursive: true });
+    await writeFile(filePath, file.content);
+  }
   execSync("git add README.md", { cwd: repoPath, stdio: "ignore" });
+  if (options?.paseoConfig) {
+    execSync("git add paseo.json", { cwd: repoPath, stdio: "ignore" });
+  }
+  for (const file of options?.files ?? []) {
+    execSync(`git add ${JSON.stringify(file.path)}`, { cwd: repoPath, stdio: "ignore" });
+  }
   execSync('git commit -m "Initial commit"', { cwd: repoPath, stdio: "ignore" });
 
-  if (withRemote) {
-    // Deterministic local remote to avoid relying on external auth/network in e2e.
-    const remoteDir = path.join(repoPath, "remote.git");
-    await mkdir(remoteDir, { recursive: true });
-    execSync(`git init --bare -b main ${remoteDir}`, { cwd: repoPath, stdio: "ignore" });
-    execSync(`git remote add origin ${remoteDir}`, { cwd: repoPath, stdio: "ignore" });
-    execSync("git push -u origin main", { cwd: repoPath, stdio: "ignore" });
+  const branchHeads: Record<string, string> = {};
+  const branches = Array.from(new Set(options?.branches ?? []));
+  for (const branch of branches) {
+    if (branch !== "main") {
+      execSync(`git checkout -b ${JSON.stringify(branch)} main`, {
+        cwd: repoPath,
+        stdio: "ignore",
+      });
+    }
+    const markerPath = `.paseo-e2e-${branch.replace(/[^a-zA-Z0-9._-]/g, "-")}.txt`;
+    await writeFile(path.join(repoPath, markerPath), `branch ${branch}\n`);
+    execSync(`git add ${JSON.stringify(markerPath)}`, { cwd: repoPath, stdio: "ignore" });
+    execSync(`git commit -m ${JSON.stringify(`Add ${branch} marker`)}`, {
+      cwd: repoPath,
+      stdio: "ignore",
+    });
+    branchHeads[branch] = execSync(`git rev-parse ${JSON.stringify(branch)}`, {
+      cwd: repoPath,
+      stdio: "pipe",
+    })
+      .toString()
+      .trim();
+    execSync("git checkout main", { cwd: repoPath, stdio: "ignore" });
   }
 
+  await configureRemote({ repoPath, withRemote, originUrl: options?.originUrl });
+
   return {
     path: repoPath,
+    branchHeads,
     cleanup: async () => {
       await rm(repoPath, { recursive: true, force: true });
     },
   };
 };
+
+export async function readWorktreeBranchInfo({ worktreePath }: { worktreePath: string }): Promise<{
+  currentBranch: string;
+  hasAncestor: (ref: string) => boolean;
+}> {
+  const currentBranch = execSync("git branch --show-current", {
+    cwd: worktreePath,
+    stdio: "pipe",
+  })
+    .toString()
+    .trim();
+
+  return {
+    currentBranch,
+    hasAncestor: (ref: string) => {
+      try {
+        execSync(`git merge-base --is-ancestor ${JSON.stringify(ref)} HEAD`, {
+          cwd: worktreePath,
+          stdio: "ignore",
+        });
+        return true;
+      } catch {
+        return false;
+      }
+    },
+  };
+}

packages/app/e2e/launcher-tab.spec.ts

@@ -0,0 +1,254 @@
+import { test, expect } from "./fixtures";
+import { createTempGitRepo } from "./helpers/workspace";
+import {
+  gotoWorkspace,
+  assertNewChatTileVisible,
+  assertTerminalTileVisible,
+  assertSingleNewTabButton,
+  pressNewTabShortcut,
+  clickNewChat,
+  clickNewTerminal,
+  countTabsOfKind,
+  getTabTestIds,
+  waitForTabWithTitle,
+  measureTileTransition,
+  sampleTabsDuringTransition,
+  terminalSurfaceLocator,
+} from "./helpers/launcher";
+import { expectComposerVisible, composerLocator } from "./helpers/composer";
+import { expectTerminalSurfaceVisible } from "./helpers/terminal-perf";
+import {
+  connectTerminalClient,
+  setupDeterministicPrompt,
+  type TerminalPerfDaemonClient,
+} from "./helpers/terminal-perf";
+
+// ─── Shared state ──────────────────────────────────────────────────────────
+
+let tempRepo: { path: string; cleanup: () => Promise<void> };
+let workspaceId: string;
+let seedClient: TerminalPerfDaemonClient;
+
+test.beforeAll(async () => {
+  tempRepo = await createTempGitRepo("launcher-e2e-");
+  seedClient = await connectTerminalClient();
+  const result = await seedClient.openProject(tempRepo.path);
+  if (!result.workspace) throw new Error(result.error ?? "Failed to seed workspace");
+  workspaceId = result.workspace.id;
+});
+
+test.afterAll(async () => {
+  if (seedClient) await seedClient.close();
+  if (tempRepo) await tempRepo.cleanup();
+});
+
+// ═══════════════════════════════════════════════════════════════════════════
+// Tab Creation Tests
+// ═══════════════════════════════════════════════════════════════════════════
+
+test.describe("Tab creation", () => {
+  test("Cmd+T opens a new agent tab with composer", async ({ page }) => {
+    await gotoWorkspace(page, workspaceId);
+
+    await pressNewTabShortcut(page);
+
+    await expectComposerVisible(page);
+  });
+
+  test("opening two new tabs creates two draft tabs", async ({ page }) => {
+    await gotoWorkspace(page, workspaceId);
+
+    const countBefore = await countTabsOfKind(page, "draft");
+
+    await pressNewTabShortcut(page);
+    await expect
+      .poll(() => countTabsOfKind(page, "draft"), { timeout: 15_000 })
+      .toBe(countBefore + 1);
+    const countAfterFirst = await countTabsOfKind(page, "draft");
+
+    // Blur the composer so the second shortcut isn't swallowed by the focused input
+    await page.evaluate(() => (document.activeElement as HTMLElement)?.blur?.());
+    await pressNewTabShortcut(page);
+    await expect
+      .poll(() => countTabsOfKind(page, "draft"), { timeout: 15_000 })
+      .toBe(countAfterFirst + 1);
+  });
+
+  test("clicking new agent tab creates a draft tab", async ({ page }) => {
+    await gotoWorkspace(page, workspaceId);
+
+    await clickNewChat(page);
+
+    await expectComposerVisible(page);
+
+    const tabsAfter = await getTabTestIds(page);
+    const draftCountAfter = tabsAfter.filter((id) => id.includes("draft")).length;
+    expect(draftCountAfter).toBeGreaterThanOrEqual(1);
+  });
+
+  test("clicking terminal button creates a standalone terminal", async ({ page }) => {
+    test.setTimeout(45_000);
+    await gotoWorkspace(page, workspaceId);
+
+    await clickNewTerminal(page);
+
+    await expectTerminalSurfaceVisible(page);
+
+    const tabsAfter = await getTabTestIds(page);
+    const terminalTabs = tabsAfter.filter((id) => id.includes("terminal"));
+    expect(terminalTabs.length).toBeGreaterThanOrEqual(1);
+  });
+
+  test("tab bar shows action buttons per pane", async ({ page }) => {
+    await gotoWorkspace(page, workspaceId);
+    await assertSingleNewTabButton(page);
+    await assertNewChatTileVisible(page);
+    await assertTerminalTileVisible(page);
+  });
+});
+
+// ═══════════════════════════════════════════════════════════════════════════
+// Terminal Title Tests
+// ═══════════════════════════════════════════════════════════════════════════
+
+test.describe("Terminal title propagation", () => {
+  // OSC title escape sequence propagation is inherently flaky — the terminal
+  // must process the sequence, emit a title change event, and the tab bar
+  // must re-render before the assertion deadline. Allow retries.
+  test.describe.configure({ retries: 2 });
+
+  let client: TerminalPerfDaemonClient;
+
+  test.beforeAll(async () => {
+    client = await connectTerminalClient();
+  });
+
+  test.afterAll(async () => {
+    if (client) await client.close();
+  });
+
+  test.skip("terminal tab title updates from OSC title escape sequence", async ({ page }) => {
+    test.setTimeout(60_000);
+
+    const result = await client.createTerminal(tempRepo.path, "title-test");
+    if (!result.terminal) throw new Error(`Failed to create terminal: ${result.error}`);
+    const terminalId = result.terminal.id;
+
+    try {
+      // Navigate to workspace and open a terminal
+      await gotoWorkspace(page, workspaceId);
+      await clickNewTerminal(page);
+
+      await expectTerminalSurfaceVisible(page);
+      await terminalSurfaceLocator(page).click();
+
+      await setupDeterministicPrompt(page);
+
+      // Send OSC 0 (set window title) escape sequence
+      const testTitle = `E2E-Title-${Date.now()}`;
+      await terminalSurfaceLocator(page).pressSequentially(`printf '\\033]0;${testTitle}\\007'\n`, {
+        delay: 0,
+      });
+
+      // Wait for the tab to reflect the new title
+      await waitForTabWithTitle(page, testTitle, 15_000);
+    } finally {
+      await client.killTerminal(terminalId).catch(() => {});
+    }
+  });
+
+  test.skip("title debouncing coalesces rapid changes", async ({ page }) => {
+    test.setTimeout(60_000);
+
+    const result = await client.createTerminal(tempRepo.path, "debounce-test");
+    if (!result.terminal) throw new Error(`Failed to create terminal: ${result.error}`);
+    const terminalId = result.terminal.id;
+
+    try {
+      await gotoWorkspace(page, workspaceId);
+      await clickNewTerminal(page);
+
+      await expectTerminalSurfaceVisible(page);
+      await terminalSurfaceLocator(page).click();
+
+      await setupDeterministicPrompt(page);
+
+      // Fire many rapid title changes — only the last should stick
+      const finalTitle = `Final-${Date.now()}`;
+      for (let i = 0; i < 5; i++) {
+        await terminalSurfaceLocator(page).pressSequentially(`printf '\\033]0;Rapid-${i}\\007'\n`, {
+          delay: 0,
+        });
+      }
+      await terminalSurfaceLocator(page).pressSequentially(
+        `printf '\\033]0;${finalTitle}\\007'\n`,
+        { delay: 0 },
+      );
+
+      // The tab should eventually settle on the final title
+      await waitForTabWithTitle(page, finalTitle, 15_000);
+    } finally {
+      await client.killTerminal(terminalId).catch(() => {});
+    }
+  });
+});
+
+// ═══════════════════════════════════════════════════════════════════════════
+// No-Flash Transition Tests
+// ═══════════════════════════════════════════════════════════════════════════
+
+test.describe("Tab transitions (no flash)", () => {
+  test("New agent tab transition has no blank intermediate tab state", async ({ page }) => {
+    await gotoWorkspace(page, workspaceId);
+
+    // Sample tabs at high frequency across the transition
+    const snapshots = await sampleTabsDuringTransition(page, () => clickNewChat(page), 2_000, 30);
+
+    // Every snapshot should have at least one tab — no blank/zero-tab frames
+    for (const snapshot of snapshots) {
+      expect(snapshot.length).toBeGreaterThanOrEqual(1);
+    }
+
+    // Tab count should never spike excessively (no duplicate flash from add-then-remove).
+    // When running in-suite, previous tests may have created tabs on the shared workspace,
+    // so we allow +2 tolerance for accumulated state and React render batching.
+    const counts = snapshots.map((s) => s.length);
+    const maxCount = Math.max(...counts);
+    const initialCount = counts[0] ?? 0;
+
+    expect(maxCount).toBeLessThanOrEqual(initialCount + 2);
+  });
+
+  test("Terminal transition completes within visual budget", async ({ page }) => {
+    test.setTimeout(30_000);
+    await gotoWorkspace(page, workspaceId);
+
+    const elapsed = await measureTileTransition(
+      page,
+      () => clickNewTerminal(page),
+      terminalSurfaceLocator(page),
+      20_000,
+    );
+
+    // Terminal surface should appear within a reasonable budget.
+    // Note: terminal creation involves a server round-trip, so we allow more time
+    // than a pure in-memory transition, but it should still be well under 5 seconds.
+    expect(elapsed).toBeLessThan(5_000);
+  });
+
+  test("New agent tab click shows composer without flash", async ({ page }) => {
+    await gotoWorkspace(page, workspaceId);
+
+    const elapsed = await measureTileTransition(
+      page,
+      () => clickNewChat(page),
+      composerLocator(page),
+      10_000,
+    );
+
+    // Draft creation is fully in-memory — should be fast
+    // We use a generous budget here because CI can be slow, but the key assertion
+    // is that no blank/flash frame appears (tested above).
+    expect(elapsed).toBeLessThan(3_000);
+  });
+});

packages/app/e2e/new-workspace.spec.ts

@@ -0,0 +1,495 @@
+import { existsSync } from "node:fs";
+import path from "node:path";
+import { buildHostWorkspaceRoute } from "@/utils/host-routes";
+import { expect, test } from "./fixtures";
+import { gotoAppShell } from "./helpers/app";
+import {
+  archiveWorkspaceFromDaemon,
+  archiveLocalWorkspaceFromDaemon,
+  assertNewWorkspaceSidebarAndHeader,
+  clickNewWorkspaceButton,
+  closeBranchPicker,
+  connectNewWorkspaceDaemonClient,
+  createWorktreeViaDaemon,
+  delayBrowserAgentCreatedStatus,
+  expectComposerGithubAttachmentPill,
+  expectPickerClosed,
+  expectPickerOpen,
+  expectPickerSelected,
+  expectStartingRefPickerTriggerPr,
+  openBranchPicker,
+  openNewWorkspaceComposer,
+  openProjectViaDaemon,
+  openStartingRefPicker,
+  selectBranchInPicker,
+  selectGitHubPrInPicker,
+  selectPickerOptionByKeyboard,
+} from "./helpers/new-workspace";
+import { createTempGitRepo, readWorktreeBranchInfo } from "./helpers/workspace";
+import {
+  expectSidebarWorkspaceSelected,
+  expectWorkspaceHeader,
+  switchWorkspaceViaSidebar,
+  waitForSidebarHydration,
+  waitForWorkspaceInSidebar,
+  workspaceLabelFromPath,
+} from "./helpers/workspace-ui";
+
+test.describe("New workspace flow", () => {
+  let client: Awaited<ReturnType<typeof connectNewWorkspaceDaemonClient>>;
+  const localWorkspaceIds = new Set<string>();
+  const createdWorktreeIds = new Set<string>();
+
+  test.describe.configure({ timeout: 240_000 });
+
+  test.beforeEach(async () => {
+    client = await connectNewWorkspaceDaemonClient();
+  });
+
+  test.afterEach(async () => {
+    if (client) {
+      for (const workspaceId of createdWorktreeIds) {
+        await archiveWorkspaceFromDaemon(client, workspaceId).catch(() => undefined);
+      }
+      for (const workspaceId of localWorkspaceIds) {
+        await archiveLocalWorkspaceFromDaemon(client, workspaceId).catch(() => undefined);
+      }
+    }
+    createdWorktreeIds.clear();
+    localWorkspaceIds.clear();
+    await client?.close().catch(() => undefined);
+  });
+
+  test("sidebar workspace navigation updates URL and header", async ({ page }) => {
+    const serverId = process.env.E2E_SERVER_ID;
+    if (!serverId) {
+      throw new Error("E2E_SERVER_ID is not set.");
+    }
+
+    const firstRepo = await createTempGitRepo("workspace-nav-a-");
+    const secondRepo = await createTempGitRepo("workspace-nav-b-");
+
+    try {
+      const firstWorkspace = await openProjectViaDaemon(client, firstRepo.path);
+      const secondWorkspace = await openProjectViaDaemon(client, secondRepo.path);
+      localWorkspaceIds.add(firstWorkspace.workspaceId);
+      localWorkspaceIds.add(secondWorkspace.workspaceId);
+
+      await gotoAppShell(page);
+      await waitForSidebarHydration(page);
+
+      await switchWorkspaceViaSidebar({
+        page,
+        serverId,
+        targetWorkspacePath: firstWorkspace.workspaceId,
+      });
+      await expectWorkspaceHeader(page, {
+        title: firstWorkspace.workspaceName,
+        subtitle: firstWorkspace.projectDisplayName,
+      });
+
+      await switchWorkspaceViaSidebar({
+        page,
+        serverId,
+        targetWorkspacePath: secondWorkspace.workspaceId,
+      });
+      await waitForWorkspaceInSidebar(page, {
+        serverId,
+        workspaceId: secondWorkspace.workspaceId,
+      });
+      await expectWorkspaceHeader(page, {
+        title: secondWorkspace.workspaceName,
+        subtitle: secondWorkspace.projectDisplayName,
+      });
+
+      await switchWorkspaceViaSidebar({
+        page,
+        serverId,
+        targetWorkspacePath: firstWorkspace.workspaceId,
+      });
+      await expectWorkspaceHeader(page, {
+        title: firstWorkspace.workspaceName,
+        subtitle: firstWorkspace.projectDisplayName,
+      });
+    } finally {
+      await secondRepo.cleanup();
+      await firstRepo.cleanup();
+    }
+  });
+
+  test("same-project workspaces switch content without requiring refresh", async ({ page }) => {
+    const serverId = process.env.E2E_SERVER_ID;
+    if (!serverId) {
+      throw new Error("E2E_SERVER_ID is not set.");
+    }
+
+    const repo = await createTempGitRepo("workspace-nav-same-project-");
+
+    try {
+      const rootWorkspace = await openProjectViaDaemon(client, repo.path);
+      const worktreeWorkspace = await createWorktreeViaDaemon(client, {
+        cwd: repo.path,
+        slug: `nav-${Date.now()}`,
+      });
+      localWorkspaceIds.add(rootWorkspace.workspaceId);
+      createdWorktreeIds.add(worktreeWorkspace.workspaceId);
+
+      await gotoAppShell(page);
+      await waitForSidebarHydration(page);
+
+      await switchWorkspaceViaSidebar({
+        page,
+        serverId,
+        targetWorkspacePath: rootWorkspace.workspaceId,
+      });
+      await expectWorkspaceHeader(page, {
+        title: rootWorkspace.workspaceName,
+        subtitle: rootWorkspace.projectDisplayName,
+      });
+      await expectSidebarWorkspaceSelected({
+        page,
+        serverId,
+        workspaceId: rootWorkspace.workspaceId,
+      });
+
+      await switchWorkspaceViaSidebar({
+        page,
+        serverId,
+        targetWorkspacePath: worktreeWorkspace.workspaceId,
+      });
+      await expectWorkspaceHeader(page, {
+        title: worktreeWorkspace.workspaceName,
+        subtitle: worktreeWorkspace.projectDisplayName,
+      });
+      await expectSidebarWorkspaceSelected({
+        page,
+        serverId,
+        workspaceId: worktreeWorkspace.workspaceId,
+      });
+      await expectSidebarWorkspaceSelected({
+        page,
+        serverId,
+        workspaceId: rootWorkspace.workspaceId,
+        selected: false,
+      });
+
+      await switchWorkspaceViaSidebar({
+        page,
+        serverId,
+        targetWorkspacePath: rootWorkspace.workspaceId,
+      });
+      await expectWorkspaceHeader(page, {
+        title: rootWorkspace.workspaceName,
+        subtitle: rootWorkspace.projectDisplayName,
+      });
+      await expectSidebarWorkspaceSelected({
+        page,
+        serverId,
+        workspaceId: rootWorkspace.workspaceId,
+      });
+      await expectSidebarWorkspaceSelected({
+        page,
+        serverId,
+        workspaceId: worktreeWorkspace.workspaceId,
+        selected: false,
+      });
+    } finally {
+      await repo.cleanup();
+    }
+  });
+
+  test("clicking new workspace redirects, renders header, shows sidebar row, and keeps one agent tab", async ({
+    page,
+  }) => {
+    const serverId = process.env.E2E_SERVER_ID;
+    if (!serverId) {
+      throw new Error("E2E_SERVER_ID is not set.");
+    }
+
+    const tempRepo = await createTempGitRepo("new-workspace-");
+
+    try {
+      const openedProject = await openProjectViaDaemon(client, tempRepo.path);
+      localWorkspaceIds.add(openedProject.workspaceId);
+
+      await gotoAppShell(page);
+      await waitForSidebarHydration(page);
+
+      await switchWorkspaceViaSidebar({
+        page,
+        serverId,
+        targetWorkspacePath: openedProject.workspaceId,
+      });
+      await expectWorkspaceHeader(page, {
+        title: openedProject.workspaceName,
+        subtitle: openedProject.projectDisplayName,
+      });
+
+      await clickNewWorkspaceButton(page, {
+        projectKey: openedProject.projectKey,
+        projectDisplayName: openedProject.projectDisplayName,
+      });
+
+      const createdWorkspace = await assertNewWorkspaceSidebarAndHeader(page, {
+        serverId,
+        previousWorkspaceId: openedProject.workspaceId,
+        projectDisplayName: openedProject.projectDisplayName,
+      });
+      createdWorktreeIds.add(createdWorkspace.workspaceId);
+
+      expect(createdWorkspace.workspaceId).not.toBe(openedProject.workspaceId);
+      await expect(page).toHaveURL(
+        buildHostWorkspaceRoute(serverId, createdWorkspace.workspaceId),
+        {
+          timeout: 30_000,
+        },
+      );
+
+      const createdWorkspaceRow = page.getByTestId(
+        `sidebar-workspace-row-${serverId}:${createdWorkspace.workspaceId}`,
+      );
+      await expect(createdWorkspaceRow).toBeVisible({ timeout: 30_000 });
+
+      await expectWorkspaceHeader(page, {
+        title: workspaceLabelFromPath(createdWorkspace.workspaceId),
+        subtitle: openedProject.projectDisplayName,
+      });
+
+      const activeWorkspaceDeckEntry = page
+        .getByTestId(`workspace-deck-entry-${serverId}:${createdWorkspace.workspaceId}`)
+        .filter({ visible: true });
+      await expect(activeWorkspaceDeckEntry).toBeVisible({ timeout: 30_000 });
+
+      const agentTabs = activeWorkspaceDeckEntry.locator('[data-testid^="workspace-tab-agent_"]');
+      await expect(agentTabs).toHaveCount(1, { timeout: 30_000 });
+
+      // Workspace setup may auto-open a setup tab that steals focus,
+      // hiding the agent panel (display:none removes it from the
+      // accessibility tree). Click the agent tab to ensure it's active.
+      await agentTabs.first().click();
+
+      const composer = page.getByRole("textbox", { name: "Message agent..." });
+      await expect(composer).toBeVisible({ timeout: 30_000 });
+    } finally {
+      await tempRepo.cleanup();
+    }
+  });
+
+  test("redirects to the optimistic draft tab before agent creation resolves", async ({ page }) => {
+    const serverId = process.env.E2E_SERVER_ID;
+    if (!serverId) {
+      throw new Error("E2E_SERVER_ID is not set.");
+    }
+
+    const tempRepo = await createTempGitRepo("new-workspace-optimistic-");
+    const agentCreatedDelay = await delayBrowserAgentCreatedStatus(page);
+
+    try {
+      const openedProject = await openProjectViaDaemon(client, tempRepo.path);
+      localWorkspaceIds.add(openedProject.workspaceId);
+
+      await gotoAppShell(page);
+      await waitForSidebarHydration(page);
+
+      await switchWorkspaceViaSidebar({
+        page,
+        serverId,
+        targetWorkspacePath: openedProject.workspaceId,
+      });
+      await expectWorkspaceHeader(page, {
+        title: openedProject.workspaceName,
+        subtitle: openedProject.projectDisplayName,
+      });
+
+      await openNewWorkspaceComposer(page, {
+        projectKey: openedProject.projectKey,
+        projectDisplayName: openedProject.projectDisplayName,
+      });
+
+      const composer = page.getByRole("textbox", { name: "Message agent..." });
+      await expect(composer).toBeVisible({ timeout: 30_000 });
+      await composer.fill("Hello from e2e");
+
+      const createButton = page
+        .getByTestId("message-input-root")
+        .getByRole("button", { name: "Create" });
+      await expect(createButton).toBeVisible({ timeout: 30_000 });
+      await createButton.click();
+
+      await agentCreatedDelay.waitForCreateRequest();
+      await agentCreatedDelay.waitForDelayedCreatedStatus();
+
+      const createdWorkspace = await assertNewWorkspaceSidebarAndHeader(page, {
+        serverId,
+        previousWorkspaceId: openedProject.workspaceId,
+        projectDisplayName: openedProject.projectDisplayName,
+      });
+      createdWorktreeIds.add(createdWorkspace.workspaceId);
+
+      await expect(page).toHaveURL(
+        buildHostWorkspaceRoute(serverId, createdWorkspace.workspaceId),
+        {
+          timeout: 30_000,
+        },
+      );
+
+      const activeWorkspaceDeckEntry = page
+        .getByTestId(`workspace-deck-entry-${serverId}:${createdWorkspace.workspaceId}`)
+        .filter({ visible: true });
+      await expect(activeWorkspaceDeckEntry).toBeVisible({ timeout: 30_000 });
+
+      const draftTabs = activeWorkspaceDeckEntry.locator('[data-testid^="workspace-tab-draft_"]');
+      await expect(draftTabs).toHaveCount(1, { timeout: 30_000 });
+      await expect(
+        activeWorkspaceDeckEntry.locator('[data-testid^="workspace-tab-agent_"]'),
+      ).toHaveCount(0);
+
+      agentCreatedDelay.release();
+      await expect(
+        activeWorkspaceDeckEntry.locator('[data-testid^="workspace-tab-agent_"]'),
+      ).toHaveCount(1, { timeout: 30_000 });
+    } finally {
+      agentCreatedDelay.release();
+      await tempRepo.cleanup();
+    }
+  });
+
+  test("selected branch becomes the base of a new workspace worktree", async ({ page }) => {
+    const serverId = process.env.E2E_SERVER_ID;
+    if (!serverId) {
+      throw new Error("E2E_SERVER_ID is not set.");
+    }
+
+    const tempRepo = await createTempGitRepo("new-workspace-ref-", {
+      branches: ["main", "dev"],
+    });
+
+    try {
+      const openedProject = await openProjectViaDaemon(client, tempRepo.path);
+      localWorkspaceIds.add(openedProject.workspaceId);
+
+      await gotoAppShell(page);
+      await waitForSidebarHydration(page);
+
+      await switchWorkspaceViaSidebar({
+        page,
+        serverId,
+        targetWorkspacePath: openedProject.workspaceId,
+      });
+      await expectWorkspaceHeader(page, {
+        title: openedProject.workspaceName,
+        subtitle: openedProject.projectDisplayName,
+      });
+
+      await openNewWorkspaceComposer(page, {
+        projectKey: openedProject.projectKey,
+        projectDisplayName: openedProject.projectDisplayName,
+      });
+      await openStartingRefPicker(page);
+      await selectBranchInPicker(page, "dev");
+
+      const createButton = page
+        .getByTestId("message-input-root")
+        .getByRole("button", { name: "Create" });
+      await expect(createButton).toBeVisible({ timeout: 30_000 });
+      await createButton.click();
+
+      const createdWorkspace = await assertNewWorkspaceSidebarAndHeader(page, {
+        serverId,
+        previousWorkspaceId: openedProject.workspaceId,
+        projectDisplayName: openedProject.projectDisplayName,
+      });
+      createdWorktreeIds.add(createdWorkspace.workspaceId);
+
+      expect(existsSync(createdWorkspace.workspaceId)).toBe(true);
+
+      const branchInfo = await readWorktreeBranchInfo({
+        worktreePath: createdWorkspace.workspaceId,
+      });
+      expect(branchInfo.currentBranch).toBe(path.basename(createdWorkspace.workspaceId));
+      expect(branchInfo.hasAncestor(tempRepo.branchHeads.main)).toBe(true);
+      expect(branchInfo.hasAncestor(tempRepo.branchHeads.dev)).toBe(true);
+    } finally {
+      await tempRepo.cleanup();
+    }
+  });
+
+  test("branch picker opens via keyboard, navigates options, and selects on Enter", async ({
+    page,
+  }) => {
+    const tempRepo = await createTempGitRepo("picker-keyboard-", { branches: ["main", "dev"] });
+
+    try {
+      const openedProject = await openProjectViaDaemon(client, tempRepo.path);
+      localWorkspaceIds.add(openedProject.workspaceId);
+
+      await gotoAppShell(page);
+      await waitForSidebarHydration(page);
+      await openNewWorkspaceComposer(page, {
+        projectKey: openedProject.projectKey,
+        projectDisplayName: openedProject.projectDisplayName,
+      });
+
+      await openBranchPicker(page);
+      await expectPickerOpen(page);
+      await selectPickerOptionByKeyboard(page, "dev");
+      await expectPickerSelected(page, "dev");
+      await expectPickerClosed(page);
+    } finally {
+      await tempRepo.cleanup();
+    }
+  });
+
+  test("branch picker closes on Escape without selecting an option", async ({ page }) => {
+    const tempRepo = await createTempGitRepo("picker-escape-");
+
+    try {
+      const openedProject = await openProjectViaDaemon(client, tempRepo.path);
+      localWorkspaceIds.add(openedProject.workspaceId);
+
+      await gotoAppShell(page);
+      await waitForSidebarHydration(page);
+      await openNewWorkspaceComposer(page, {
+        projectKey: openedProject.projectKey,
+        projectDisplayName: openedProject.projectDisplayName,
+      });
+
+      await openBranchPicker(page);
+      await expectPickerOpen(page);
+      await closeBranchPicker(page);
+      await expectPickerClosed(page);
+    } finally {
+      await tempRepo.cleanup();
+    }
+  });
+
+  test("selected GitHub PR shows PR context in the trigger and composer", async ({ page }) => {
+    const tempRepo = await createTempGitRepo("new-workspace-pr-ref-");
+
+    try {
+      const openedProject = await openProjectViaDaemon(client, tempRepo.path);
+      localWorkspaceIds.add(openedProject.workspaceId);
+
+      await gotoAppShell(page);
+      await waitForSidebarHydration(page);
+      await openNewWorkspaceComposer(page, {
+        projectKey: openedProject.projectKey,
+        projectDisplayName: openedProject.projectDisplayName,
+      });
+      await openStartingRefPicker(page);
+      await selectGitHubPrInPicker(page, 515);
+
+      await expectStartingRefPickerTriggerPr(page, {
+        number: 515,
+        title: "Review selected start ref",
+        headRef: "feature/start-from-pr",
+      });
+      await expectComposerGithubAttachmentPill(page, {
+        number: 515,
+        title: "Review selected start ref",
+      });
+    } finally {
+      await tempRepo.cleanup();
+    }
+  });
+});