chore(release): cut 0.1.69

Translate Paseo's header-keyed AskUserQuestion answers into the full question-text keys Claude expects before resolving question permissions.

Also preserve the original question payload when UI callbacks return an answers-only updatedInput, and lock the behavior down with provider-flow regression coverage in claude-agent tests.

.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

@@ -4,6 +4,7 @@ on:
   push:
     tags:
       - "v*"
+      - "android-v*"
   workflow_dispatch:
     inputs:
       tag:
@@ -16,7 +17,7 @@ concurrency:
   cancel-in-progress: false
 
 env:
-  RELEASE_TAG: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.tag || github.ref_name }}
+  SOURCE_TAG: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.tag || github.ref_name }}
 
 jobs:
   publish-android-apk:
@@ -31,6 +32,36 @@ jobs:
           fetch-depth: 0
           ref: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.tag || github.ref }}
 
+      - 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: ${{ secrets.GITHUB_TOKEN }}
+        shell: bash
+        run: |
+          set -euo pipefail
+
+          if gh release view "$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 "${release_args[@]}"; then
+            echo "Release creation raced with another workflow; continuing."
+          fi
+
       - name: Setup Node
         uses: actions/setup-node@v4
         with:
@@ -94,24 +125,6 @@ jobs:
           echo "asset_name=$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 +1,278 @@
+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:
+    runs-on: ubuntu-latest
+    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: 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 }}
+
+  server-tests-windows:
+    runs-on: windows-latest
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - 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: Run Windows-critical server tests
+        working-directory: packages/server
+        run: >
+          npx vitest run
+          src/utils/executable.test.ts
+          src/utils/spawn.launch-regression.test.ts
+          src/utils/spawn.percent-escape.test.ts
+          src/utils/spawn.test.ts
+          src/utils/process-tree.test.ts
+          src/utils/run-git-command.test.ts
+          src/utils/checkout-git-rev-parse.test.ts
+          src/terminal/worker-terminal-manager.test.ts
+          src/server/agent/provider-registry.test.ts
+          src/server/agent/provider-launch-config.test.ts
+          src/server/agent/provider-snapshot-manager.test.ts
+          src/server/agent/providers/claude-agent.spawn.test.ts
+          src/server/agent/providers/provider-windows-launch.test.ts
+          src/server/agent/providers/provider-availability.test.ts
+          src/server/workspace-registry-model.test.ts
+          src/server/persisted-config.test.ts
+          src/server/bootstrap-provider-availability.test.ts
+
+  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 @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:
+    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 agent CLIs for provider tests
+        run: npm install -g @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"

.github/workflows/deploy-app.yml

@@ -3,7 +3,10 @@ name: Deploy App
 on:
   push:
     tags:
-      - 'v*'
+      - "v*"
+      - "!v*-beta.*"
+      - "app-v*"
+      - "!app-v*-beta.*"
   workflow_dispatch:
 
 jobs:
@@ -15,16 +18,19 @@ 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 }}
 
+      - name: Build highlight dependency
+        run: npm run build --workspace=@getpaseo/highlight
+
       - name: Typecheck
         run: npm run typecheck --workspace=@getpaseo/app
 

.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

@@ -4,15 +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]
   workflow_dispatch:
 
 jobs:
   deploy:
+    if: ${{ github.event_name != 'release' || (!github.event.release.prerelease && !github.event.release.draft) }}
     runs-on: ubuntu-latest
 
     steps:
@@ -20,8 +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

@@ -24,6 +24,24 @@ on:
           - 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 }}
@@ -31,18 +49,59 @@ concurrency:
 
 env:
   SOURCE_TAG: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.tag || github.ref_name }}
+  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:
         include:
           - runner: macos-14
-            rust_target: aarch64-apple-darwin
+            electron_arch: arm64
           - runner: macos-15-intel
-            rust_target: x86_64-apple-darwin
+            electron_arch: x64
     permissions:
       contents: write
       packages: read
@@ -52,62 +111,11 @@ 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="${SOURCE_TAG}"
-          release_tag="$source_tag"
-          for prefix in desktop-windows-v desktop-linux-v desktop-macos-v desktop-v; do
-            if [[ "$source_tag" == ${prefix}* ]]; then
-              release_tag="v${source_tag#${prefix}}"
-              break
-            fi
-          done
-          echo "RELEASE_TAG=$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
-
-      - name: Set desktop version from tag
-        shell: bash
-        run: |
-          node <<'NODE'
-          const fs = require('node:fs');
-          const path = require('node:path');
-          const version = process.env.DESKTOP_VERSION;
-          if (!version) throw new Error('DESKTOP_VERSION env var is missing');
-          console.log(`Setting desktop version to ${version}`);
-
-          const tauriConfPath = path.join('packages', 'desktop', 'src-tauri', 'tauri.conf.json');
-          const tauriConfText = fs.readFileSync(tauriConfPath, 'utf8');
-          const tauriRe = /("version"\s*:\s*")([^"]+)(")/;
-          if (!tauriRe.test(tauriConfText)) throw new Error('Failed to find version in tauri.conf.json');
-          fs.writeFileSync(tauriConfPath, tauriConfText.replace(tauriRe, `$1${version}$3`));
-
-          const cargoTomlPath = path.join('packages', 'desktop', 'src-tauri', 'Cargo.toml');
-          const lines = fs.readFileSync(cargoTomlPath, 'utf8').split(/\r?\n/);
-          let inPackage = false, updated = false;
-          const result = lines.map((line) => {
-            if (/^\[package\]\s*$/.test(line)) inPackage = true;
-            else if (inPackage && /^\[/.test(line)) inPackage = false;
-            if (inPackage && /^version\s*=\s*".*"\s*$/.test(line)) {
-              updated = true;
-              return `version = "${version}"`;
-            }
-            return line;
-          });
-          if (!updated) throw new Error('Failed to update Cargo.toml version');
-          fs.writeFileSync(cargoTomlPath, result.join('\n') + '\n');
-          NODE
+        run: node scripts/emit-release-env.mjs --source-tag "$SOURCE_TAG" >> "$GITHUB_ENV"
 
       - name: Setup Node
         uses: actions/setup-node@v4
@@ -118,156 +126,88 @@ jobs:
           registry-url: "https://npm.pkg.github.com"
           scope: "@boudra"
 
-      - name: Install Rust stable
-        uses: dtolnay/rust-toolchain@stable
-        with:
-          targets: ${{ matrix.rust_target }}
-
-      - name: Restore Rust cache
-        uses: Swatinem/rust-cache@v2
-        with:
-          shared-key: desktop-release-macos-${{ matrix.rust_target }}
-          workspaces: |
-            .
-            packages/desktop/src-tauri -> target
-
       - name: Install JS dependencies
         run: npm ci
         env:
           NODE_AUTH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
 
-      - name: Build web app for Tauri
-        run: npm run build:web --workspace=@getpaseo/app
+      - name: Set desktop package version from tag
+        shell: bash
+        run: |
+          node <<'NODE'
+          const fs = require('node:fs');
+          const path = require('node:path');
 
-      - name: Build managed runtime
-        run: npm run prepare:managed-runtime --workspace=@getpaseo/desktop
+          const version = process.env.DESKTOP_VERSION;
+          if (!version) throw new Error('DESKTOP_VERSION env var is missing');
 
-      - name: Validate managed runtime bundle
-        run: npm run validate:managed-runtime --workspace=@getpaseo/desktop
+          const packageJsonPath = path.join(process.env.DESKTOP_PACKAGE_PATH, 'package.json');
+          const packageJson = JSON.parse(fs.readFileSync(packageJsonPath, 'utf8'));
+          packageJson.version = version;
+          fs.writeFileSync(packageJsonPath, `${JSON.stringify(packageJson, null, 2)}\n`);
+          NODE
 
-      - name: Import Apple code-signing certificate
-        uses: apple-actions/import-codesign-certs@v3
-        with:
-          p12-file-base64: ${{ secrets.APPLE_CERTIFICATE }}
-          p12-password: ${{ secrets.APPLE_CERTIFICATE_PASSWORD }}
-
-      - name: Sign bundled managed runtime
+      - name: Build desktop release
+        shell: bash
         env:
-          APPLE_SIGNING_IDENTITY: ${{ secrets.APPLE_SIGNING_IDENTITY }}
-        run: node ./packages/desktop/scripts/sign-managed-runtime-macos.mjs
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          EP_GH_IGNORE_TIME: true
+          CSC_LINK: ${{ secrets.APPLE_CERTIFICATE }}
+          CSC_KEY_PASSWORD: ${{ secrets.APPLE_CERTIFICATE_PASSWORD }}
+          APPLE_ID: ${{ secrets.APPLE_ID }}
+          APPLE_APP_SPECIFIC_PASSWORD: ${{ secrets.APPLE_PASSWORD }}
+          APPLE_TEAM_ID: ${{ secrets.APPLE_TEAM_ID }}
+          PASEO_DESKTOP_SMOKE: "1"
+        run: |
+          set -euo pipefail
+          build_args=(-- --publish never --mac --${{ matrix.electron_arch }})
+          build_args+=("-c.publish.releaseType=$RELEASE_TYPE")
+          build_args+=("-c.publish.channel=$RELEASE_CHANNEL")
+          npm run build:desktop "${build_args[@]}"
 
-      - name: Detect existing GitHub release state
-        if: env.IS_SMOKE_TAG != 'true'
+      - name: Upload desktop artifacts to release
+        if: env.SHOULD_PUBLISH == 'true' && env.IS_SMOKE_TAG != 'true'
         env:
           GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         shell: bash
         run: |
           set -euo pipefail
-          if release_draft="$(gh release view "$RELEASE_TAG" --repo "${{ github.repository }}" --json isDraft --jq '.isDraft' 2>/dev/null)"; then
-            :
-          else
-            release_draft="false"
+          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 $release_dir"
+            exit 1
           fi
-          echo "RELEASE_DRAFT=$release_draft" >> "$GITHUB_ENV"
+          gh release upload "$RELEASE_TAG" "${files[@]}" --clobber --repo "${{ github.repository }}"
 
-      - name: Build and publish macOS Tauri release
-        if: env.IS_SMOKE_TAG != 'true'
-        uses: tauri-apps/tauri-action@v0
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          APPLE_CERTIFICATE: ${{ secrets.APPLE_CERTIFICATE }}
-          APPLE_CERTIFICATE_PASSWORD: ${{ secrets.APPLE_CERTIFICATE_PASSWORD }}
-          APPLE_SIGNING_IDENTITY: ${{ secrets.APPLE_SIGNING_IDENTITY }}
-          APPLE_ID: ${{ secrets.APPLE_ID }}
-          APPLE_PASSWORD: ${{ secrets.APPLE_PASSWORD }}
-          APPLE_TEAM_ID: ${{ secrets.APPLE_TEAM_ID }}
-          TAURI_SIGNING_PRIVATE_KEY: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY }}
-          TAURI_SIGNING_PRIVATE_KEY_PASSWORD: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY_PASSWORD }}
+      - name: Upload manifest artifact
+        if: env.SHOULD_PUBLISH == 'true' && env.IS_SMOKE_TAG != 'true'
+        uses: actions/upload-artifact@v4
         with:
-          projectPath: packages/desktop
-          tagName: ${{ env.RELEASE_TAG }}
-          releaseName: Paseo ${{ env.RELEASE_TAG }}
-          releaseBody: See the assets to download and install this version.
-          releaseDraft: ${{ env.RELEASE_DRAFT }}
-          prerelease: false
-          args: --target ${{ matrix.rust_target }}
-
-      - name: Build macOS app (smoke only)
-        if: env.IS_SMOKE_TAG == 'true'
-        run: npm run tauri --workspace=@getpaseo/desktop build -- --target ${{ matrix.rust_target }} --no-bundle
+          name: mac-manifest-${{ 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
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-22.04
 
     steps:
       - 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="${SOURCE_TAG}"
-          release_tag="$source_tag"
-          for prefix in desktop-windows-v desktop-linux-v desktop-macos-v desktop-v; do
-            if [[ "$source_tag" == ${prefix}* ]]; then
-              release_tag="v${source_tag#${prefix}}"
-              break
-            fi
-          done
-          echo "RELEASE_TAG=$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
-
-      - name: Set desktop version from tag
-        shell: bash
-        run: |
-          node <<'NODE'
-          const fs = require('node:fs');
-          const path = require('node:path');
-          const version = process.env.DESKTOP_VERSION;
-          if (!version) throw new Error('DESKTOP_VERSION env var is missing');
-          console.log(`Setting desktop version to ${version}`);
-
-          const tauriConfPath = path.join('packages', 'desktop', 'src-tauri', 'tauri.conf.json');
-          const tauriConfText = fs.readFileSync(tauriConfPath, 'utf8');
-          const tauriRe = /("version"\s*:\s*")([^"]+)(")/;
-          if (!tauriRe.test(tauriConfText)) throw new Error('Failed to find version in tauri.conf.json');
-          fs.writeFileSync(tauriConfPath, tauriConfText.replace(tauriRe, `$1${version}$3`));
-
-          const cargoTomlPath = path.join('packages', 'desktop', 'src-tauri', 'Cargo.toml');
-          const lines = fs.readFileSync(cargoTomlPath, 'utf8').split(/\r?\n/);
-          let inPackage = false, updated = false;
-          const result = lines.map((line) => {
-            if (/^\[package\]\s*$/.test(line)) inPackage = true;
-            else if (inPackage && /^\[/.test(line)) inPackage = false;
-            if (inPackage && /^version\s*=\s*".*"\s*$/.test(line)) {
-              updated = true;
-              return `version = "${version}"`;
-            }
-            return line;
-          });
-          if (!updated) throw new Error('Failed to update Cargo.toml version');
-          fs.writeFileSync(cargoTomlPath, result.join('\n') + '\n');
-          NODE
-
-      - name: Install Linux packaging dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y libwebkit2gtk-4.1-dev libgtk-3-dev libappindicator3-dev librsvg2-dev patchelf
+        run: node scripts/emit-release-env.mjs --source-tag "$SOURCE_TAG" >> "$GITHUB_ENV"
 
       - name: Setup Node
         uses: actions/setup-node@v4
@@ -278,67 +218,72 @@ jobs:
           registry-url: "https://npm.pkg.github.com"
           scope: "@boudra"
 
-      - name: Install Rust stable
-        uses: dtolnay/rust-toolchain@stable
-
-      - name: Restore Rust cache
-        uses: Swatinem/rust-cache@v2
-        with:
-          shared-key: desktop-release-linux
-          workspaces: |
-            .
-            packages/desktop/src-tauri -> target
-
       - name: Install JS dependencies
         run: npm ci
         env:
           NODE_AUTH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
 
-      - name: Build web app for Tauri
-        run: npm run build:web --workspace=@getpaseo/app
+      - name: Set desktop package version from tag
+        shell: bash
+        run: |
+          node <<'NODE'
+          const fs = require('node:fs');
+          const path = require('node:path');
 
-      - name: Build managed runtime
-        run: npm run prepare:managed-runtime --workspace=@getpaseo/desktop
+          const version = process.env.DESKTOP_VERSION;
+          if (!version) throw new Error('DESKTOP_VERSION env var is missing');
 
-      - name: Validate managed runtime bundle
-        run: npm run validate:managed-runtime --workspace=@getpaseo/desktop
+          const packageJsonPath = path.join(process.env.DESKTOP_PACKAGE_PATH, 'package.json');
+          const packageJson = JSON.parse(fs.readFileSync(packageJsonPath, 'utf8'));
+          packageJson.version = version;
+          fs.writeFileSync(packageJsonPath, `${JSON.stringify(packageJson, null, 2)}\n`);
+          NODE
 
-      - name: Detect existing GitHub release state
-        if: env.IS_SMOKE_TAG != 'true'
+      - 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: ${{ secrets.GITHUB_TOKEN }}
+          EP_GH_IGNORE_TIME: true
+          PASEO_DESKTOP_SMOKE: "1"
+        run: |
+          set -euo pipefail
+          build_args=(-- --publish never --linux --x64)
+          build_args+=("-c.publish.releaseType=$RELEASE_TYPE")
+          build_args+=("-c.publish.channel=$RELEASE_CHANNEL")
+          npm run build:desktop "${build_args[@]}"
+
+      - name: Upload desktop artifacts to release
+        if: env.SHOULD_PUBLISH == 'true' && env.IS_SMOKE_TAG != 'true'
         env:
           GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         shell: bash
         run: |
           set -euo pipefail
-          if release_draft="$(gh release view "$RELEASE_TAG" --repo "${{ github.repository }}" --json isDraft --jq '.isDraft' 2>/dev/null)"; then
-            :
-          else
-            release_draft="false"
+          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 $release_dir"
+            exit 1
           fi
-          echo "RELEASE_DRAFT=$release_draft" >> "$GITHUB_ENV"
+          gh release upload "$RELEASE_TAG" "${files[@]}" --clobber --repo "${{ github.repository }}"
 
-      - name: Build and publish Linux Tauri release
-        if: env.IS_SMOKE_TAG != 'true'
-        uses: tauri-apps/tauri-action@v0
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          TAURI_SIGNING_PRIVATE_KEY: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY }}
-          TAURI_SIGNING_PRIVATE_KEY_PASSWORD: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY_PASSWORD }}
+      - name: Upload manifest artifact
+        if: env.SHOULD_PUBLISH == 'true' && env.IS_SMOKE_TAG != 'true'
+        uses: actions/upload-artifact@v4
         with:
-          projectPath: packages/desktop
-          tagName: ${{ env.RELEASE_TAG }}
-          releaseName: Paseo ${{ env.RELEASE_TAG }}
-          releaseBody: See the assets to download and install this version.
-          releaseDraft: ${{ env.RELEASE_DRAFT }}
-          prerelease: false
-          args: --bundles appimage
-
-      - name: Build Linux app (smoke only)
-        if: env.IS_SMOKE_TAG == 'true'
-        run: npm run tauri --workspace=@getpaseo/desktop build -- --no-bundle
+          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
@@ -348,62 +293,11 @@ 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="${SOURCE_TAG}"
-          release_tag="$source_tag"
-          for prefix in desktop-windows-v desktop-linux-v desktop-macos-v desktop-v; do
-            if [[ "$source_tag" == ${prefix}* ]]; then
-              release_tag="v${source_tag#${prefix}}"
-              break
-            fi
-          done
-          echo "RELEASE_TAG=$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
-
-      - name: Set desktop version from tag
-        shell: bash
-        run: |
-          node <<'NODE'
-          const fs = require('node:fs');
-          const path = require('node:path');
-          const version = process.env.DESKTOP_VERSION;
-          if (!version) throw new Error('DESKTOP_VERSION env var is missing');
-          console.log(`Setting desktop version to ${version}`);
-
-          const tauriConfPath = path.join('packages', 'desktop', 'src-tauri', 'tauri.conf.json');
-          const tauriConfText = fs.readFileSync(tauriConfPath, 'utf8');
-          const tauriRe = /("version"\s*:\s*")([^"]+)(")/;
-          if (!tauriRe.test(tauriConfText)) throw new Error('Failed to find version in tauri.conf.json');
-          fs.writeFileSync(tauriConfPath, tauriConfText.replace(tauriRe, `$1${version}$3`));
-
-          const cargoTomlPath = path.join('packages', 'desktop', 'src-tauri', 'Cargo.toml');
-          const lines = fs.readFileSync(cargoTomlPath, 'utf8').split(/\r?\n/);
-          let inPackage = false, updated = false;
-          const result = lines.map((line) => {
-            if (/^\[package\]\s*$/.test(line)) inPackage = true;
-            else if (inPackage && /^\[/.test(line)) inPackage = false;
-            if (inPackage && /^version\s*=\s*".*"\s*$/.test(line)) {
-              updated = true;
-              return `version = "${version}"`;
-            }
-            return line;
-          });
-          if (!updated) throw new Error('Failed to update Cargo.toml version');
-          fs.writeFileSync(cargoTomlPath, result.join('\n') + '\n');
-          NODE
+        run: node scripts/emit-release-env.mjs --source-tag "$SOURCE_TAG" >> "$GITHUB_ENV"
 
       - name: Setup Node
         uses: actions/setup-node@v4
@@ -414,64 +308,178 @@ jobs:
           registry-url: "https://npm.pkg.github.com"
           scope: "@boudra"
 
-      - name: Install Rust stable
-        uses: dtolnay/rust-toolchain@stable
+      - name: Install JS dependencies
+        run: npm ci
+        env:
+          NODE_AUTH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
 
-      - name: Restore Rust cache
-        uses: Swatinem/rust-cache@v2
+      - name: Set desktop package version from tag
+        shell: bash
+        run: |
+          node <<'NODE'
+          const fs = require('node:fs');
+          const path = require('node:path');
+
+          const version = process.env.DESKTOP_VERSION;
+          if (!version) throw new Error('DESKTOP_VERSION env var is missing');
+
+          const packageJsonPath = path.join(process.env.DESKTOP_PACKAGE_PATH, 'package.json');
+          const packageJson = JSON.parse(fs.readFileSync(packageJsonPath, 'utf8'));
+          packageJson.version = version;
+          fs.writeFileSync(packageJsonPath, `${JSON.stringify(packageJson, null, 2)}\n`);
+          NODE
+
+      - name: Build desktop release
+        shell: bash
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          EP_GH_IGNORE_TIME: true
+          PASEO_DESKTOP_SMOKE: "1"
+        run: |
+          set -euo pipefail
+          build_args=(-- --publish never --win --x64 --arm64)
+          build_args+=("-c.publish.releaseType=$RELEASE_TYPE")
+          build_args+=("-c.publish.channel=$RELEASE_CHANNEL")
+          npm run build:desktop "${build_args[@]}"
+
+      - name: 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:
-          shared-key: desktop-release-windows
-          workspaces: |
-            .
-            packages/desktop/src-tauri -> target
+          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: Build web app for Tauri
-        run: npm run build:web --workspace=@getpaseo/app
+      - 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: Build managed runtime
-        run: npm run prepare:managed-runtime --workspace=@getpaseo/desktop
+      - 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: Validate managed runtime bundle
-        run: npm run validate:managed-runtime --workspace=@getpaseo/desktop
+      - 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: Detect existing GitHub release state
+      - 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
-            :
-          else
-            release_draft="false"
+          cd release-manifests
+          manifests_dir="$PWD/final"
+          mkdir -p "$manifests_dir"
+
+          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
-          echo "RELEASE_DRAFT=$release_draft" >> "$GITHUB_ENV"
 
-      - name: Build and publish Windows Tauri release
-        if: env.IS_SMOKE_TAG != 'true'
-        uses: tauri-apps/tauri-action@v0
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          TAURI_SIGNING_PRIVATE_KEY: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY }}
-          TAURI_SIGNING_PRIVATE_KEY_PASSWORD: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY_PASSWORD }}
-        with:
-          projectPath: packages/desktop
-          tagName: ${{ env.RELEASE_TAG }}
-          releaseName: Paseo ${{ env.RELEASE_TAG }}
-          releaseBody: See the assets to download and install this version.
-          releaseDraft: ${{ env.RELEASE_DRAFT }}
-          prerelease: false
-          args: --bundles nsis,msi
+          if [[ "${{ needs.publish-linux.result }}" == "success" ]]; then
+            cp "linux-manifest/${RELEASE_CHANNEL}-linux.yml" "$manifests_dir/"
+          fi
 
-      - name: Build Windows app (smoke only)
-        if: env.IS_SMOKE_TAG == 'true'
-        run: npm run tauri --workspace=@getpaseo/desktop build -- --no-bundle
-        env:
-          TAURI_SIGNING_PRIVATE_KEY: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY }}
-          TAURI_SIGNING_PRIVATE_KEY_PASSWORD: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY_PASSWORD }}
+          if [[ "${{ 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

@@ -0,0 +1,45 @@
+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

@@ -0,0 +1,59 @@
+name: Nix Build
+
+on:
+  push:
+    branches: [main]
+    paths:
+      - "nix/**"
+      - "flake.nix"
+      - "flake.lock"
+      - "package.json"
+      - "package-lock.json"
+      - "packages/highlight/**"
+      - "packages/server/**"
+      - "packages/relay/**"
+      - "packages/cli/**"
+  pull_request:
+    branches: [main]
+    paths:
+      - "nix/**"
+      - "flake.nix"
+      - "flake.lock"
+      - "package.json"
+      - "package-lock.json"
+      - "packages/highlight/**"
+      - "packages/server/**"
+      - "packages/relay/**"
+      - "packages/cli/**"
+
+permissions:
+  contents: read
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    continue-on-error: false
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: cachix/install-nix-action@v31
+        with:
+          nix_path: nixpkgs=channel:nixos-unstable
+
+      - 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
+        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

.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: Build highlight dependency
+        run: npm run build --workspace=@getpaseo/highlight
+
       - name: Build relay dependency
         run: npm run build --workspace=@getpaseo/relay
 
@@ -44,3 +47,6 @@ jobs:
 
       - name: Test
         run: npm run test --workspace=@getpaseo/server
+        env:
+          CLAUDE_CODE_OAUTH_TOKEN: ${{ secrets.CLAUDE_CODE_OAUTH_TOKEN }}
+          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}

.gitignore

@@ -46,6 +46,9 @@ test-results/
 # Vercel
 .vercel/
 
+# Expo
+.expo/
+
 # Misc
 *.pem
 .vercel
@@ -71,6 +74,8 @@ valknut-report.json/
 **/.paseo-provider-history/
 .claude/settings.local.json
 **/.claude/settings.local.json
+.claude/scheduled_tasks.lock
+.claude/worktrees/
 .plans/
 packages/server/src/server/fixtures/dictation/dictation-debug-largest.wav
 packages/server/src/server/fixtures/dictation/dictation-debug-largest.transcript.txt

.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"]
+}

.oxlintrc.json

@@ -0,0 +1,89 @@
+{
+  "$schema": "./node_modules/oxlint/configuration_schema.json",
+  "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/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"
+      }
+    }
+  ]
+}

.prettierignore

@@ -1,21 +0,0 @@
-# Dependencies
-node_modules
-
-# Build outputs
-dist
-.next
-.expo
-build
-*.tsbuildinfo
-
-# Coverage
-coverage
-
-# Lock files
-*.lock
-package-lock.json
-
-# Generated
-android
-ios
-.turbo

.prettierrc

@@ -1,7 +0,0 @@
-{
-  "semi": false,
-  "singleQuote": true,
-  "trailingComma": "es5",
-  "tabWidth": 2,
-  "printWidth": 100
-}

CLAUDE.md

@@ -1,267 +1,122 @@
 # CLAUDE.md
 
-This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
-
-## Project Overview
-
-Paseo is a mobile app for monitoring and controlling your local AI coding agents from anywhere. Your dev environment, in your pocket.
-
-**Key features:**
-- Real-time streaming of agent output
-- Voice commands for hands-free interaction
-- Push notifications when tasks complete
-- Multi-agent orchestration across projects
-
-**Not a cloud sandbox** - Paseo connects directly to your actual development environment. Your code stays on your machine.
+Paseo is a mobile app for monitoring and controlling your local AI coding agents from anywhere. Your dev environment, in your pocket. Connects directly to your actual development environment — your code stays on your machine.
 
 **Supported agents:** Claude Code, Codex, and OpenCode.
 
-## Monorepo Structure
+## Repository map
 
 This is an npm workspace monorepo:
 
-- **packages/server**: The Paseo daemon that runs on your machine. Manages agent processes, provides WebSocket API for real-time streaming, and exposes an MCP server for agent control.
-- **packages/app**: Cross-platform client (Expo). Connects to one or more servers, displays agent output, handles voice input, and sends push notifications.
-- **packages/cli**: The `paseo` CLI that is used to manage the deamon, and acts as a client to it with  Docker-style commands like `paseo run/ls/logs/wait`
-- **packages/website**: Marketing site at paseo.sh (TanStack Router + Cloudflare Workers).
+- `packages/server` — Daemon: agent lifecycle, WebSocket API, MCP server
+- `packages/app` — Mobile + web client (Expo)
+- `packages/cli` — Docker-style CLI (`paseo run/ls/logs/wait`)
+- `packages/relay` — E2E encrypted relay for remote access
+- `packages/desktop` — Electron desktop wrapper
+- `packages/website` — Marketing site (paseo.sh)
 
-## Development Server
+## Documentation
 
-The `npm run dev` script automatically picks an available port for the development server.
+| 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/custom-providers.md](docs/custom-providers.md) | Custom provider config: Z.AI, Alibaba/Qwen, ACP agents, profiles, custom binaries |
+| [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                     |
 
-When running in a worktree or alongside the main checkout, set `PASEO_HOME` to isolate state:
+## Quick start
 
 ```bash
-PASEO_HOME=~/.paseo-blue npm run dev
+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
 ```
 
-- `PASEO_HOME` – path for runtime state (agent data, sockets, etc.). Defaults to `~/.paseo`; set this to a unique directory when running a secondary server instance.
+See [docs/development.md](docs/development.md) for full setup, build sync requirements, and debugging.
 
-For trace+ logs, check $PASEO_HOME/daemon.log
+## Critical rules
 
-## Running and checking logs
+- **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.
+- **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`
+- **NEVER make breaking changes to WebSocket or message schemas.** The primary compatibility path is old mobile app clients talking to newly updated daemons. Users update desktop and daemon first, then keep running the old app for a while. Every schema change MUST be backward-compatible for old clients against new daemons:
+  - New fields: always `.optional()` with a sensible default or `.transform()` fallback.
+  - Never change a field from optional to required.
+  - Never remove a field — deprecate it (keep accepting it, stop sending it).
+  - Never narrow a field's type (e.g. `string` → `enum`, `nullable` → non-null).
+  - Test with: "does a 6-month-old client still parse this?" and "does a 6-month-old daemon still send something this client accepts?"
 
-Both the server and Expo app are running in a Tmux session. See CLAUDE.local.md for system-specific session details.
+## 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
 
-### Daemon and CLI
-
-The Paseo daemon communicates via WebSocket. In the main checkout:
-- Daemon runs at `localhost:6767`
-- Expo app at `localhost:8081`
-- State lives in `$PASEO_HOME`
-
-In worktrees or when running `npm run dev`, ports and home directories may differ. Never assume the defaults.
-
-Use `npm run cli` to run the local CLI (instead of the globally linked `paseo` which points to the main checkout). Always run `npm run cli -- --help` or load the `/paseo` skill before using it - do not guess commands.
-
-Use `--host <host:port>` to point the CLI at a different daemon (e.g., `--host localhost:7777`).
-
-### Relay build sync (important)
-
-When changing `packages/relay/src/*`, rebuild relay before running/debugging the daemon:
-
-```bash
-npm run build --workspace=@getpaseo/relay
-```
-
-Reason: Node daemon imports `@getpaseo/relay` from `packages/relay/dist/*` (`node` export path), not directly from `src/*`.
-
-### Server build sync for CLI (important)
-
-When changing `packages/server/src/client/*` (especially `daemon-client.ts`) or shared WS protocol types, rebuild server before running/debugging CLI commands:
-
-```bash
-npm run build --workspace=@getpaseo/server
-```
-
-Reason: local CLI imports `@getpaseo/server` via package exports that resolve to `packages/server/dist/*` first. If `dist` is stale, CLI can speak an old protocol (for example, sending `session` before `hello`) and fail with handshake warnings/timeouts.
-
-### Quick reference CLI commands
-
-```bash
-npm run cli -- ls -a -g              # List all agents globally
-npm run cli -- ls -a -g --json       # Same, as JSON
-npm run cli -- inspect <id>          # Show detailed agent info
-npm run cli -- logs <id>             # View agent timeline
-npm run cli -- daemon status         # Check daemon status
-```
-
-### Agent state
-
-Agent data is stored at:
-```
-$PASEO_HOME/agents/{cwd-with-dashes}/{agent-id}.json
-```
-
-To find an agent by ID:
-```bash
-find $PASEO_HOME/agents -name "{agent-id}.json"
-```
-
-To find an agent by title or other content:
-```bash
-rg -l "some title text" $PASEO_HOME/agents/
-rg -l "spiteful-toad" $PASEO_HOME/agents/
-```
-
-### Provider session files
-
-Get the session ID from the agent JSON file (`persistence.sessionId`), then:
-
-**Claude sessions:**
-```
-~/.claude/projects/{cwd-with-dashes}/{session-id}.jsonl
-```
-
-**Codex sessions:**
-```
-~/.codex/sessions/{YYYY}/{MM}/{DD}/rollout-{timestamp}-{session-id}.jsonl
-```
-
-## Android
-
-Take screenshots like this: `adb exec-out screencap -p > screenshot.png`
-
-### Android variants (vanilla Expo)
-
-Use `APP_VARIANT` in `packages/app/app.config.js` to control app name + package ID (no custom Gradle flavor plugin):
-
-- `production` -> app name `Paseo`, package `sh.paseo`
-- `development` -> app name `Paseo Debug`, package `sh.paseo.debug`
-
-EAS profiles live in `packages/app/eas.json` as `development`, `production`, and `production-apk`.
-
-`development` uses Android `debug`.
-
-### Local build + install (Android device)
-
-From `packages/app`:
-
-```bash
-# development (debug)
-APP_VARIANT=development npx expo prebuild --platform android --non-interactive
-APP_VARIANT=development npx expo run:android --variant=debug
-
-# production (release)
-APP_VARIANT=production npx expo prebuild --platform android --non-interactive
-APP_VARIANT=production npx expo run:android --variant=release
-
-# clean native project (when needed)
-npx expo prebuild --platform android --clean --non-interactive
-```
-
-From repo root:
-
-```bash
-npm run android:development
-npm run android:production
-npm run android:clean
-```
-
-`npm run android:release` is an alias for `npm run android:production`.
-
-### Cloud build + submit (EAS Workflows)
-
-Tag pushes like `v0.1.0` trigger `packages/app/.eas/workflows/release-mobile.yml` on Expo servers.
-Tag pushes like `v0.1.0` also trigger `.github/workflows/android-apk-release.yml` on GitHub Actions to publish an APK asset on the matching GitHub Release.
-
-That workflow does:
-- Build iOS with the `production` profile
-- Build Android with the `production` profile
-- Submit each build with the `production` submit profile
-
-Useful commands:
-
-```bash
-# List recent mobile workflow runs
-cd packages/app && npx eas workflow:runs --workflow release-mobile.yml --limit 10
-
-# Inspect one run (jobs, status, outputs)
-cd packages/app && npx eas workflow:view <run-id>
-
-# Stream logs for all steps in one failed job
-cd packages/app && npx eas workflow:logs <job-id> --non-interactive --all-steps
-```
-
-## Testing with Playwright MCP
-
-**CRITICAL:** When asked to test the app, you MUST use the Playwright MCP connecting to Metro at `http://localhost:8081`.
-
-Use the Playwright MCP to test the app in Metro web. Navigate to `http://localhost:8081` to interact with the app UI.
-
-**Important:** 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 navigation breaks the state.
-
-## Expo troubleshooting
-
-Run `npx expo-doctor` to diagnose version mismatches and native module issues.
-
-## Release playbook
-
-Use the scripted release flow from repo root. Avoid manual version bumps, manual tags, or ad hoc publish commands unless debugging.
-
-```bash
-# Recommended: full patch release (bump, check, publish, push branch+tag)
-npm run release:patch
-
-# Manual, step-by-step fallback:
-npm run version:all:patch  # npm version across all workspaces (creates commit + local tag)
-npm run release:check
-npm run release:publish
-npm run release:push       # pushes HEAD and current version tag (triggers desktop + Android APK + EAS mobile workflows)
-```
-
-### Draft release flow
-
-```bash
-# Stage a draft GitHub release with assets, but do not publish npm yet.
-npm run draft-release:patch
-
-# Publish npm and promote the same GitHub draft release to final.
-npm run release:finalize
-```
-
-Behavior:
-- `draft-release:patch` bumps the version, runs release checks, pushes `HEAD` and the new `v*` tag, and creates the matching GitHub Release as a draft so desktop assets, APK uploads, and synced notes attach to that same draft release.
-- `release:finalize` requires that the current tag already has a GitHub draft release, publishes the npm packages for that exact version, and promotes the same GitHub Release from draft to published.
-- Use the same semver tag for both draft and final states; do not cut a second tag just to publish the release.
-
-Notes:
-- `version:all:*` bumps the root package version and runs the root `version` lifecycle script to sync workspace versions and internal `@getpaseo/*` dependency versions before the release commit/tag is created.
-- `release:prepare` refreshes workspace `node_modules` links to prevent stale local package types during release checks.
-- If `release:publish` fails after a successful publish of one workspace, re-run `npm run release:publish`; npm will skip already-published versions and continue where possible.
-- If a user asks to "release paseo" (without specifying major/minor), treat it as a patch release and run `npm run release:patch`.
-- All workspaces share one version by design. Keep versions synchronized and release together.
-- The website Mac download CTA URL is derived from `packages/website/package.json` version at build time, so no manual update is required after release.
-
-Release completion checklist:
-- Manually update CHANGELOG.md with release notes, between current release vs previous one, use Git commands to figure out what changed. The notes are user-facing:
-    - Ask yourself, what do Paseo users want to know about?
-    - Include: New features, bug fixes
-    - Don't include: Refactors or code changes that are not noticeable by users
-- `npm run release:patch` completes successfully.
-- GitHub `Desktop Release` workflow for the new `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 (Expo queues can take longer on the free plan).
-
-## Orchestrator Mode
-
-- **When agent control tool calls fail**, make sure you list agents before trying to launch another one. It could just be a wait timeout.
-- **Always prefix agent titles** so we can tell which ones are running under you (e.g., "🎭 Feature Implementation", "🎭 Design Discussion").
-- **Launch agents in the most permissive mode**: Use full access or bypass permissions mode.
-- **Set cwd to the repository root** - The agent's working directory should usually be the repo root
-
-**CRITICAL: ALWAYS RUN TYPECHECK AFTER EVERY CHANGE.**
-
-## Agent Authentication
-
-All agent providers (Claude, Codex, OpenCode) handle their own authentication outside of environment variables. They are authenticated without providing any extra configuration—Paseo does not manage API keys or tokens for agents.
-
-**Do not add auth checks to tests.** If auth fails for whatever reason, let the user know instead of patching the code or adding conditional skips.
-
-## NEVER DO THESE THINGS
-
-- **NEVER restart the main Paseo daemon on port 6767 without permission** - This is the production daemon that launches and manages agents. If you are reading this, you are probably running as an agent under it. Restarting it will kill your own process and all other running agents. The daemon is managed by the user in Tmux.
-- **NEVER assume a timeout means the service needs restarting** - Timeouts can be transient network issues, not service failures
-- **NEVER add authentication checks to tests** - Agent providers handle their own auth. If tests fail due to auth issues, report it rather than adding conditional skips or env var checks
+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

@@ -4,39 +4,124 @@
 
 <h1 align="center">Paseo</h1>
 
-<p align="center">Manage coding agents from your phone and desktop.</p>
+<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">
-  <img src="https://paseo.sh/paseo-mockup.png" alt="Paseo app screenshot" width="100%">
+  <img src="https://paseo.sh/hero-mockup.png" alt="Paseo app screenshot" width="100%">
+</p>
+
+<p align="center">
+  <img src="https://paseo.sh/mobile-mockup.png" alt="Paseo mobile app" width="100%">
 </p>
 
 ---
 
-> [!WARNING]
-> **Early development** — Features may break or change without notice. Use at your own risk.
+Run agents in parallel on your own machines. Ship from your phone or your desk.
 
-Paseo is a self-hosted daemon for Claude Code, Codex, and OpenCode. Agents run on your machine with your full dev environment. Connect from phone, desktop, or web.
+- **Self-hosted:** Agents run on your machine with your full dev environment. Use your tools, your configs, and your skills.
+- **Multi-provider:** Claude Code, Codex, and OpenCode through the same interface. Pick the right model for each job.
+- **Voice control:** Dictate tasks or talk through problems in voice mode. Hands-free when you need it.
+- **Cross-device:** iOS, Android, desktop, web, and CLI. Start work at your desk, check in from your phone, script it from the terminal.
+- **Privacy-first:** Paseo doesn't have any telemetry, tracking, or forced log-ins.
 
 ## Getting Started
 
+Paseo runs a local server called the daemon that manages your coding agents. Clients like the desktop app, mobile app, web app, and CLI connect to it.
+
+### Prerequisites
+
+You need at least one agent CLI installed and configured with your credentials:
+
+- [Claude Code](https://docs.anthropic.com/en/docs/claude-code)
+- [Codex](https://github.com/openai/codex)
+- [OpenCode](https://github.com/anomalyco/opencode)
+
+### Desktop app (recommended)
+
+Download it from [paseo.sh/download](https://paseo.sh/download) or the [GitHub releases page](https://github.com/getpaseo/paseo/releases). Open the app and the daemon starts automatically. Nothing else to install.
+
+To connect from your phone, scan the QR code shown in Settings.
+
+### CLI / headless
+
+Install the CLI and start Paseo:
+
 ```bash
 npm install -g @getpaseo/cli
 paseo
 ```
 
-Then open the app and connect to your daemon.
+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)
 
+## CLI
+
+Everything you can do in the app, you can do from the terminal.
+
+```bash
+paseo run --provider claude/opus-4.6 "implement user authentication"
+paseo run --provider codex/gpt-5.4 --worktree feature-x "implement feature X"
+
+paseo ls                           # list running agents
+paseo attach abc123                # stream live output
+paseo send abc123 "also add tests" # follow-up task
+
+# run on a remote daemon
+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)
+
+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.
+
+```bash
+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
+```
+
 ## 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
-- `packages/desktop`: Tauri desktop app
+- `packages/desktop`: Electron desktop app
 - `packages/relay`: Relay package for remote connectivity
 - `packages/website`: Marketing site and documentation (`paseo.sh`)
 
@@ -49,12 +134,26 @@ npm run dev
 # run individual surfaces
 npm run dev:server
 npm run dev:app
+npm run dev:desktop
 npm run dev:website
 
+# build the daemon
+npm run build:daemon
+
 # repo-wide checks
 npm run typecheck
 ```
 
+<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
 
-MIT
+AGPL-3.0

SECURITY.md

@@ -22,7 +22,7 @@ The relay is designed to be untrusted. All traffic between your phone and daemon
 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.
+4. Both sides perform an ECDH key exchange to derive a shared secret. All subsequent messages are encrypted with XSalsa20-Poly1305 (NaCl box).
 
 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.
 
@@ -31,19 +31,31 @@ The relay sees only: IP addresses, timing, message sizes, and session IDs. It ca
 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
+- **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`. The local control plane is trusted by network reachability, not by an additional authentication token.
+
+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.
+
+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.
+
+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 incoming requests against configured hostnames. Requests with unrecognized hosts are rejected.
 
 ## Agent authentication
 
@@ -51,4 +63,4 @@ Paseo wraps agent CLIs (Claude Code, Codex, OpenCode) but does not manage their
 
 ## Reporting vulnerabilities
 
-If you discover a security vulnerability, please report it privately by emailing mo@faro.so. Do not open a public issue.
+If you discover a security vulnerability, please report it privately by emailing hello@moboudra.com. Do not open a public issue.

app.json

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

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:${port}/ws`,
+  appVersion: "0.1.54", // see gotcha #1
+});
+await client.connect();
+await client.fetchAgents({ subscribe: { subscriptionId: "test" } });
+
+// ... do your testing ...
+
+await client.close();
+await daemon.stop();
+await rm(paseoHomeRoot, { recursive: true, force: true });
+await rm(staticDir, { recursive: true, force: true });
+```
+
+Run with:
+
+```bash
+npx tsx packages/server/src/server/your-script.ts
+```
+
+## Using the test helper
+
+For simpler cases, `createTestPaseoDaemon` + `DaemonClient` handles temp dirs and port selection:
+
+```typescript
+import { createTestPaseoDaemon } from "./test-utils/paseo-daemon.js";
+import { DaemonClient } from "./test-utils/daemon-client.js";
+
+const daemon = await createTestPaseoDaemon();
+const client = new DaemonClient({
+  url: `ws://127.0.0.1:${daemon.port}/ws`,
+  appVersion: "0.1.54",
+});
+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: "0.1.54",
+});
+```
+
+### 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/android.md

@@ -0,0 +1,69 @@
+# Android
+
+## App variants
+
+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`       |
+| `development` | Paseo Debug | `sh.paseo.debug` |
+
+EAS profiles: `development`, `production`, and `production-apk` in `packages/app/eas.json`.
+
+`development` uses Android `debug`.
+
+## Local build + install
+
+From repo root:
+
+```bash
+npm run android:development    # Debug build
+npm run android:production     # Release build
+npm run android:clear          # Remove generated Android project
+```
+
+Or from `packages/app`:
+
+```bash
+# Debug
+APP_VARIANT=development npx expo prebuild --platform android --non-interactive
+APP_VARIANT=development npx expo run:android --variant=debug
+
+# Release
+APP_VARIANT=production npx expo prebuild --platform android --non-interactive
+APP_VARIANT=production npx expo run:android --variant=release
+
+# Clear generated Android project
+rm -rf android
+```
+
+## Screenshots
+
+```bash
+adb exec-out screencap -p > screenshot.png
+```
+
+## Cloud build + submit (EAS)
+
+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.
+
+### Useful commands
+
+```bash
+cd packages/app
+
+# List recent workflow runs
+npx eas workflow:runs --workflow release-mobile.yml --limit 10
+
+# Inspect a run
+npx eas workflow:view <run-id>
+
+# Stream logs for a failed job
+npx eas workflow:logs <job-id> --non-interactive --all-steps
+```

docs/architecture.md

@@ -0,0 +1,209 @@
+# 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
+
+### 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
+
+```
+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

@@ -0,0 +1,188 @@
+# 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/custom-providers.md

@@ -0,0 +1,587 @@
+# 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"
+        },
+        "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"
+        },
+        "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 @anthropic-ai/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"
+        },
+        "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"
+        },
+        "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 +1,428 @@
+# Data Model
+
+Paseo uses **file-based JSON persistence** instead of a traditional database. All data is validated at runtime with Zod schemas and written atomically (write to temp file, then rename). There are no migrations — schemas use optional fields with defaults for forward compatibility.
+
+All server-side stores live under `$PASEO_HOME` (defaults to `~/.paseo`).
+
+---
+
+## Directory layout
+
+```
+$PASEO_HOME/
+├── config.json                          # Daemon configuration
+├── agents/
+│   └── {project-dir}/
+│       └── {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
+```
+
+---
+
+## 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 `{}`)                                        |
+| `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                                           |
+| `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?`             |
+| `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[],
+    mcp: { enabled: boolean },
+    cors: { allowedOrigins: string[] },
+    relay: { enabled: boolean, endpoint: string, publicEndpoint: string }
+  },
+  app: {
+    baseUrl: string
+  },
+  providers: {
+    openai: { apiKey: string },
+    local: { modelsDir: string }
+  },
+  agents: {
+    providers: {
+      [provider: string]: {
+        command: { mode: "default" } | { mode: "append", args: string[] } | { mode: "replace", argv: string[] },
+        env: Record<string, string>
+      }
+    }
+  },
+  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.
+
+| 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.
+
+| 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                              |
+| `workerProvider`        | `string?`                                           | Override provider for workers              |
+| `workerModel`           | `string?`                                           | Override model for workers                 |
+| `verifierProvider`      | `string?`                                           | Override provider for verifiers            |
+| `verifierModel`         | `string?`                                           | Override model 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?` (ISO 8601) | Soft-delete timestamp          |
+
+---
+
+## 7. Workspace Registry
+
+**Path:** `$PASEO_HOME/projects/workspaces.json`
+
+Array of workspace records. A workspace is a specific working directory within a project.
+
+| Field         | Type                                            | Description             |
+| ------------- | ----------------------------------------------- | ----------------------- |
+| `workspaceId` | `string`                                        | Primary key             |
+| `projectId`   | `string`                                        | FK to Project.projectId |
+| `cwd`         | `string`                                        | Filesystem path         |
+| `kind`        | `"local_checkout" \| "worktree" \| "directory"` |                         |
+| `displayName` | `string`                                        |                         |
+| `createdAt`   | `string` (ISO 8601)                             |                         |
+| `updatedAt`   | `string` (ISO 8601)                             |                         |
+| `archivedAt`  | `string?` (ISO 8601)                            | Soft-delete timestamp   |
+
+---
+
+## 8. Push Token Store
+
+**Path:** `$PASEO_HOME/push-tokens.json`
+
+```json
+{
+  "tokens": ["ExponentPushToken[...]", ...]
+}
+```
+
+Simple set of Expo push notification tokens. No schema validation — just an array of strings.
+
+---
+
+## 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-system.md

@@ -0,0 +1,247 @@
+# Design system
+
+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:31-34`), 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:3052-3057`).
+- **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: `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:1056-1062`, `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 daemon version mismatch 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/daemon-version-mismatch-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/daemon-version-mismatch-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/design.md

@@ -0,0 +1,73 @@
+# 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

@@ -0,0 +1,198 @@
+# 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`.
+- In git worktrees, `npm run dev` derives a stable home like `~/.paseo-<worktree-name>`.
+  On first run, it seeds that home from `~/.paseo` by copying agent/project JSON metadata
+  and `config.json`; actual checkout/worktree directories are not copied.
+- `PASEO_DEV_SEED_HOME=/path/to/home npm run dev` seeds from a different source home.
+- `PASEO_DEV_RESET_HOME=1 npm run dev` clears and reseeds the derived worktree home.
+
+### 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 `$PASEO_HOME/daemon.log` for trace-level logs.
+
+### Database queries
+
+Run arbitrary SQL against the SQLite database:
+
+```bash
+# Show table row counts
+npm run db:query
+
+# Run any SQL
+npm run db:query -- "SELECT agent_id, title, last_status FROM agent_snapshots"
+npm run db:query -- "SELECT agent_id, seq, item_kind FROM agent_timeline_rows ORDER BY committed_at DESC LIMIT 10"
+
+# Point at a specific DB directory
+npm run db:query -- --db /path/to/db "SELECT ..."
+```
+
+Auto-detects the running dev daemon's database from `/tmp/paseo-dev.*`, `PASEO_HOME`, or `~/.paseo/db`.
+Pass either a DB directory or a `paseo.sqlite` file to `--db`. The script opens the database directly in read-only mode.
+
+## paseo.json service scripts
+
+`worktree.setup` and `worktree.teardown` accept either a multiline shell script or an array
+of commands. Both run sequentially.
+
+```json
+{
+  "worktree": {
+    "setup": "npm ci\ncp \"$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=$PASEO_PORT npm run dev:web"
+    }
+  }
+}
+```
+
+## Build sync gotchas
+
+### Relay → Daemon
+
+When changing `packages/relay/src/*`, rebuild before running the daemon:
+
+```bash
+npm run build --workspace=@getpaseo/relay
+```
+
+The Node daemon imports `@getpaseo/relay` from `packages/relay/dist/*`, not `src/*`.
+
+### Server → CLI
+
+When changing `packages/server/src/client/*` (especially `daemon-client.ts`) or shared WS protocol types, rebuild before running CLI commands:
+
+```bash
+npm run build --workspace=@getpaseo/server
+```
+
+The CLI imports `@getpaseo/server` via package exports resolving to `dist/*`. Stale `dist` means the CLI speaks an old protocol and fails with handshake warnings or timeouts.
+
+## CLI reference
+
+Use `npm run cli` to run the local CLI (instead of the globally installed `paseo` which points to the main checkout).
+
+```bash
+npm run cli -- ls -a -g              # List all agents globally
+npm run cli -- ls -a -g --json       # Same, as JSON
+npm run cli -- inspect <id>          # Show detailed agent info
+npm run cli -- logs <id>             # View agent timeline
+npm run cli -- daemon status         # Check daemon status
+```
+
+Use `--host <host:port>` to point the CLI at a different daemon:
+
+```bash
+npm run cli -- --host localhost:7777 ls -a
+```
+
+## Agent state
+
+Agent data lives at:
+
+```
+$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/file-icons.md

@@ -0,0 +1,109 @@
+# 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
+
+## 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 +1,45 @@
+# 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 a project". Code: `ProjectSummary` (`packages/app/src/utils/projects.ts:17`), `projectKey` (`packages/server/src/server/workspace-registry-model.ts:88`). 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:2128`). 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:9`).
+- **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:597`). Forbidden: "Task", "Job", "Run".
+- **Daemon** — Local Paseo server process; identified by `serverId`. UI: "Daemon" (system contexts only). Code: `serverId` (`packages/server/src/shared/messages.ts:1886`), `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:36`). Forbidden: "Connection" (means `HostConnection`, not host).
+- **Placement** — One workspace's relationship to its project (projectKey, projectName, git checkout snapshot). Internal. Code: `ProjectPlacementPayload` (`packages/server/src/shared/messages.ts:2063`).
+- **Branch** — Plain git branch. UI: "Switch branch". Code: `currentBranch` (`packages/server/src/shared/messages.ts:2027`); `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:2042`); 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:36`). Never user-facing.
+- **Provider** — Agent backend (Claude Code, Codex, OpenCode). UI: "Provider". Code: `ProviderSnapshotEntry` (`packages/server/src/shared/messages.ts:193`).
+- **Model** — A specific LLM offered by a provider. UI: "Model" / "Select model". Code: `AgentModelDefinition` (`packages/server/src/shared/messages.ts:182`).
+- **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:249`).
+- **Attachment** — GitHub PR or Issue bound to an agent prompt. UI: "Attach issue or PR". Code: `AgentAttachment` (`packages/server/src/shared/messages.ts:736`).
+- **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:576`); (b) **git merge conflict** (no current UI string).
+
+## Open question — per-host project entry (TBD)
+
+A project aggregates workspaces across daemons. The projects screen and project-settings screen need a name for "one row in the project list per (project, daemon)". `ProjectPlacementPayload` is **per-workspace**, so it isn't this thing — but the noun "placement" could naturally extend (`ProjectDaemonPlacement`, or just "placements grouped by daemon"). The in-progress `ProjectCheckout` type (`packages/app/src/utils/projects.ts:4`) is also per-workspace today, even though the settings UI selector treats it per-daemon (matched on `serverId` alone) — so the code is currently inconsistent with itself.
+
+Candidates: (1) extend "placement" to the (project, daemon) bundle, (2) drop the wrapper type and use `{ host, project, workspaces[] }` plus descriptive UI copy ("project · host X · 2 online"). Recommendation: option (2) — no new noun, use existing terms compositionally. Do not introduce "Checkout" / `ProjectCheckout` as the canonical name.
+
+## Rename before landing (in-progress `Checkout*` references)
+
+- `packages/app/src/utils/projects.ts:4,20,22,23,46,85,104,120,124,126,146,148,155` — `ProjectCheckout`, `checkouts`, `checkoutCount`, `onlineCheckoutCount`, `buildCheckoutTarget`, `compareCheckouts`.
+- `packages/app/src/screens/projects-screen.tsx:73,74,77,90` — `checkoutLabel`, `${checkoutCount} checkout(s)`, `onlineSuffix`.
+- `packages/app/src/screens/project-settings-screen.tsx:17,41,46,55,76,81,110-152,323,443-600` — `ProjectCheckout` import, `CheckoutSelection`, `CheckoutSelector`, `CheckoutOption`, `usableCheckouts`, `onlineUsableCheckouts`, `selectedCheckout`, `selectedCheckoutKey`, "no online checkout", "no checkout with a valid server", `testID="checkout-selector"`, `testID="checkout-option-*"`, a11y `"Edit X checkout"`.
+- `packages/app/src/screens/project-settings-screen.test.tsx:185,195,196,366,369,378,389,412,415,431,432,456,464,478,481,490,501,504,521` — `checkouts`, `checkoutCount`, `onlineCheckoutCount`, "no checkouts are online", "checkout selector", "online checkout".
+- `packages/app/src/utils/projects.test.ts:107,108,111,259` — `checkoutCount`, `onlineCheckoutCount`, `checkouts`.
+
+(Out of scope for this rename: `ProjectCheckoutLite*Payload` and the git-`checkout` family in `packages/server/src/utils/checkout-*.ts`. Those refer to _git_ checkout state, not the rejected (project, daemon) sense.)
+
+## Inconsistencies (documented, not papered over)
+
+- CLI `--host <host>` description `"Daemon host target"` (`packages/cli/src/commands/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:2137`) while `PersistedWorkspaceKind` does not (`packages/server/src/server/workspace-registry-model.ts:9`).
+- In-progress `ProjectCheckout` (`packages/app/src/utils/projects.ts:4`) is per-workspace, but `project-settings-screen.tsx` selects by `serverId` alone — same daemon with multiple workspaces will produce duplicate React keys in the selector.

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 `add-host-modal.tsx` and `pair-link-modal.tsx` for the pattern. 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

@@ -0,0 +1,76 @@
+# Product
+
+What Paseo is, who it's for, and where it's going.
+
+## What is Paseo
+
+Paseo is a next-generation development environment built around agents. One interface to run, monitor, and interact with coding agents across desktop, mobile, terminal, and web.
+
+The development workflow is shifting from manually editing files to orchestrating agents that do the editing. Paseo is built for that workflow.
+
+## Core philosophy
+
+Freedom and flexibility. Every design decision follows from this:
+
+- **Multi-provider** — Use any coding agent harness. Pick the right model for each job, switch freely as the landscape shifts. No vendor-lock in.
+- **Cross-device** — Desktop, mobile, web, CLI. Start work at your desk, check progress from your phone, script from the terminal.
+- **Self-hosted** — The daemon runs on your machine. Your code, your keys, your environment. No inference markup, no cloud dependency.
+- **Respectful** - No telemetry, no forced cloud, no forced accounts
+- **Open source** — AGPL-3.0. Users can inspect, fork, and contribute.
+- **BYOK** — Bring your own keys. Use your subsidized plans and first-party provider pricing. Paseo adds zero cost on top.
+
+## How it works
+
+### Projects and workspaces
+
+Projects are grouped in the sidebar, detected automatically from your filesystem and tagged by git remote when available.
+
+Each project opens as a workspace. For git projects, the default workspace is the main checkout. Users can create additional workspaces, which are isolated copies (git worktrees) where agents work without affecting main.
+
+### 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
+
+### The daemon
+
+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
+
+## Target user
+
+Anyone who builds software:
+
+- Care about owning their tools and their data
+- Use multiple AI providers and want to switch freely
+- Run agents on real tasks across real projects
+- Want to work from multiple devices
+
+## What compounds over time
+
+- **Trust** — Showing up daily, shipping in public, being open source. Earned slowly, lost quickly.
+- **Community contributions** — Code, packaging, skills, agent configs. Contributors become advocates.
+- **Ecosystem** — Skills, integrations, shared configs. Community-built content that makes the platform more valuable.
+
+## Strategic bets
+
+1. **Models commoditize.** Value moves to the orchestration layer. The best model changes monthly — the workflow layer stays.
+2. **Multi-provider wins.** No single provider stays on top. Developers want the best model for each task.
+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)
+
+- 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

docs/providers.md

@@ -0,0 +1,366 @@
+# 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`. 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.
+
+Existing ACP providers: `claude-acp`, `copilot`.
+
+### Direct
+
+Implement the `AgentClient` and `AgentSession` interfaces yourself. This gives full control but requires you to handle process management, streaming, permissions, and session persistence from scratch.
+
+Existing direct providers: `claude`, `codex`, `opencode`.
+
+---
+
+## 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:
+
+```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: runtimeSettings?.["my-provider"],
+    }),
+};
+```
+
+### 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`:
+
+```ts
+import { MyProviderIcon } from "@/components/icons/my-provider-icon";
+
+const PROVIDER_ICONS: Record<string, typeof Bot> = {
+  claude: ClaudeIcon as unknown as typeof Bot,
+  codex: CodexIcon as unknown as typeof Bot,
+  "my-provider": MyProviderIcon as unknown as typeof Bot,
+};
+```
+
+If no icon is registered, the app 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()`:
+
+```ts
+case "my-provider":
+  return (
+    isCommandAvailable("my-agent-binary") &&
+    Boolean(process.env.MY_PROVIDER_API_KEY)
+  );
+```
+
+Add to the `allProviders` array:
+
+```ts
+export const allProviders: AgentProvider[] = [
+  "claude",
+  "claude-acp",
+  "codex",
+  "copilot",
+  "opencode",
+  "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
+
+**`AgentClient`** -- factory for sessions and model listing:
+
+```ts
+interface AgentClient {
+  readonly provider: AgentProvider;
+  readonly capabilities: AgentCapabilityFlags;
+  createSession(
+    config: AgentSessionConfig,
+    launchContext?: AgentLaunchContext,
+  ): Promise<AgentSession>;
+  resumeSession(
+    handle: AgentPersistenceHandle,
+    overrides?: Partial<AgentSessionConfig>,
+    launchContext?: AgentLaunchContext,
+  ): Promise<AgentSession>;
+  listModels(options?: ListModelsOptions): Promise<AgentModelDefinition[]>;
+  isAvailable(): Promise<boolean>;
+  // Optional:
+  listPersistedAgents?(options?: ListPersistedAgentsOptions): Promise<PersistedAgentDescriptor[]>;
+}
+```
+
+**`AgentSession`** -- a running agent conversation:
+
+```ts
+interface AgentSession {
+  readonly provider: AgentProvider;
+  readonly id: string | null;
+  readonly capabilities: AgentCapabilityFlags;
+  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<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>;
+}
+```
+
+### 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 +1,317 @@
+# 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.
+- Make sure local `npm run typecheck` passes on that commit.
+- 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, APK, and EAS mobile 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.
+
+## 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`.
+
+## 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

@@ -0,0 +1,126 @@
+# Testing
+
+## Philosophy
+
+Tests prove behavior, not structure. Every test should answer: "what user-visible or API-visible behavior does this verify?"
+
+## Test-driven development
+
+Work in vertical slices: one test, one implementation, repeat. Each test responds to what you learned from the previous cycle.
+
+```
+RIGHT (vertical):
+  RED→GREEN: test1→impl1
+  RED→GREEN: test2→impl2
+  RED→GREEN: test3→impl3
+
+WRONG (horizontal):
+  RED:   test1, test2, test3, test4, test5
+  GREEN: impl1, impl2, impl3, impl4, impl5
+```
+
+Writing all tests first then all implementation produces bad tests — you end up testing imagined behavior instead of actual behavior.
+
+## Determinism first
+
+Tests must produce the same result every run:
+
+- No conditional assertions or branching paths
+- No reliance on timing, randomness, or network jitter
+- No weak assertions (`toBeTruthy`, `toBeDefined`)
+- Assert the full intended behavior, not fragments
+
+```typescript
+// Bad: conditional and weak
+it("creates a tool call", async () => {
+  const result = await createToolCall(input);
+  if (result.ok) {
+    expect(result.id).toBeDefined();
+  }
+});
+
+// Good: deterministic and explicit
+it("returns timeout error when provider times out", async () => {
+  const result = await createToolCall(input);
+  expect(result).toEqual({
+    ok: false,
+    error: { code: "PROVIDER_TIMEOUT", waitedMs: 30000 },
+  });
+});
+```
+
+## Flaky tests are a bug
+
+Never remove a test because it's flaky. Find the variance source (time, randomness, race condition, shared state, non-deterministic output, environment drift) and fix it.
+
+## Real dependencies over mocks
+
+Mocks are not the default. They require an explicit decision.
+
+- **Database**: real test database, not a mock
+- **APIs**: real APIs with test/sandbox credentials, not request mocks
+- **File system**: temporary directory that gets cleaned up, not fs mocks
+
+Ask: "will this still hold with real dependencies at runtime?" If no, don't mock.
+
+### Use swappable adapters instead
+
+When you need test isolation, design code so dependencies are injectable:
+
+```typescript
+interface EmailSender {
+  send(to: string, body: string): Promise<void>;
+}
+
+// Production
+const realSender: EmailSender = { send: sendgrid.send };
+
+// Test: in-memory adapter
+function createTestEmailSender() {
+  const sent: Array<{ to: string; body: string }> = [];
+  return {
+    send: async (to: string, body: string) => {
+      sent.push({ to, body });
+    },
+    sent,
+  };
+}
+```
+
+## End-to-end means end-to-end
+
+When a test is labeled end-to-end, it calls the real service. No environment variable gates, no conditional skipping, no mocking the external dependency.
+
+## Test organization
+
+- Collocate tests with implementation: `thing.ts` + `thing.test.ts`
+- Extract complex setup into reusable helpers
+- Test bodies should read like plain English
+- Build a vocabulary of test helpers that make complex flows simple
+
+## 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.
+
+## Debugging with tests
+
+Use the test as your debugging ground:
+
+1. Add temporary logging to the code under test
+2. Run the test, observe actual values
+3. Trace the flow end-to-end through test output
+4. Confirm each assumption with actual output
+5. Remove logging when done
+
+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
+- Setup requires too much global state
+
+Aim for deep modules: small interface, deep implementation. Fewer methods = fewer tests needed, simpler params = simpler setup.

docs/unistyles.md

@@ -0,0 +1,302 @@
+# 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 / `AgentStreamSection` / `AgentComposerSection`
+- `Composer` and `MessageInput`
+- `WorkspaceScreen` shell, tabs row, 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.
+
+When the content container itself needs themed behavior, wrap the component with [`withUnistyles`](https://www.unistyl.es/v3/references/with-unistyles):
+
+```tsx
+import { ScrollView } from "react-native";
+import { StyleSheet, withUnistyles } from "react-native-unistyles";
+
+const ThemedScrollView = withUnistyles(ScrollView);
+
+<ThemedScrollView style={styles.scrollView} contentContainerStyle={styles.contentContainer} />;
+```
+
+`withUnistyles` extracts dependency metadata from both `style` and `contentContainerStyle`, subscribes to the relevant theme/runtime changes, and re-renders only that wrapped component when needed. Its [auto-mapping behavior for `style` and `contentContainerStyle`](https://www.unistyl.es/v3/references/with-unistyles#auto-mapping-for-style-and-contentcontainerstyle-props) is the reason it fixes themed `ScrollView` content containers. Reach for it when wrapper-view layout would be awkward or when a third-party component needs theme-aware non-`style` props mapped through Unistyles.
+
+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 }}`. `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.
+
+Use `useUnistyles()` inside the component instead:
+
+```tsx
+const { theme } = useUnistyles();
+
+<ChevronDown size={theme.iconSize.md} color={theme.colors.foregroundMuted} />;
+```
+
+Importing `baseColors`, theme-name constants, or `type Theme` is fine when the value is intentionally static or type-only.
+
+## 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)

flake.lock

@@ -0,0 +1,27 @@
+{
+  "nodes": {
+    "nixpkgs": {
+      "locked": {
+        "lastModified": 1772963539,
+        "narHash": "sha256-9jVDGZnvCckTGdYT53d/EfznygLskyLQXYwJLKMPsZs=",
+        "owner": "NixOS",
+        "repo": "nixpkgs",
+        "rev": "9dcb002ca1690658be4a04645215baea8b95f31d",
+        "type": "github"
+      },
+      "original": {
+        "owner": "NixOS",
+        "ref": "nixos-unstable",
+        "repo": "nixpkgs",
+        "type": "github"
+      }
+    },
+    "root": {
+      "inputs": {
+        "nixpkgs": "nixpkgs"
+      }
+    }
+  },
+  "root": "root",
+  "version": 7
+}

flake.nix

@@ -0,0 +1,59 @@
+{
+  description = "Paseo - self-hosted daemon for AI coding agents";
+
+  inputs = {
+    nixpkgs.url = "github:NixOS/nixpkgs/nixos-unstable";
+  };
+
+  outputs =
+    {
+      self,
+      nixpkgs,
+    }:
+    let
+      supportedSystems = [
+        "x86_64-linux"
+        "aarch64-linux"
+        "x86_64-darwin"
+        "aarch64-darwin"
+      ];
+      forAllSystems = nixpkgs.lib.genAttrs supportedSystems;
+      pkgsFor = system: import nixpkgs { inherit system; };
+    in
+    {
+      packages = forAllSystems (
+        system:
+        let
+          pkgs = pkgsFor system;
+          paseo = pkgs.callPackage ./nix/package.nix { };
+        in
+        {
+          default = paseo;
+          paseo = paseo;
+        }
+      );
+
+      nixosModules.default = self.nixosModules.paseo;
+      nixosModules.paseo =
+        { pkgs, lib, ... }:
+        {
+          imports = [ ./nix/module.nix ];
+          services.paseo.package = lib.mkDefault self.packages.${pkgs.stdenv.hostPlatform.system}.default;
+        };
+
+      devShells = forAllSystems (
+        system:
+        let
+          pkgs = pkgsFor system;
+        in
+        {
+          default = pkgs.mkShell {
+            packages = [
+              pkgs.nodejs_22
+              pkgs.python3
+            ];
+          };
+        }
+      );
+    };
+}

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

@@ -0,0 +1,176 @@
+{
+  config,
+  lib,
+  pkgs,
+  ...
+}:
+
+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";
+
+    package = lib.mkPackageOption pkgs "paseo" { };
+
+    user = lib.mkOption {
+      type = lib.types.str;
+      default = "paseo";
+      description = "User account under which Paseo runs.";
+    };
+
+    group = lib.mkOption {
+      type = lib.types.str;
+      default = "paseo";
+      description = "Group under which Paseo runs.";
+    };
+
+    dataDir = lib.mkOption {
+      type = lib.types.str;
+      default =
+        if cfg.user == "paseo"
+        then "/var/lib/paseo"
+        else "/home/${cfg.user}/.paseo";
+      defaultText = lib.literalExpression ''
+        if cfg.user == "paseo"
+        then "/var/lib/paseo"
+        else "/home/''${cfg.user}/.paseo"
+      '';
+      description = "Directory for Paseo state (PASEO_HOME). Stores agent data, config, and logs.";
+    };
+
+    port = lib.mkOption {
+      type = lib.types.port;
+      default = 6767;
+      description = "Port for the Paseo daemon to listen on.";
+    };
+
+    listenAddress = lib.mkOption {
+      type = lib.types.str;
+      default = "127.0.0.1";
+      description = "Address for the Paseo daemon to bind to.";
+    };
+
+    openFirewall = lib.mkOption {
+      type = lib.types.bool;
+      default = false;
+      description = "Whether to open the firewall for the Paseo daemon port.";
+    };
+
+    hostnames = lib.mkOption {
+      type = lib.types.either (lib.types.enum [ true ]) (lib.types.listOf lib.types.str);
+      default = [ ];
+      example = [ ".example.com" "myhost.local" ];
+      description = ''
+        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
+        (e.g. `".example.com"` matches `example.com` and `foo.example.com`).
+
+        Set to `true` to allow any host (not recommended).
+      '';
+    };
+
+    relay = {
+      enable = lib.mkOption {
+        type = lib.types.bool;
+        default = true;
+        description = "Whether to enable the relay connection for remote access via app.paseo.sh.";
+      };
+    };
+
+    inheritUserEnvironment = lib.mkOption {
+      type = lib.types.bool;
+      default = cfg.user != "paseo";
+      defaultText = lib.literalExpression ''cfg.user != "paseo"'';
+      description = ''
+        Whether to include the user's profile PATH in the service environment.
+
+        When Paseo runs as a real user (not the default system user), AI agents
+        need access to the user's tools (git, ssh, etc.). This adds the user's
+        NixOS profile and system paths so agents can use them without manually
+        setting PATH.
+
+        Enabled by default when `user` is set to a non-default value.
+      '';
+    };
+
+    environment = lib.mkOption {
+      type = lib.types.attrsOf lib.types.str;
+      default = { };
+      example = lib.literalExpression ''
+        {
+          PASEO_RELAY_ENDPOINT = "relay.paseo.sh:443";
+        }
+      '';
+      description = "Extra environment variables for the Paseo daemon.";
+    };
+  };
+
+  config = lib.mkIf cfg.enable {
+    users.users.${cfg.user} = lib.mkIf (cfg.user == "paseo") {
+      isSystemUser = true;
+      group = cfg.group;
+      home = cfg.dataDir;
+    };
+
+    users.groups.${cfg.group} = lib.mkIf (cfg.group == "paseo") { };
+
+    systemd.tmpfiles.rules = [
+      "d ${cfg.dataDir} 0700 ${cfg.user} ${cfg.group} - -"
+    ];
+
+    systemd.services.paseo = {
+      description = "Paseo - self-hosted daemon for AI coding agents";
+      after = [ "network.target" ];
+      wantedBy = [ "multi-user.target" ];
+
+      environment = {
+        NODE_ENV = "production";
+        PASEO_HOME = cfg.dataDir;
+        PASEO_LISTEN = "${cfg.listenAddress}:${toString cfg.port}";
+      } // lib.optionalAttrs cfg.inheritUserEnvironment {
+        # mkForce overrides the default PATH from NixOS's systemd module (which
+        # only includes store paths for coreutils/grep/sed/systemd). Our PATH
+        # includes /run/current-system/sw/bin which is a superset of those.
+        PATH = lib.mkForce (lib.concatStringsSep ":" [
+          "/etc/profiles/per-user/${cfg.user}/bin"
+          "/run/current-system/sw/bin"
+          "/run/wrappers/bin"
+          "/nix/var/nix/profiles/default/bin"
+        ]);
+      } // 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 = {
+        Type = "simple";
+        User = cfg.user;
+        Group = cfg.group;
+
+        ExecStart =
+          "${cfg.package}/bin/paseo-server"
+          + lib.optionalString (!cfg.relay.enable) " --no-relay";
+
+        Restart = "on-failure";
+        RestartSec = 5;
+
+        # Graceful shutdown (server handles SIGTERM with a 10s timeout)
+        KillSignal = "SIGTERM";
+        TimeoutStopSec = 15;
+      };
+    };
+
+    environment.systemPackages = [ cfg.package ];
+
+    networking.firewall.allowedTCPPorts = lib.mkIf cfg.openFirewall [ cfg.port ];
+  };
+}

nix/package.nix

@@ -0,0 +1,144 @@
+{
+  lib,
+  stdenv,
+  buildNpmPackage,
+  nodejs_22,
+  python3,
+  makeWrapper,
+  # node-pty needs libuv headers on Linux
+  libuv,
+}:
+
+buildNpmPackage rec {
+  pname = "paseo";
+  version = (builtins.fromJSON (builtins.readFile ../package.json)).version;
+
+  src = lib.cleanSourceWith {
+    src = ./..;
+    filter = path: type:
+      let
+        baseName = builtins.baseNameOf path;
+        relPath = lib.removePrefix (toString ./..) path;
+      in
+      # Exclude non-daemon workspace contents (keep package.json for workspace resolution)
+      !(lib.hasPrefix "/packages/app/src" relPath)
+      && !(lib.hasPrefix "/packages/app/assets" relPath)
+      && !(lib.hasPrefix "/packages/app/android" relPath)
+      && !(lib.hasPrefix "/packages/app/ios" relPath)
+      && !(lib.hasPrefix "/packages/website/src" relPath)
+      && !(lib.hasPrefix "/packages/website/public" relPath)
+      && !(lib.hasPrefix "/packages/desktop/src" relPath)
+      && !(lib.hasPrefix "/packages/desktop/src-tauri" relPath)
+      # Exclude test fixtures and debug files
+      && !(lib.hasSuffix ".test.ts" baseName)
+      && !(lib.hasSuffix ".e2e.test.ts" baseName)
+      && baseName != "node_modules"
+      && baseName != ".git"
+      && baseName != ".paseo"
+      && baseName != ".DS_Store";
+  };
+
+  nodejs = nodejs_22;
+
+  # 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-J0SsiBBQYhsFzmwCg4NLOLhCXS1vqiegfJqPdoHWSKM=";
+
+  # 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).
+  # We manually rebuild only node-pty in buildPhase.
+  npmRebuildFlags = [ "--ignore-scripts" ];
+
+  nativeBuildInputs = [
+    python3 # for node-gyp (node-pty compilation)
+    makeWrapper
+  ];
+
+  buildInputs = lib.optionals stdenv.hostPlatform.isLinux [
+    libuv
+  ];
+
+  # Don't use the default npm build hook — we need a custom build sequence
+  dontNpmBuild = true;
+
+  buildPhase = ''
+    runHook preBuild
+
+    # Rebuild only node-pty (native addon for terminal emulation).
+    # Speech-related native modules (sherpa-onnx, onnxruntime-node) are
+    # intentionally left unbuilt — they're lazily loaded and gracefully
+    # degrade when unavailable.
+    npm rebuild node-pty
+
+    # Build all daemon packages in dependency order (defined in package.json)
+    npm run build:daemon
+
+    runHook postBuild
+  '';
+
+  installPhase = ''
+    runHook preInstall
+
+    mkdir -p $out/lib/paseo
+
+    # Copy root package metadata
+    cp package.json $out/lib/paseo/
+
+    # Copy node_modules (preserving workspace symlinks)
+    cp -a node_modules $out/lib/paseo/
+
+    # Auto-detect which @getpaseo/* packages were built by build:daemon
+    # (they'll have a dist/ directory). Copy those and remove the rest.
+    for link in $out/lib/paseo/node_modules/@getpaseo/*; do
+      name=$(basename "$link")
+      if [ -d "packages/$name/dist" ]; then
+        mkdir -p "$out/lib/paseo/packages/$name"
+        cp "packages/$name/package.json" "$out/lib/paseo/packages/$name/"
+        cp -a "packages/$name/dist" "$out/lib/paseo/packages/$name/"
+        if [ -d "packages/$name/node_modules" ]; then
+          cp -a "packages/$name/node_modules" "$out/lib/paseo/packages/$name/"
+        fi
+      else
+        rm -f "$link"
+      fi
+    done
+
+    # Copy CLI bin entry
+    mkdir -p $out/lib/paseo/packages/cli/bin
+    cp packages/cli/bin/paseo $out/lib/paseo/packages/cli/bin/
+
+    # Copy extra server files referenced at runtime
+    for f in agent-prompt.md .env.example; do
+      if [ -f packages/server/$f ]; then
+        cp packages/server/$f $out/lib/paseo/packages/server/
+      fi
+    done
+
+    # Copy server scripts (including supervisor-entrypoint) needed by CLI
+    if [ -d packages/server/dist/scripts ]; then
+      mkdir -p $out/lib/paseo/packages/server/dist/scripts
+      cp -a packages/server/dist/scripts/* $out/lib/paseo/packages/server/dist/scripts/
+    fi
+
+    # 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/scripts/supervisor-entrypoint.js" \
+      --set NODE_ENV production
+
+    # Create wrapper for the CLI
+    makeWrapper ${nodejs}/bin/node $out/bin/paseo \
+      --add-flags "$out/lib/paseo/packages/cli/dist/index.js" \
+      --set NODE_PATH "$out/lib/paseo/node_modules"
+
+    runHook postInstall
+  '';
+
+  meta = {
+    description = "Self-hosted daemon for Claude Code, Codex, and OpenCode";
+    homepage = "https://github.com/getpaseo/paseo";
+    license = lib.licenses.agpl3Plus;
+    mainProgram = "paseo";
+    platforms = lib.platforms.linux ++ lib.platforms.darwin;
+  };
+}

package.json

@@ -1,8 +1,29 @@
 {
   "name": "paseo",
-  "version": "0.1.24",
+  "version": "0.1.69",
   "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",
     "packages/server",
     "packages/app",
     "packages/relay",
@@ -12,72 +33,82 @@
   ],
   "scripts": {
     "dev": "./scripts/dev.sh",
-    "dev:server": "NODE_ENV=development tsx packages/server/scripts/daemon-runner.ts --dev",
+    "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:daemon": "npm run build --workspace=@getpaseo/relay && npm run build --workspace=@getpaseo/server && npm run build --workspace=@getpaseo/cli",
+    "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": "prettier --write .",
-    "format:check": "prettier --check .",
+    "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",
     "android:production": "npm run android:production --workspace=@getpaseo/app",
     "android:release": "npm run android:production --workspace=@getpaseo/app",
+    "android:clear": "npm run android:clear --workspace=@getpaseo/app",
     "android:clean": "npm run android:clean --workspace=@getpaseo/app",
     "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/relay && npm run typecheck --workspace=@getpaseo/server && npm run typecheck --workspace=@getpaseo/cli && npm run build --workspace=@getpaseo/relay && npm run build --workspace=@getpaseo/server && npm run build --workspace=@getpaseo/cli && 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/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/relay --access public && npm publish --workspace=@getpaseo/server --access public && npm publish --workspace=@getpaseo/cli --access public",
+    "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 version:all:patch && npm run release:check && npm run draft-release:push",
-    "draft-release:minor": "npm run version:all:minor && npm run release:check && npm run draft-release:push",
-    "draft-release:major": "npm run version:all:major && npm run release:check && npm run draft-release:push",
-    "release:finalize": "node scripts/finalize-current-release.mjs",
-    "release:patch": "npm run version:all:patch && npm run release:check && npm run release:publish && npm run release:push",
-    "release:minor": "npm run version:all:minor && npm run release:check && npm run release:publish && npm run release:push",
-    "release:major": "npm run version:all:major && npm run release:check && npm run release:publish && npm run release:push"
+    "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": {
+    "@types/ws": "^8.5.14",
+    "@typescript/native-preview": "7.0.0-dev.20260423.1",
     "concurrently": "^9.2.1",
-    "prettier": "^3.5.3",
+    "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",
     "patch-package": "^8.0.1",
-    "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",
-  "keywords": [
-    "openai",
-    "realtime",
-    "voice",
-    "voice-assistant",
-    "development",
-    "mcp"
-  ],
-  "author": "moboudra",
-  "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"
+    "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/.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/app.config.js

@@ -65,8 +65,7 @@ export default {
     ios: {
       supportsTablet: true,
       infoPlist: {
-        NSMicrophoneUsageDescription:
-          "This app needs access to the microphone for voice commands.",
+        NSMicrophoneUsageDescription: "This app needs access to the microphone for voice commands.",
         ITSAppUsesNonExemptEncryption: false,
       },
       bundleIdentifier: variant.packageId,
@@ -92,14 +91,15 @@ export default {
         "android.permission.CAMERA",
       ],
       package: variant.packageId,
-      ...(variant.googleServicesFile
-        ? { googleServicesFile: variant.googleServicesFile }
-        : {}),
+      ...(variant.googleServicesFile ? { googleServicesFile: variant.googleServicesFile } : {}),
     },
     web: {
       output: "single",
       favicon: "./assets/images/favicon.png",
     },
+    autolinking: {
+      searchPaths: ["../../node_modules", "./node_modules"],
+    },
     plugins: [
       "expo-router",
       [
@@ -143,6 +143,7 @@ export default {
     experiments: {
       typedRoutes: true,
       reactCompiler: true,
+      autolinkingModuleResolution: true,
     },
     extra: {
       router: {},

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

@@ -1,183 +0,0 @@
-import { test, expect, type Page } from "./fixtures";
-import { createTempGitRepo } from "./helpers/workspace";
-import {
-  connectDaemonClient,
-  createReplyTurn,
-  expectDetachedFromBottom,
-  expectNearBottom,
-  getChatContainerKey,
-  readScrollMetrics,
-  scrollUpFromBottom,
-  seedBottomAnchorAgent,
-  waitForAgentReady,
-  waitForContentGrowth,
-} from "./helpers/agent-bottom-anchor";
-
-test.describe.configure({ timeout: 180000 });
-
-async function openWorkspaceAgentTab(page: Page, agentId: string) {
-  const tab = page.getByTestId(`workspace-tab-agent_${agentId}`).first();
-  await expect(tab).toBeVisible({ timeout: 30000 });
-  await tab.click();
-}
-
-test("direct load and refresh land at the bottom for history-backed chats", async ({
-  page,
-}) => {
-  const repo = await createTempGitRepo("paseo-e2e-bottom-anchor-direct-");
-  const client = await connectDaemonClient();
-
-  try {
-    const agent = await seedBottomAnchorAgent({
-      client,
-      cwd: repo.path,
-      title: `bottom-anchor-direct-${Date.now()}`,
-      turnCount: 4,
-    });
-
-    await page.goto(agent.url, { waitUntil: "domcontentloaded" });
-    await openWorkspaceAgentTab(page, agent.id);
-    await waitForAgentReady(page, agent.expectedTailText);
-    await expectNearBottom(page);
-
-    await page.reload({ waitUntil: "commit" });
-    await openWorkspaceAgentTab(page, agent.id);
-    await waitForAgentReady(page, agent.expectedTailText);
-    await expectNearBottom(page);
-  } finally {
-    await client.close().catch(() => undefined);
-    await repo.cleanup();
-  }
-});
-
-test("revisiting a loaded chat restores bottom anchoring", async ({
-  page,
-}) => {
-  const repo = await createTempGitRepo("paseo-e2e-bottom-anchor-switch-");
-  const client = await connectDaemonClient();
-
-  try {
-    const agent = await seedBottomAnchorAgent({
-      client,
-      cwd: repo.path,
-      title: `bottom-anchor-switch-${Date.now()}`,
-      turnCount: 4,
-    });
-
-    await page.goto(agent.url, { waitUntil: "domcontentloaded" });
-    await openWorkspaceAgentTab(page, agent.id);
-    await waitForAgentReady(page, agent.expectedTailText);
-    await expectNearBottom(page);
-
-    await page.getByTestId("sidebar-new-agent").first().click();
-    await expect(page.getByRole("textbox", { name: "Message agent..." }).first()).toBeVisible({
-      timeout: 30000,
-    });
-
-    await openWorkspaceAgentTab(page, agent.id);
-    await waitForAgentReady(page, agent.expectedTailText);
-    await expectNearBottom(page);
-  } finally {
-    await client.close().catch(() => undefined);
-    await repo.cleanup();
-  }
-});
-
-test("sticky mode stays pinned through composer growth and viewport resize, but detached mode does not fight streamed updates", async ({
-  page,
-}) => {
-  const repo = await createTempGitRepo("paseo-e2e-bottom-anchor-sticky-");
-  const client = await connectDaemonClient();
-
-  try {
-    const agent = await seedBottomAnchorAgent({
-      client,
-      cwd: repo.path,
-      title: `bottom-anchor-sticky-${Date.now()}`,
-      turnCount: 10,
-    });
-
-    await page.setViewportSize({ width: 1320, height: 920 });
-    await page.goto(agent.url, { waitUntil: "domcontentloaded" });
-    await openWorkspaceAgentTab(page, agent.id);
-    await waitForAgentReady(page, agent.expectedTailText);
-    await expectNearBottom(page);
-
-    const composer = page.getByRole("textbox", { name: "Message agent..." }).first();
-    await composer.click();
-    for (let index = 0; index < 6; index += 1) {
-      await composer.pressSequentially(`composer growth line ${index + 1}`);
-      if (index < 5) {
-        await page.keyboard.press("Shift+Enter");
-      }
-    }
-    await expectNearBottom(page);
-    await expect(page.getByTestId("scroll-to-bottom-button")).toHaveCount(0);
-
-    await page.setViewportSize({ width: 820, height: 760 });
-    await expectNearBottom(page);
-
-    await scrollUpFromBottom(page, 720);
-    await expectDetachedFromBottom(page);
-    const beforeExternalUpdate = await readScrollMetrics(page);
-
-    const externalTurn = createReplyTurn(`external-stream-${Date.now()}`);
-    await client.sendAgentMessage(agent.id, externalTurn.message);
-    await waitForContentGrowth(page, beforeExternalUpdate.contentHeight);
-    const finish = await client.waitForFinish(agent.id, 120000);
-    expect(finish.status).toBe("idle");
-    await expectDetachedFromBottom(page);
-  } finally {
-    await client.close().catch(() => undefined);
-    await repo.cleanup();
-  }
-});
-
-test("web partial virtualization keeps bottom anchoring stable across direct load, refresh, and resize", async ({
-  page,
-}) => {
-  await page.addInitScript(() => {
-    (window as typeof window & {
-      __PASEO_E2E_WEB_PARTIAL_VIRTUALIZATION_THRESHOLD?: number;
-      __PASEO_E2E_WEB_MOUNTED_RECENT_STREAM_ITEMS?: number;
-    }).__PASEO_E2E_WEB_PARTIAL_VIRTUALIZATION_THRESHOLD = 6;
-    (window as typeof window & {
-      __PASEO_E2E_WEB_PARTIAL_VIRTUALIZATION_THRESHOLD?: number;
-      __PASEO_E2E_WEB_MOUNTED_RECENT_STREAM_ITEMS?: number;
-    }).__PASEO_E2E_WEB_MOUNTED_RECENT_STREAM_ITEMS = 4;
-  });
-
-  const repo = await createTempGitRepo("paseo-e2e-bottom-anchor-virtualized-");
-  const client = await connectDaemonClient();
-
-  try {
-    const agent = await seedBottomAnchorAgent({
-      client,
-      cwd: repo.path,
-      title: `bottom-anchor-virtualized-${Date.now()}`,
-      turnCount: 4,
-    });
-
-    await page.goto(agent.url, { waitUntil: "domcontentloaded" });
-    await openWorkspaceAgentTab(page, agent.id);
-    await waitForAgentReady(page, agent.expectedTailText);
-    await expect
-      .poll(async () => await getChatContainerKey(page))
-      .toBe("web-partial-virtualized");
-    await expectNearBottom(page);
-
-    await page.reload({ waitUntil: "commit" });
-    await openWorkspaceAgentTab(page, agent.id);
-    await waitForAgentReady(page, agent.expectedTailText);
-    await expect
-      .poll(async () => await getChatContainerKey(page))
-      .toBe("web-partial-virtualized");
-    await expectNearBottom(page);
-
-    await page.setViewportSize({ width: 780, height: 720 });
-    await expectNearBottom(page);
-  } finally {
-    await client.close().catch(() => undefined);
-    await repo.cleanup();
-  }
-});

packages/app/e2e/agent-details-sheet.spec.ts

@@ -1,35 +0,0 @@
-import { test, expect } from "./fixtures";
-import { createAgent, ensureHostSelected, gotoHome, setWorkingDirectory } from "./helpers/app";
-import { createTempGitRepo } from "./helpers/workspace";
-
-test("agent details sheet shows IDs and copy toast", async ({ page }) => {
-  test.setTimeout(120_000);
-  const repo = await createTempGitRepo();
-  const prompt = "Respond with exactly: Hello";
-
-  try {
-    await gotoHome(page);
-    await setWorkingDirectory(page, repo.path);
-    await ensureHostSelected(page);
-    await createAgent(page, prompt);
-
-    await page.getByTestId("agent-overflow-menu").click();
-    await expect(page.getByTestId("agent-details-sheet")).toBeVisible();
-
-    await expect(page.getByTestId("agent-details-agent-id")).toBeVisible();
-    await expect(page.getByTestId("agent-details-agent-id-value")).not.toHaveText(
-      "Not available"
-    );
-
-    await expect(page.getByTestId("agent-details-persistence-session-id")).toBeVisible();
-    await expect(
-      page.getByTestId("agent-details-persistence-session-id-value")
-    ).not.toHaveText("Not available", { timeout: 90_000 });
-
-    await page.getByTestId("agent-details-agent-id").click();
-    await expect(page.getByTestId("app-toast")).toBeVisible();
-    await expect(page.getByTestId("app-toast-message")).toHaveText(/copied/i);
-  } finally {
-    await repo.cleanup();
-  }
-});

packages/app/e2e/agent-scroll-submit-firefox.spec.ts

@@ -1,197 +0,0 @@
-import { expect, test, type Page } from "@playwright/test";
-
-const SERVER_ID =
-  process.env.PLAYWRIGHT_REPRO_SERVER_ID ?? "srv_ETXtcjYRGrCI";
-const AGENT_ID =
-  process.env.PLAYWRIGHT_REPRO_AGENT_ID ??
-  "3533e6c3-c0b3-4310-b85c-eb07cbb501a0";
-const APP_BASE_URL = process.env.PLAYWRIGHT_REPRO_BASE_URL ?? "http://localhost:8081";
-const DAEMON_ENDPOINT =
-  process.env.PLAYWRIGHT_REPRO_DAEMON_ENDPOINT ?? "127.0.0.1:6767";
-const SUBMIT_TEXT = process.env.PLAYWRIGHT_REPRO_MESSAGE ?? "hello";
-const AGENT_URL = `${APP_BASE_URL}/h/${SERVER_ID}/agent/${AGENT_ID}`;
-const NEAR_BOTTOM_THRESHOLD_PX = 64;
-
-type ScrollMetrics = {
-  offsetY: number;
-  contentHeight: number;
-  viewportHeight: number;
-  distanceFromBottom: number;
-};
-
-test.use({ browserName: "firefox" });
-
-function seedDaemonRegistryScript(params: {
-  serverId: string;
-  endpoint: string;
-  nowIso: string;
-}) {
-  const daemon = {
-    serverId: params.serverId,
-    label: "localhost",
-    connections: [
-      {
-        id: `direct:${params.endpoint}`,
-        type: "direct",
-        endpoint: params.endpoint,
-      },
-    ],
-    preferredConnectionId: `direct:${params.endpoint}`,
-    createdAt: params.nowIso,
-    updatedAt: params.nowIso,
-  };
-
-  localStorage.setItem("@paseo:e2e", "1");
-  localStorage.setItem("@paseo:daemon-registry", JSON.stringify([daemon]));
-  localStorage.setItem(
-    "@paseo:create-agent-preferences",
-    JSON.stringify({
-      serverId: params.serverId,
-      provider: "codex",
-      providerPreferences: {
-        claude: { model: "haiku" },
-        codex: { model: "gpt-5.1-codex-mini", thinkingOptionId: "low" },
-      },
-    })
-  );
-}
-
-async function readScrollMetrics(page: Page): Promise<ScrollMetrics> {
-  return page.getByTestId("agent-chat-scroll").evaluate((root: Element) => {
-    const rootElement = root as HTMLElement;
-    const candidates = [rootElement, ...Array.from(rootElement.querySelectorAll("*"))];
-    const scrollElement =
-      candidates.find(
-        (element) =>
-          element instanceof HTMLElement &&
-          element.scrollHeight - element.clientHeight > 1
-      ) ?? rootElement;
-
-    const offsetY = Math.max(0, scrollElement.scrollTop);
-    const contentHeight = Math.max(0, scrollElement.scrollHeight);
-    const viewportHeight = Math.max(0, scrollElement.clientHeight);
-    const distanceFromBottom = Math.max(
-      0,
-      contentHeight - (offsetY + viewportHeight)
-    );
-
-    return {
-      offsetY,
-      contentHeight,
-      viewportHeight,
-      distanceFromBottom,
-    };
-  });
-}
-
-async function scrollUpFromBottom(
-  page: Page,
-  pixels: number
-): Promise<void> {
-  await page.getByTestId("agent-chat-scroll").evaluate(
-    (root: Element, amount: number) => {
-      const rootElement = root as HTMLElement;
-      const candidates = [
-        rootElement,
-        ...Array.from(rootElement.querySelectorAll("*")),
-      ];
-      const scrollElement =
-        candidates.find(
-          (element) =>
-            element instanceof HTMLElement &&
-            element.scrollHeight - element.clientHeight > 1
-        ) ?? rootElement;
-
-      const bottomOffset = Math.max(
-        0,
-        scrollElement.scrollHeight - scrollElement.clientHeight
-      );
-      scrollElement.scrollTop = Math.max(0, bottomOffset - amount);
-    },
-    pixels
-  );
-}
-
-test("repro: submit while scrolled up should stay anchored to bottom (Firefox)", async ({
-  page,
-}, testInfo) => {
-  const userAgent = await page.evaluate(() => navigator.userAgent);
-  expect(userAgent).toContain("Firefox");
-
-  await page.addInitScript(seedDaemonRegistryScript, {
-    serverId: SERVER_ID,
-    endpoint: DAEMON_ENDPOINT,
-    nowIso: new Date().toISOString(),
-  });
-
-  await page.goto(AGENT_URL, { waitUntil: "domcontentloaded" });
-  await expect(page.getByTestId("agent-chat-scroll")).toBeVisible({
-    timeout: 60_000,
-  });
-  await expect(page.getByRole("textbox", { name: "Message agent..." })).toBeVisible({
-    timeout: 60_000,
-  });
-
-  // Require enough history so the repro actually scrolls away from bottom.
-  await expect
-    .poll(async () => {
-      const metrics = await readScrollMetrics(page);
-      return Math.max(0, metrics.contentHeight - metrics.viewportHeight);
-    })
-    .toBeGreaterThan(300);
-
-  await scrollUpFromBottom(page, 900);
-  await page.waitForTimeout(250);
-  const beforeSubmit = await readScrollMetrics(page);
-  const beforePath = testInfo.outputPath("before-submit.png");
-  await page.screenshot({ path: beforePath, fullPage: true });
-  await testInfo.attach("before-submit", {
-    path: beforePath,
-    contentType: "image/png",
-  });
-
-  await page.getByRole("textbox", { name: "Message agent..." }).fill(SUBMIT_TEXT);
-  await page.getByRole("textbox", { name: "Message agent..." }).press("Enter");
-  await page.waitForTimeout(1200);
-
-  const afterSubmit = await readScrollMetrics(page);
-  const afterPath = testInfo.outputPath("after-submit.png");
-  await page.screenshot({ path: afterPath, fullPage: true });
-  await testInfo.attach("after-submit", {
-    path: afterPath,
-    contentType: "image/png",
-  });
-  await testInfo.attach("firefox-scroll-metrics", {
-    body: JSON.stringify(
-      {
-        url: AGENT_URL,
-        submitText: SUBMIT_TEXT,
-        thresholdPx: NEAR_BOTTOM_THRESHOLD_PX,
-        beforeScreenshot: beforePath,
-        afterScreenshot: afterPath,
-        beforeSubmit,
-        afterSubmit,
-      },
-      null,
-      2
-    ),
-    contentType: "application/json",
-  });
-
-  console.log(
-    `[firefox-scroll-repro] ${JSON.stringify(
-      {
-        submitText: SUBMIT_TEXT,
-        beforeSubmit,
-        afterSubmit,
-      },
-      null,
-      2
-    )}`
-  );
-
-  expect(beforeSubmit.distanceFromBottom).toBeGreaterThan(NEAR_BOTTOM_THRESHOLD_PX);
-  expect(afterSubmit.distanceFromBottom).toBeLessThanOrEqual(
-    NEAR_BOTTOM_THRESHOLD_PX
-  );
-});

packages/app/e2e/agent-timeline-hydration.spec.ts

@@ -1,27 +0,0 @@
-import { test, expect } from './fixtures';
-import { createAgent, ensureHostSelected, gotoHome, setWorkingDirectory } from './helpers/app';
-import { createTempGitRepo } from './helpers/workspace';
-
-test('agent timeline hydrates after reload via fetch_agent_timeline_request', async ({ page }) => {
-  const repo = await createTempGitRepo();
-  const prompt = 'Respond with exactly: TIMELINE_HYDRATION_OK';
-
-  try {
-    await gotoHome(page);
-    await setWorkingDirectory(page, repo.path);
-    await ensureHostSelected(page);
-    await createAgent(page, prompt);
-    await expect(page.getByText(prompt, { exact: true }).first()).toBeVisible({
-      timeout: 30000,
-    });
-
-    await page.reload({ waitUntil: 'commit' });
-    await expect(page).toHaveURL(/\/agent\//);
-    await expect(page.getByTestId('agent-loading')).toHaveCount(0, { timeout: 30000 });
-    await expect(page.getByText(prompt, { exact: true }).first()).toBeVisible({
-      timeout: 30000,
-    });
-  } finally {
-    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/checkout-ship.spec.ts

@@ -1,440 +0,0 @@
-import path from 'node:path';
-import { appendFile, mkdtemp, rm, writeFile, realpath } from 'node:fs/promises';
-import { execSync } from 'node:child_process';
-import { tmpdir } from 'node:os';
-import { test, expect, type Page } from './fixtures';
-import {
-  createAgent,
-  ensureHostSelected,
-  gotoHome,
-  setWorkingDirectory,
-} from './helpers/app';
-import { createTempGitRepo } from './helpers/workspace';
-
-test.describe.configure({ mode: 'serial', timeout: 120000 });
-
-function escapeRegex(value: string): string {
-  return value.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
-}
-
-function getChangesScope(page: Page) {
-  return page.locator('[data-testid="explorer-content-area"]:visible').first();
-}
-
-function getChangesHeader(page: Page) {
-  return getChangesScope(page).getByTestId('changes-header');
-}
-
-async function ensureExplorerTabsVisible(page: Page) {
-  const changesTab = page.getByTestId('explorer-tab-changes').first();
-  if (await changesTab.isVisible().catch(() => false)) {
-    return;
-  }
-
-  const toggle = page
-    .getByRole('button', { name: /open explorer|close explorer|toggle explorer/i })
-    .first();
-  await expect(toggle).toBeVisible({ timeout: 10000 });
-  for (let attempt = 0; attempt < 4; attempt += 1) {
-    if (await changesTab.isVisible().catch(() => false)) {
-      return;
-    }
-    await toggle.click();
-    await page.waitForTimeout(200);
-  }
-  await expect(changesTab).toBeVisible({ timeout: 30000 });
-}
-
-async function selectChangesView(page: Page, view: 'working' | 'base') {
-  // Defensive: close any open dropdown menus (their backdrops intercept clicks).
-  const primaryBackdrop = page.getByTestId('changes-primary-cta-menu-backdrop');
-  if (await primaryBackdrop.isVisible().catch(() => false)) {
-    await primaryBackdrop.click({ force: true });
-    await expect(primaryBackdrop).toHaveCount(0);
-  }
-  const overflowBackdrop = page.getByTestId('changes-overflow-content-backdrop');
-  if (await overflowBackdrop.isVisible().catch(() => false)) {
-    await overflowBackdrop.click({ force: true });
-    await expect(overflowBackdrop).toHaveCount(0);
-  }
-  const diffModeBackdrop = page.getByTestId('changes-diff-status-menu-backdrop');
-  if (await diffModeBackdrop.isVisible().catch(() => false)) {
-    await diffModeBackdrop.click({ force: true });
-    await expect(diffModeBackdrop).toHaveCount(0);
-  }
-
-  const scope = getChangesScope(page);
-  const modeToggle = scope.getByTestId('changes-diff-status').first();
-  const expected = view === 'working' ? 'Uncommitted' : 'Committed';
-  if (!(await modeToggle.isVisible().catch(() => false))) {
-    return;
-  }
-  const current = ((await modeToggle.innerText().catch(() => '')) ?? '').trim();
-  if (current !== expected) {
-    await modeToggle.click();
-    const menu = page.getByTestId('changes-diff-status-menu');
-    await expect(menu).toBeVisible({ timeout: 10000 });
-    const optionTestId =
-      view === 'working' ? 'changes-diff-mode-uncommitted' : 'changes-diff-mode-committed';
-    await page.getByTestId(optionTestId).click({ force: true });
-  }
-  await expect(modeToggle).toContainText(expected, { timeout: 10000 });
-}
-
-async function openChangesPrimaryMenu(page: Page) {
-  const scope = getChangesScope(page);
-  const caret = scope.getByTestId('changes-primary-cta-caret').first();
-  await expect(caret).toBeVisible();
-  await caret.click();
-  // Menu content is rendered via a portal, so don't scope it to the explorer content area.
-  await expect(page.getByTestId('changes-primary-cta-menu')).toBeVisible();
-}
-
-async function openChangesPanel(page: Page, options?: { expectGit?: boolean }) {
-  await ensureExplorerTabsVisible(page);
-  const changesHeader = getChangesHeader(page);
-  if (!(await changesHeader.isVisible())) {
-    await page.getByTestId('explorer-tab-changes').first().click();
-  }
-  await expect(changesHeader).toBeVisible({ timeout: 30000 });
-  if (options?.expectGit === false) {
-    return;
-  }
-  const changesScope = getChangesScope(page);
-  await expect(changesScope.getByTestId('changes-not-git')).toHaveCount(0, {
-    timeout: 30000,
-  });
-  await expect(changesScope.getByTestId('changes-branch')).not.toHaveText('Not a git repository', {
-    timeout: 30000,
-  });
-}
-
-async function waitForAgentTurnToSettle(page: Page, timeout = 90000) {
-  const stopButton = page.getByRole('button', { name: /stop agent|stop/i }).first();
-  if (!(await stopButton.isVisible().catch(() => false))) {
-    return;
-  }
-  await expect(stopButton).not.toBeVisible({ timeout });
-}
-
-async function createAgentAndWait(page: Page, message: string) {
-  await createAgent(page, message);
-}
-
-async function selectAttachWorktree(page: Page, branchName: string) {
-  const trigger = page.getByTestId('worktree-select-trigger').first();
-  await expect(trigger).toBeVisible({ timeout: 30000 });
-  await trigger.click();
-
-  const menu = page.getByTestId('combobox-desktop-container').first();
-  await expect(menu).toBeVisible({ timeout: 10000 });
-  const searchInput = page.getByRole('textbox', { name: /search worktrees/i }).first();
-  if (await searchInput.isVisible().catch(() => false)) {
-    await searchInput.fill(branchName);
-  }
-
-  const preferredOption = menu
-    .getByText(new RegExp(`^${escapeRegex(branchName)}$`, 'i'))
-    .first();
-  await expect(preferredOption).toBeVisible({ timeout: 10000 });
-  await preferredOption.click({ force: true });
-  await expect(menu).toHaveCount(0);
-  await expect(trigger).toContainText(branchName, { timeout: 30000 });
-}
-
-async function enableCreateWorktree(page: Page) {
-  const trigger = page.getByTestId('worktree-select-trigger').first();
-  await expect(trigger).toBeVisible({ timeout: 30000 });
-
-  const currentValue = ((await trigger.innerText().catch(() => '')) ?? '').trim();
-  if (/Create new worktree/i.test(currentValue)) {
-    await expect(page.getByTestId('worktree-base-branch-trigger')).toBeVisible({
-      timeout: 30000,
-    });
-    return;
-  }
-
-  await trigger.click();
-  const menu = page.getByTestId('combobox-desktop-container').first();
-  await expect(menu).toBeVisible({ timeout: 10000 });
-  const createOption = menu.getByText('Create new worktree', { exact: true }).first();
-  await expect(createOption).toBeVisible({ timeout: 10000 });
-  await createOption.click({ force: true });
-  await expect(menu).toHaveCount(0);
-  await expect(trigger).toContainText('Create new worktree', { timeout: 30000 });
-  await expect(page.getByTestId('worktree-base-branch-trigger')).toBeVisible({
-    timeout: 30000,
-  });
-}
-
-async function refreshUncommittedMode(page: Page) {
-  await selectChangesView(page, 'base');
-  await selectChangesView(page, 'working');
-}
-
-async function refreshChangesTab(page: Page) {
-  await ensureExplorerTabsVisible(page);
-  await page.getByTestId('explorer-tab-files').first().click();
-  await page.getByTestId('explorer-tab-changes').first().click();
-}
-
-function normalizeTmpPath(value: string) {
-  if (value.startsWith('/var/')) {
-    return `/private${value}`;
-  }
-  return value;
-}
-
-type GitWorktreeEntry = {
-  worktreePath: string;
-  branchRef: string | null;
-};
-
-function parseGitWorktreeList(raw: string): GitWorktreeEntry[] {
-  const blocks = raw
-    .split(/\n\s*\n/g)
-    .map((block) => block.trim())
-    .filter((block) => block.length > 0);
-
-  const entries: GitWorktreeEntry[] = [];
-  for (const block of blocks) {
-    const worktreeMatch = block.match(/^worktree (.+)$/m);
-    if (!worktreeMatch) {
-      continue;
-    }
-    const branchMatch = block.match(/^branch (.+)$/m);
-    entries.push({
-      worktreePath: worktreeMatch[1].trim(),
-      branchRef: branchMatch ? branchMatch[1].trim() : null,
-    });
-  }
-  return entries;
-}
-
-async function waitForCreatedWorktree(repoPath: string, timeoutMs = 30000) {
-  const deadline = Date.now() + timeoutMs;
-  const normalizedRepoPath = normalizeTmpPath(repoPath);
-
-  while (Date.now() < deadline) {
-    try {
-      const output = execSync('git worktree list --porcelain', {
-        cwd: repoPath,
-        encoding: 'utf8',
-      });
-      const entries = parseGitWorktreeList(output);
-      const candidate = entries.find((entry) => {
-        const normalizedWorktreePath = normalizeTmpPath(entry.worktreePath);
-        if (normalizedWorktreePath === normalizedRepoPath) {
-          return false;
-        }
-        if (!entry.branchRef) {
-          return false;
-        }
-        return !/\/main$/i.test(entry.branchRef);
-      });
-
-      if (candidate) {
-        const branchName = candidate.branchRef?.split('/').filter(Boolean).pop();
-        if (branchName) {
-          return {
-            worktreePath: candidate.worktreePath,
-            branchName,
-          };
-        }
-      }
-    } catch {
-      // Ignore transient git worktree read errors while polling.
-    }
-    await new Promise((resolve) => setTimeout(resolve, 250));
-  }
-
-  throw new Error(`Timed out waiting for a non-main worktree under ${repoPath}`);
-}
-
-test('checkout-first Changes panel ship loop', async ({ page }) => {
-  const repo = await createTempGitRepo('paseo-e2e-', { withRemote: true });
-  const nonGitDir = await mkdtemp(path.join(tmpdir(), 'paseo-e2e-non-git-'));
-
-  try {
-    await gotoHome(page);
-    await setWorkingDirectory(page, repo.path);
-    await ensureHostSelected(page);
-
-    await enableCreateWorktree(page);
-    await createAgentAndWait(page, 'Respond with exactly: READY');
-    await waitForAgentTurnToSettle(page);
-
-    await openChangesPanel(page);
-    const branchLabelLocator = getChangesScope(page).getByTestId('changes-branch');
-    await expect
-      .poll(async () => (await branchLabelLocator.innerText()).trim(), { timeout: 30000 })
-      .not.toBe('Unknown');
-    const branchNameFromUi = (await branchLabelLocator.innerText()).trim();
-    expect(branchNameFromUi.length).toBeGreaterThan(0);
-
-    const { worktreePath: firstCwd, branchName: worktreeBranch } = await waitForCreatedWorktree(
-      repo.path
-    );
-    expect(worktreeBranch.length).toBeGreaterThan(0);
-    const [resolvedCwd, resolvedRepo] = await Promise.all([
-      realpath(firstCwd).catch(() => firstCwd),
-      realpath(repo.path).catch(() => repo.path),
-    ]);
-    const normalizedRepo = normalizeTmpPath(resolvedRepo);
-    const normalizedCwd = normalizeTmpPath(resolvedCwd);
-    const expectedMarker = `${path.sep}worktrees${path.sep}`;
-    expect(normalizedCwd.includes(expectedMarker)).toBeTruthy();
-
-    await page.getByTestId('sidebar-new-agent').click();
-    await expect(page).toHaveURL(/\/h\/[^/]+\/agent(\?|$)/);
-
-    await setWorkingDirectory(page, repo.path);
-    await ensureHostSelected(page);
-    await selectAttachWorktree(page, worktreeBranch);
-    await createAgentAndWait(page, 'Respond with exactly: READY2');
-    await waitForAgentTurnToSettle(page);
-    await openChangesPanel(page);
-
-    const readmePath = path.join(firstCwd, 'README.md');
-    await appendFile(readmePath, '\nFirst change\n');
-
-    await refreshUncommittedMode(page);
-    await expect(getChangesScope(page).getByText('README.md', { exact: true })).toBeVisible({
-      timeout: 30000,
-    });
-    await getChangesScope(page).getByTestId('diff-file-0-toggle').first().click();
-    await expect(page.getByText('First change')).toBeVisible();
-    const primaryCta = getChangesScope(page).getByTestId('changes-primary-cta').first();
-    await expect(primaryCta).toBeVisible();
-    await expect(primaryCta).toContainText('Commit');
-
-    await primaryCta.click();
-    await expect
-      .poll(() => {
-        try {
-          return execSync('git status --porcelain', {
-            cwd: firstCwd,
-            encoding: 'utf8',
-            env: { ...process.env, GIT_OPTIONAL_LOCKS: '0' },
-          }).trim();
-        } catch {
-          return null;
-        }
-      }, { timeout: 30000 })
-      .toBe('');
-    await openChangesPanel(page);
-    await selectChangesView(page, 'working');
-    await expect(getChangesScope(page).getByText('No uncommitted changes')).toBeVisible({
-      timeout: 30000,
-    });
-    await expect(getChangesScope(page).getByTestId('changes-primary-cta')).not.toContainText('Commit');
-
-    await selectChangesView(page, 'base');
-    await expect(getChangesScope(page).getByText('README.md', { exact: true })).toBeVisible({
-      timeout: 30000,
-    });
-
-    // Push once from the menu so the branch has an origin/<branch> ref.
-    await openChangesPrimaryMenu(page);
-    await page.getByTestId('changes-menu-push').click();
-    await expect
-      .poll(() => {
-        try {
-          execSync(`git show-ref --verify --quiet refs/remotes/origin/${worktreeBranch}`, { cwd: firstCwd });
-          return true;
-        } catch {
-          return false;
-        }
-      }, { timeout: 30000 })
-      .toBe(true);
-
-    const notesPath = path.join(firstCwd, 'notes.txt');
-    await writeFile(notesPath, 'Second change\n');
-
-    await refreshUncommittedMode(page);
-    await refreshChangesTab(page);
-    await expect(getChangesScope(page).getByText('notes.txt', { exact: true })).toBeVisible({
-      timeout: 30000,
-    });
-    await expect(getChangesScope(page).getByText('README.md', { exact: true })).toHaveCount(0);
-    await expect(getChangesScope(page).getByTestId('changes-primary-cta')).toContainText('Commit');
-
-    await selectChangesView(page, 'base');
-    await expect(getChangesScope(page).getByText('README.md', { exact: true })).toBeVisible({
-      timeout: 30000,
-    });
-
-    await getChangesScope(page).getByTestId('changes-primary-cta').click();
-    await expect
-      .poll(() => {
-        try {
-          return execSync('git status --porcelain', {
-            cwd: firstCwd,
-            encoding: 'utf8',
-            env: { ...process.env, GIT_OPTIONAL_LOCKS: '0' },
-          }).trim();
-        } catch {
-          return null;
-        }
-      }, { timeout: 30000 })
-      .toBe('');
-    await openChangesPanel(page);
-    await selectChangesView(page, 'working');
-    await expect(getChangesScope(page).getByText('No uncommitted changes')).toBeVisible({ timeout: 30000 });
-    await expect(getChangesScope(page).getByTestId('changes-primary-cta')).not.toContainText('Commit');
-
-    await selectChangesView(page, 'base');
-    await expect(getChangesScope(page).getByText('README.md', { exact: true })).toBeVisible({
-      timeout: 30000,
-    });
-    await expect(getChangesScope(page).getByText('notes.txt', { exact: true })).toBeVisible({
-      timeout: 30000,
-    });
-
-    // Push is now the primary action (origin/<branch> exists and we're ahead of it).
-    const pushPrimary = getChangesScope(page).getByTestId('changes-primary-cta').first();
-    await expect(pushPrimary).toContainText(/push/i, { timeout: 30000 });
-    await pushPrimary.click();
-    // Regression check: the primary CTA stays in place while pushing.
-    await expect(pushPrimary).toBeVisible();
-    await page.waitForTimeout(50);
-    await expect(pushPrimary).toBeVisible();
-
-    await expect
-      .poll(() => {
-        try {
-          const count = execSync(
-            `git rev-list --count origin/${worktreeBranch}..${worktreeBranch}`,
-            { cwd: firstCwd, encoding: 'utf8' }
-          ).trim();
-          return Number.parseInt(count, 10);
-        } catch {
-          return null;
-        }
-      }, { timeout: 30000 })
-      .toBe(0);
-
-    // Merge to base in the main worktree (worktree branches can't always check out base refs in-place).
-    // This avoids UI flakiness around ship actions while still validating the diff panel end-to-end.
-    execSync("git checkout main", { cwd: repo.path });
-    execSync(`git -c commit.gpgsign=false merge --no-edit ${worktreeBranch}`, { cwd: repo.path });
-    execSync("git push", { cwd: repo.path });
-
-    await selectChangesView(page, 'base');
-    await expect(getChangesScope(page).getByTestId('changes-diff-status')).toContainText(
-      'Committed',
-      { timeout: 30000 }
-    );
-    await refreshChangesTab(page);
-
-    // Post-ship UI behavior is implementation-dependent (archive can be promoted into
-    // primary flow or hidden behind menu variants), so continue from a fresh draft.
-    await page.getByTestId('sidebar-new-agent').click();
-    await expect(page).toHaveURL(/\/h\/[^/]+\/agent(?:\?|$)/, { timeout: 30000 });
-
-    await expect(page.getByRole('textbox', { name: 'Message agent...' })).toBeEditable();
-  } finally {
-    await rm(nonGitDir, { recursive: true, force: true });
-    await repo.cleanup();
-  }
-});

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/app";
+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/create-agent.spec.ts

@@ -1,27 +0,0 @@
-import { test, expect } from './fixtures';
-import { createAgentInRepo } from './helpers/app';
-import { createTempGitRepo } from './helpers/workspace';
-
-test('create agent in a temp repo', async ({ page }) => {
-  const repo = await createTempGitRepo();
-  const prompt = "Respond with exactly: Hello";
-
-  try {
-    await createAgentInRepo(page, { directory: repo.path, prompt });
-
-    // Verify user message is shown in the stream
-    await expect(page.getByText(prompt, { exact: true })).toBeVisible();
-
-    // Verify we used the seeded fast model (do not fall back to other defaults).
-    const modelPicker = page.getByRole("button", { name: /select agent model/i }).first();
-    await expect(modelPicker).toBeVisible({ timeout: 30000 });
-    await expect(modelPicker).toContainText(/gpt-5\.1-codex-mini/i);
-
-    // Verify the assistant response is rendered.
-    await expect(page.getByText("Hello", { exact: true }).first()).toBeVisible({
-      timeout: 30000,
-    });
-  } finally {
-    await repo.cleanup();
-  }
-});

packages/app/e2e/daemon-connectivity.spec.ts

@@ -1,19 +0,0 @@
-import { test, expect } from './fixtures';
-import { gotoHome, openSettings } from './helpers/app';
-
-test('daemon is connected in settings', async ({ page }) => {
-  const daemonPort = process.env.E2E_DAEMON_PORT;
-  const serverId = process.env.E2E_SERVER_ID;
-  if (!daemonPort) {
-    throw new Error('E2E_DAEMON_PORT is not set (expected from globalSetup).');
-  }
-  if (!serverId) {
-    throw new Error('E2E_SERVER_ID is not set (expected from globalSetup).');
-  }
-
-  await gotoHome(page);
-  await openSettings(page);
-
-  await expect(page.getByText(`127.0.0.1:${daemonPort}`)).toBeVisible();
-  await expect(page.getByTestId(`daemon-card-${serverId}`).getByText('Online', { exact: true })).toBeVisible();
-});

packages/app/e2e/delete-agent-persists.spec.ts

@@ -1,64 +0,0 @@
-import { test, expect } from './fixtures';
-import { ensureHostSelected, gotoHome, setWorkingDirectory } from './helpers/app';
-import { createTempGitRepo } from './helpers/workspace';
-
-test('deleting an agent persists after reload', async ({ page }) => {
-  const repo = await createTempGitRepo();
-  const nonce = Math.random().toString(36).slice(2, 10);
-  const prompt = `respond-ready-${nonce}`;
-
-  try {
-    await gotoHome(page);
-    await setWorkingDirectory(page, repo.path);
-    await ensureHostSelected(page);
-
-    // Create agent (via message input) so it shows up in the sidebar list.
-    const input = page.getByRole('textbox', { name: 'Message agent...' });
-    await expect(input).toBeEditable();
-    await input.fill(prompt);
-    await input.press('Enter');
-    await expect(page).toHaveURL(/\/h\/[^/]+\/agent\/[^/?#]+(?:\?|$)/, {
-      timeout: 30000,
-    });
-    // Wait for the initial turn to complete so the agent can be archived (web uses a hover action).
-    const stopOrCancel = page.getByRole('button', { name: /Stop agent|Canceling agent/ });
-    await stopOrCancel.first().waitFor({ state: 'visible', timeout: 30000 }).catch(() => undefined);
-    await expect(stopOrCancel).toHaveCount(0, { timeout: 120000 });
-
-    const match =
-      page.url().match(/\/h\/([^/]+)\/agent\/([^/?#]+)/) ??
-      page.url().match(/\/agent\/([^/]+)\/([^/?#]+)/);
-    if (!match) {
-      throw new Error(`Expected /h/:serverId/agent/:agentId URL, got ${page.url()}`);
-    }
-    const serverId = decodeURIComponent(match[1]);
-    const agentId = decodeURIComponent(match[2]);
-
-    // Return home and delete via long-press in the agent list.
-    await gotoHome(page);
-    const rowTestId = `agent-row-${serverId}-${agentId}`;
-    const agentRow = page.getByTestId(rowTestId).first();
-    await expect(agentRow).toBeVisible({ timeout: 30000 });
-
-    // Web UX: hover shows a quick-archive icon. First click enters confirm state; second click archives.
-    await agentRow.hover();
-    const quickArchive = page.getByTestId(`agent-archive-${serverId}-${agentId}`).first();
-    await expect(quickArchive).toBeVisible({ timeout: 10000 });
-    await quickArchive.click({ force: true });
-    const confirmArchive = page
-      .getByTestId(`agent-archive-confirm-${serverId}-${agentId}`)
-      .first();
-    await expect(confirmArchive).toBeVisible({ timeout: 10000 });
-    await confirmArchive.click({ force: true });
-
-    // Ensure deletion finished before reload (avoids races).
-    await expect(page.getByTestId(rowTestId)).toHaveCount(0, { timeout: 30000 });
-
-    // A full reload should not bring the agent back.
-    await page.reload();
-    await expect(page.getByRole('textbox', { name: 'Message agent...' })).toBeVisible();
-    await expect(page.getByTestId(rowTestId)).toHaveCount(0, { timeout: 30000 });
-  } finally {
-    await repo.cleanup();
-  }
-});

packages/app/e2e/dictation-web.spec.ts

@@ -1,342 +0,0 @@
-import { test, expect } from './fixtures';
-import {
-  createAgent,
-  createAgentInRepo,
-  ensureHostSelected,
-  gotoHome,
-  openSettings,
-  setWorkingDirectory,
-} from './helpers/app';
-import { createTempGitRepo } from './helpers/workspace';
-import type { Page } from '@playwright/test';
-import { readFile } from 'node:fs/promises';
-import path from 'node:path';
-import {
-  buildHostWorkspaceAgentRoute,
-  buildHostWorkspaceRoute,
-} from '@/utils/host-routes';
-import {
-  ensureWorkspaceAgentPaneVisible,
-  waitForWorkspaceTabsVisible,
-} from './helpers/workspace-tabs';
-
-function escapeRegex(value: string): string {
-  return value.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
-}
-
-async function addFakeMicrophone(page: Page) {
-  const fixturePath = path.resolve(__dirname, 'fixtures', 'recording.wav');
-  const base64Audio = (await readFile(fixturePath)).toString('base64');
-  const mimeType = 'audio/wav';
-
-  return page.addInitScript(({ base64Audio, mimeType }) => {
-    const mic = {
-      active: 0,
-      getUserMediaCalls: 0,
-      stopCalls: 0,
-      lastRecorder: null as null | { state: string },
-    };
-    (window as any).__mic = mic;
-
-    (window as any).isSecureContext = true;
-
-    const nav = navigator as any;
-    if (!nav.mediaDevices) {
-      nav.mediaDevices = {};
-    }
-    nav.mediaDevices.getUserMedia = async () => {
-      mic.getUserMediaCalls += 1;
-      mic.active += 1;
-      const track = {
-        stop: () => {
-          mic.stopCalls += 1;
-          mic.active = Math.max(0, mic.active - 1);
-        },
-      };
-      return {
-        getTracks: () => [track],
-      };
-    };
-
-    const AudioContextCtor =
-      (window as any).AudioContext ?? (window as any).webkitAudioContext;
-    if (AudioContextCtor?.prototype?.createMediaStreamSource) {
-      const nativeCreateMediaStreamSource =
-        AudioContextCtor.prototype.createMediaStreamSource;
-      AudioContextCtor.prototype.createMediaStreamSource =
-        function patchedCreateMediaStreamSource(stream: unknown) {
-          const isFakeStream =
-            !!stream &&
-            typeof (stream as { getTracks?: unknown }).getTracks === 'function' &&
-            typeof (stream as { id?: unknown }).id === 'undefined';
-          if (isFakeStream) {
-            throw new Error('Force recorder fallback for fake microphone stream');
-          }
-          return nativeCreateMediaStreamSource.call(this, stream);
-        };
-    }
-
-    const blobFromBase64 = (base64: string, mimeType: string): Blob => {
-      const binaryString = atob(base64);
-      const bytes = new Uint8Array(binaryString.length);
-      for (let i = 0; i < binaryString.length; i++) {
-        bytes[i] = binaryString.charCodeAt(i);
-      }
-      return new Blob([bytes], { type: mimeType });
-    };
-
-    class FakeMediaRecorder extends EventTarget {
-      public static isTypeSupported() {
-        return true;
-      }
-
-      public state: 'inactive' | 'recording' = 'inactive';
-      public mimeType: string;
-      public ondataavailable: ((event: { data: Blob }) => void) | null = null;
-      public onerror: ((event: unknown) => void) | null = null;
-
-      constructor(_stream: unknown, options?: MediaRecorderOptions) {
-        super();
-        this.mimeType = options?.mimeType ?? 'audio/webm';
-        mic.lastRecorder = this;
-      }
-
-      public start() {
-        this.state = 'recording';
-      }
-
-      public stop() {
-        if (this.state !== 'recording') {
-          throw new Error('Not recording');
-        }
-        this.state = 'inactive';
-        try {
-          this.ondataavailable?.({
-            data: blobFromBase64(base64Audio, mimeType),
-          });
-        } catch (err) {
-          this.onerror?.(err);
-        }
-        this.dispatchEvent(new Event('stop'));
-      }
-    }
-
-    (window as any).MediaRecorder = FakeMediaRecorder;
-  }, { base64Audio, mimeType });
-}
-
-async function expectComposerReady(page: Page) {
-  await expect(page.getByRole('textbox', { name: 'Message agent...' }).first()).toBeEditable();
-}
-
-async function expectDictationStarted(page: Page) {
-  await expect
-    .poll(async () => page.evaluate(() => (window as any).__mic.active as number))
-    .toBe(1);
-}
-
-async function expectDictationStopped(page: Page) {
-  await expect
-    .poll(async () => page.evaluate(() => (window as any).__mic.active as number))
-    .toBe(0);
-}
-
-test('dictation hotkey works on a workspace agent tab', async ({ page }) => {
-  await addFakeMicrophone(page);
-
-  const repo = await createTempGitRepo();
-  try {
-    const serverId = process.env.E2E_SERVER_ID;
-    if (!serverId) {
-      throw new Error('E2E_SERVER_ID is not set.');
-    }
-
-    await createAgentInRepo(page, {
-      directory: repo.path,
-      prompt: 'Respond with exactly: Hello',
-    });
-    await page.goto(buildHostWorkspaceRoute(serverId, repo.path));
-    await waitForWorkspaceTabsVisible(page);
-    await ensureWorkspaceAgentPaneVisible(page);
-    await expect(page).toHaveURL(/\/workspace\//);
-    await expectComposerReady(page);
-
-    await page.keyboard.press('Control+d');
-    await expectDictationStarted(page);
-
-    const calls = await page.evaluate(() => (window as any).__mic.getUserMediaCalls as number);
-    expect(calls).toBe(1);
-
-    await page.keyboard.press('Escape');
-    await expectDictationStopped(page);
-  } finally {
-    await repo.cleanup();
-  }
-});
-
-test('dictation hotkey works on a workspace draft tab', async ({ page }) => {
-  await addFakeMicrophone(page);
-
-  const repo = await createTempGitRepo();
-  try {
-    const serverId = process.env.E2E_SERVER_ID;
-    if (!serverId) {
-      throw new Error('E2E_SERVER_ID is not set.');
-    }
-
-    await createAgentInRepo(page, {
-      directory: repo.path,
-      prompt: 'Respond with exactly: Hello',
-    });
-    await page.goto(buildHostWorkspaceRoute(serverId, repo.path));
-    await waitForWorkspaceTabsVisible(page);
-    await ensureWorkspaceAgentPaneVisible(page);
-
-    await page.getByTestId('workspace-new-agent-tab').first().click();
-    await expectComposerReady(page);
-
-    await page.keyboard.press('Control+d');
-    await expectDictationStarted(page);
-
-    const calls = await page.evaluate(() => (window as any).__mic.getUserMediaCalls as number);
-    expect(calls).toBe(1);
-
-    await page.keyboard.press('Escape');
-    await expectDictationStopped(page);
-  } finally {
-    await repo.cleanup();
-  }
-});
-
-test('dictation hotkeys do not trigger on background screens', async ({ page }) => {
-  await addFakeMicrophone(page);
-
-  const repo = await createTempGitRepo();
-  try {
-    await gotoHome(page);
-    await ensureHostSelected(page);
-    await setWorkingDirectory(page, repo.path);
-    await createAgent(page, 'Respond with exactly: Hello');
-    await expect(page).toHaveURL(/\/workspace\//);
-    await expectComposerReady(page);
-
-    await openSettings(page);
-    await page.keyboard.press('Control+d');
-    await page.waitForTimeout(200);
-
-    const calls = await page.evaluate(() => (window as any).__mic.getUserMediaCalls as number);
-    const active = await page.evaluate(() => (window as any).__mic.active as number);
-
-    expect(calls).toBe(0);
-    expect(active).toBe(0);
-  } finally {
-    await repo.cleanup();
-  }
-});
-
-test('dictation transcribes fixture via real STT', async ({ page }) => {
-  await addFakeMicrophone(page);
-
-  const repo = await createTempGitRepo();
-  try {
-    await gotoHome(page);
-    await ensureHostSelected(page);
-    await setWorkingDirectory(page, repo.path);
-    await createAgent(page, 'Respond with exactly: Hello');
-    await expect(page).toHaveURL(/\/workspace\//);
-    await expectComposerReady(page);
-
-    await page.keyboard.press('Control+d');
-    await expectDictationStarted(page);
-
-    const initialCopyMessageCount = await page
-      .getByRole('button', { name: 'Copy message' })
-      .count();
-
-    await page.keyboard.press('Control+d');
-    await expectDictationStopped(page);
-
-    await expect
-      .poll(
-        async () => page.getByRole('button', { name: 'Copy message' }).count(),
-        { timeout: 60_000 }
-      )
-      .toBeGreaterThan(initialCopyMessageCount);
-  } finally {
-    await repo.cleanup();
-  }
-});
-
-test('cancel stops mic even if recorder is already inactive', async ({ page }) => {
-  await addFakeMicrophone(page);
-
-  const repo = await createTempGitRepo();
-  try {
-    await gotoHome(page);
-    await ensureHostSelected(page);
-    await setWorkingDirectory(page, repo.path);
-    await createAgent(page, 'Respond with exactly: Hello');
-
-    await expect(page).toHaveURL(/\/workspace\//);
-    await expectComposerReady(page);
-
-    await page.keyboard.press('Control+d');
-    await expectDictationStarted(page);
-
-    await page.evaluate(() => {
-      const mic = (window as any).__mic as { lastRecorder: null | { state: string } };
-      if (mic.lastRecorder) {
-        mic.lastRecorder.state = 'inactive';
-      }
-    });
-
-    await page.keyboard.press('Escape');
-    await expectDictationStopped(page);
-  } finally {
-    await repo.cleanup();
-  }
-});
-
-test('dictation confirm+send does not dispatch after navigating away', async ({ page }) => {
-  await addFakeMicrophone(page);
-
-  const repo = await createTempGitRepo();
-  try {
-    await gotoHome(page);
-    await setWorkingDirectory(page, repo.path);
-    await ensureHostSelected(page);
-    await createAgent(page, 'Respond with exactly: Hello');
-
-    await expect(page).toHaveURL(/\/workspace\//);
-    const match = page.url().match(/\/h\/([^/]+)\/workspace\/[^?]+(?:\?open=agent%3A|\\?open=agent:)([^/?#&]+)/);
-    if (!match) {
-      throw new Error(`Expected workspace agent URL, got ${page.url()}`);
-    }
-    const serverId = decodeURIComponent(match[1]!);
-    const agentId = decodeURIComponent(match[2]!);
-    await expectComposerReady(page);
-    const initialCopyMessageCount = await page.getByRole('button', { name: 'Copy message' }).count();
-
-    await page.keyboard.press('Control+d');
-    await expectDictationStarted(page);
-
-    await page.keyboard.press('Control+d');
-
-    const newAgentButton = page.getByTestId('sidebar-new-agent');
-    await expect(newAgentButton).toBeVisible();
-    await newAgentButton.click();
-    await expect(page).toHaveURL(/\/h\/[^/]+\/new-agent(\?|$)/);
-
-    await page.waitForTimeout(10_000);
-
-    await page.goto(buildHostWorkspaceAgentRoute(serverId, repo.path, agentId));
-    await expect(page).toHaveURL(
-      new RegExp(`/h/${escapeRegex(serverId)}/workspace/[^?]+\\?open=agent(?:%3A|:)${escapeRegex(agentId)}(?:$|&)`)
-    );
-
-    await expect(page.getByRole('button', { name: 'Copy message' })).toHaveCount(initialCopyMessageCount);
-    await expect(page.getByTestId('agent-chat-scroll').getByText(/this is a voice note\./i)).toHaveCount(0);
-  } finally {
-    await repo.cleanup();
-  }
-});

packages/app/e2e/draft-explorer-sidebar.spec.ts

@@ -1,42 +0,0 @@
-import { test, expect } from "./fixtures";
-import { ensureHostSelected, gotoHome, setWorkingDirectory } from "./helpers/app";
-import { createTempGitRepo } from "./helpers/workspace";
-
-test("draft enables explorer after selecting a working directory", async ({ page }) => {
-  const repo = await createTempGitRepo("paseo-e2e-draft-explorer-");
-
-  try {
-    await gotoHome(page);
-    await ensureHostSelected(page);
-
-    const newAgentButton = page.getByTestId("sidebar-new-agent").first();
-    await expect(newAgentButton).toBeVisible({ timeout: 30000 });
-    await newAgentButton.click();
-    await expect(page).toHaveURL(/\/h\/[^/]+\/agent(\?|$)/, { timeout: 30000 });
-
-    await setWorkingDirectory(page, repo.path);
-
-    const toggle = page
-      .getByRole("button", {
-        name: /open explorer|close explorer|toggle explorer/i,
-      })
-      .first();
-    await expect(toggle).toBeVisible({ timeout: 30000 });
-
-    await toggle.click();
-    await expect(
-      page.locator('[data-testid="explorer-header"]:visible').first()
-    ).toBeVisible({ timeout: 30000 });
-
-    const terminalsTab = page
-      .locator('[data-testid="explorer-tab-terminals"]:visible')
-      .first();
-    await expect(terminalsTab).toBeVisible({ timeout: 30000 });
-    await terminalsTab.click();
-    await expect(
-      page.locator('[data-testid="terminal-surface"]:visible').first()
-    ).toBeVisible({ timeout: 30000 });
-  } finally {
-    await repo.cleanup();
-  }
-});

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 = String(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

@@ -1,13 +1,14 @@
-import { test as base, expect, type Page } from '@playwright/test';
+import { test as base, expect, type Page } from "@playwright/test";
+import { buildCreateAgentPreferences, buildSeededHost } from "./helpers/daemon-registry";
 
 // Extend base test to provide dynamic baseURL from global-setup
 const test = base.extend({
-  baseURL: async ({}, use) => {
+  baseURL: async ({}, provide) => {
     const metroPort = process.env.E2E_METRO_PORT;
     if (!metroPort) {
-      throw new Error('E2E_METRO_PORT not set - globalSetup must run first');
+      throw new Error("E2E_METRO_PORT not set - globalSetup must run first");
     }
-    await use(`http://localhost:${metroPort}`);
+    await provide(`http://localhost:${metroPort}`);
   },
 });
 
@@ -18,19 +19,19 @@ test.beforeEach(async ({ page }) => {
   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.'
+      "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') {
+  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.'
+      "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.'
+      "E2E_METRO_PORT is not set. Ensure Playwright `globalSetup` starts Metro and exports E2E_METRO_PORT.",
     );
   }
 
@@ -38,17 +39,17 @@ test.beforeEach(async ({ page }) => {
   // 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.' });
+    await ws.close({ code: 1008, reason: "Blocked connection to localhost:6767 during e2e." });
   });
 
   const entries: string[] = [];
   consoleEntries.set(page, entries);
 
-  page.on('console', (message) => {
+  page.on("console", (message) => {
     entries.push(`[console:${message.type()}] ${message.text()}`);
   });
 
-  page.on('pageerror', (error) => {
+  page.on("pageerror", (error) => {
     entries.push(`[pageerror] ${error.message}`);
   });
 
@@ -56,57 +57,38 @@ test.beforeEach(async ({ page }) => {
   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.');
+    throw new Error("E2E_SERVER_ID is not set - expected from Playwright globalSetup.");
   }
-  const testDaemon = {
+  const testDaemon = buildSeededHost({
     serverId,
-    label: 'localhost',
-    connections: [
-      {
-        id: `direct:127.0.0.1:${daemonPort}`,
-        type: 'direct',
-        endpoint: `127.0.0.1:${daemonPort}`,
-      },
-    ],
-    preferredConnectionId: `direct:127.0.0.1:${daemonPort}`,
-    createdAt: nowIso,
-    updatedAt: nowIso,
-  };
-
-  const createAgentPreferences = {
-    // Ensure create flow never uses a remembered host from the developer's real app.
-    serverId: testDaemon.serverId,
-    // Keep e2e fast/cheap by default.
-    provider: 'codex',
-    providerPreferences: {
-      claude: { model: 'haiku' },
-      codex: { model: 'gpt-5.1-codex-mini', thinkingOptionId: 'low' },
-    },
-  };
+    endpoint: `127.0.0.1:${daemonPort}`,
+    nowIso,
+  });
+  const createAgentPreferences = buildCreateAgentPreferences(testDaemon.serverId);
 
   await page.addInitScript(
-    ({ daemon, preferences, seedNonce }) => {
+    ({ 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 disableOnceKey = "@paseo:e2e-disable-default-seed-once";
       const disableValue = localStorage.getItem(disableOnceKey);
       if (disableValue) {
         localStorage.removeItem(disableOnceKey);
-        if (disableValue === seedNonce) {
+        if (disableValue === nonce) {
           return;
         }
       }
 
-      localStorage.setItem('@paseo:e2e', '1');
-      localStorage.setItem('@paseo:e2e-seed-nonce', seedNonce);
+      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));
+      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 }
+    { daemon: testDaemon, preferences: createAgentPreferences, seedNonce },
   );
 });
 
@@ -120,9 +102,9 @@ test.afterEach(async ({ page }, testInfo) => {
     return;
   }
 
-  await testInfo.attach('browser-console', {
-    body: entries.join('\n'),
-    contentType: 'text/plain',
+  await testInfo.attach("browser-console", {
+    body: entries.join("\n"),
+    contentType: "text/plain",
   });
 });
 

packages/app/e2e/git-diff-sticky-headers.spec.ts

@@ -1,119 +0,0 @@
-import path from 'node:path';
-import { appendFile } from 'node:fs/promises';
-import { test, expect, type Page } from './fixtures';
-import { createAgent, ensureHostSelected, gotoHome, setWorkingDirectory } from './helpers/app';
-import { createTempGitRepo } from './helpers/workspace';
-
-test.describe.configure({ timeout: 90000 });
-
-function getChangesScope(page: Page) {
-  return page.locator('[data-testid="explorer-content-area"]:visible').first();
-}
-
-async function ensureExplorerTabsVisible(page: Page) {
-  const changesTab = page.getByTestId('explorer-tab-changes').first();
-  if (await changesTab.isVisible().catch(() => false)) {
-    return;
-  }
-
-  const toggle = page
-    .getByRole('button', { name: /open explorer|close explorer|toggle explorer/i })
-    .first();
-  await expect(toggle).toBeVisible({ timeout: 10000 });
-  for (let attempt = 0; attempt < 4; attempt += 1) {
-    if (await changesTab.isVisible().catch(() => false)) {
-      return;
-    }
-    await toggle.click();
-    await page.waitForTimeout(200);
-  }
-  await expect(changesTab).toBeVisible({ timeout: 30000 });
-}
-
-async function openChangesPanel(page: Page) {
-  await ensureExplorerTabsVisible(page);
-  const changesHeader = getChangesScope(page).getByTestId('changes-header');
-  if (!(await changesHeader.isVisible())) {
-    await page.getByTestId('explorer-tab-changes').first().click();
-  }
-  await expect(changesHeader).toBeVisible();
-}
-
-async function refreshUncommittedMode(page: Page) {
-  const scope = getChangesScope(page);
-  const toggle = scope.getByTestId('changes-diff-status').first();
-  await expect(toggle).toBeVisible({ timeout: 30000 });
-
-  const diffModeBackdrop = page.getByTestId('changes-diff-status-menu-backdrop');
-  if (await diffModeBackdrop.isVisible().catch(() => false)) {
-    await diffModeBackdrop.click({ force: true });
-    await expect(diffModeBackdrop).toHaveCount(0);
-  }
-
-  const currentLabel = (await toggle.innerText()).trim();
-  await toggle.click({ force: true });
-  await expect(page.getByTestId('changes-diff-status-menu')).toBeVisible({ timeout: 10000 });
-  const firstTarget = currentLabel === 'Uncommitted' ? 'changes-diff-mode-committed' : 'changes-diff-mode-uncommitted';
-  await page.getByTestId(firstTarget).click({ force: true });
-  await expect.poll(async () => (await toggle.innerText()).trim()).not.toBe(currentLabel);
-
-  const nextLabel = (await toggle.innerText()).trim();
-  await toggle.click({ force: true });
-  await expect(page.getByTestId('changes-diff-status-menu')).toBeVisible({ timeout: 10000 });
-  const secondTarget = nextLabel === 'Uncommitted' ? 'changes-diff-mode-committed' : 'changes-diff-mode-uncommitted';
-  await page.getByTestId(secondTarget).click({ force: true });
-  await expect.poll(async () => (await toggle.innerText()).trim()).not.toBe(nextLabel);
-}
-
-async function createAgentAndWait(page: Page, message: string) {
-  await createAgent(page, message);
-}
-
-test('keeps file header sticky while scrolling within a long diff', async ({ page }) => {
-  const repo = await createTempGitRepo('paseo-e2e-sticky-');
-
-  try {
-    await gotoHome(page);
-    await setWorkingDirectory(page, repo.path);
-    await ensureHostSelected(page);
-    await createAgentAndWait(page, 'Respond with exactly: READY');
-
-    await openChangesPanel(page);
-
-    const readmePath = path.join(repo.path, 'README.md');
-    const lines = Array.from({ length: 400 }, (_, idx) => `Sticky header line ${idx}\n`).join('');
-    await appendFile(readmePath, `\n${lines}`);
-
-    await refreshUncommittedMode(page);
-
-    const scope = getChangesScope(page);
-    await expect(scope.getByText('README.md', { exact: true })).toBeVisible({ timeout: 30000 });
-
-    const fileToggle = scope.getByTestId('diff-file-0-toggle').first();
-    await fileToggle.click();
-
-    const markerLine = scope.getByText('Sticky header line 250').first();
-    await expect(markerLine).toBeVisible({ timeout: 30000 });
-
-    const scroll = scope.getByTestId('git-diff-scroll').first();
-    await expect(scroll).toBeVisible();
-
-    await expect.poll(async () => {
-      return await scroll.evaluate((el) => (el.scrollHeight ?? 0) > (el.clientHeight ?? 0));
-    }).toBe(true);
-
-    await scroll.hover();
-    for (let i = 0; i < 12; i++) {
-      await page.mouse.wheel(0, 700);
-    }
-
-    await expect.poll(async () => {
-      return await scroll.evaluate((el) => el.scrollTop ?? 0);
-    }).toBeGreaterThan(0);
-
-    await expect(scope.getByText('Sticky header line 390').first()).toBeVisible({ timeout: 30000 });
-    await expect(fileToggle).toBeVisible();
-  } finally {
-    await repo.cleanup();
-  }
-});

packages/app/e2e/global-draft-create-status-controls.spec.ts

@@ -1,47 +0,0 @@
-import { test, expect } from './fixtures'
-import { createTempGitRepo } from './helpers/workspace'
-import { ensureHostSelected, gotoHome, setWorkingDirectory } from './helpers/app'
-
-function escapeRegex(value: string): string {
-  return value.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')
-}
-
-test('global draft create uses input status controls and preserves optimistic flow', async ({ page }) => {
-  const repo = await createTempGitRepo('paseo-e2e-global-draft-')
-  const prompt = `global draft prompt ${Date.now()}`
-
-  try {
-    await gotoHome(page)
-    await ensureHostSelected(page)
-    await setWorkingDirectory(page, repo.path)
-
-    await expect(page.getByTestId('working-directory-select').first()).toBeVisible({ timeout: 30_000 })
-
-    const providerSelector = page.getByTestId('agent-provider-selector').first()
-    const modeSelector = page.getByTestId('agent-mode-selector').first()
-    const modelSelector = page.getByTestId('agent-model-selector').first()
-    const thinkingSelector = page.getByTestId('agent-thinking-selector').first()
-
-    await expect(providerSelector).toBeVisible({ timeout: 30_000 })
-    await expect(modeSelector).toBeVisible({ timeout: 30_000 })
-    await expect(modelSelector).toBeVisible({ timeout: 30_000 })
-    await expect(thinkingSelector).toBeVisible({ timeout: 30_000 })
-
-    const selectedMode = ((await modeSelector.innerText()) ?? '').trim()
-    const selectedModel = ((await modelSelector.innerText()) ?? '').trim()
-    const selectedThinking = ((await thinkingSelector.innerText()) ?? '').trim()
-
-    const composer = page.getByRole('textbox', { name: 'Message agent...' }).first()
-    await composer.fill(prompt)
-    await composer.press('Enter')
-
-    await expect(page.getByText(prompt, { exact: true }).first()).toBeVisible({ timeout: 5_000 })
-    await expect(page).toHaveURL(/\/workspace\//, { timeout: 30_000 })
-
-    await expect(modelSelector).toContainText(new RegExp(escapeRegex(selectedModel), 'i'))
-    await expect(modeSelector).toContainText(new RegExp(escapeRegex(selectedMode), 'i'))
-    await expect(thinkingSelector).toContainText(new RegExp(escapeRegex(selectedThinking), 'i'))
-  } finally {
-    await repo.cleanup()
-  }
-})

packages/app/e2e/global-draft-to-existing-workspace-redirect.spec.ts

@@ -1,219 +0,0 @@
-import { test, expect } from './fixtures'
-import type { Page } from '@playwright/test'
-import { ensureHostSelected, gotoHome, setWorkingDirectory } from './helpers/app'
-import { createTempGitRepo } from './helpers/workspace'
-import { getWorkspaceTabTestIds } from './helpers/workspace-tabs'
-
-function parseWorkspaceContextFromUrl(rawUrl: string): { workspaceToken: string | null; agentId: string | null } {
-  const url = new URL(rawUrl)
-  const pathMatch = url.pathname.match(/\/workspace\/([^/?#]+)(?:\/agent\/([^/?#]+))?/) ?? null
-  const workspaceToken = pathMatch?.[1] ?? null
-  let agentId = pathMatch?.[2] ?? null
-
-  if (!agentId) {
-    const open = url.searchParams.get('open')
-    const openMatch = open?.match(/^agent:(.+)$/) ?? null
-    if (openMatch?.[1]) {
-      agentId = openMatch[1]
-    }
-  }
-
-  return { workspaceToken, agentId }
-}
-
-function getOpenAgentIdFromUrl(rawUrl: string): string | null {
-  const url = new URL(rawUrl)
-  const open = url.searchParams.get('open')
-  const openMatch = open?.match(/^agent:(.+)$/) ?? null
-  return openMatch?.[1] ?? null
-}
-
-async function preferSlowerThinkingOption(page: Page): Promise<void> {
-  await page.keyboard.press('Escape').catch(() => undefined)
-  const thinkingTrigger = page.getByTestId('agent-thinking-selector').first()
-  if (!(await thinkingTrigger.isVisible().catch(() => false))) {
-    return
-  }
-  await thinkingTrigger.click({ force: true })
-  const menu = page.getByTestId('agent-thinking-menu').first()
-  if (!(await menu.isVisible().catch(() => false))) {
-    return
-  }
-
-  const preferred = ['high', 'max', 'deep', 'long', 'medium']
-  for (const label of preferred) {
-    const option = menu.getByRole('button', { name: new RegExp(`^${label}$`, 'i') }).first()
-    if (await option.isVisible().catch(() => false)) {
-      await option.click({ force: true })
-      await expect(menu).not.toBeVisible({ timeout: 5_000 })
-      return
-    }
-  }
-
-  const options = menu.getByRole('button')
-  const count = await options.count()
-  if (count > 0) {
-    await options.last().click({ force: true })
-  }
-  await expect(menu).not.toBeVisible({ timeout: 5_000 })
-}
-
-async function isAnyStopVisible(page: Page): Promise<boolean> {
-  const candidates = page.getByRole('button', { name: /stop|cancel/i })
-  const count = await candidates.count()
-  for (let index = 0; index < count; index += 1) {
-    if (await candidates.nth(index).isVisible().catch(() => false)) {
-      return true
-    }
-  }
-  return false
-}
-
-async function isRunningControlVisible(page: Page): Promise<boolean> {
-  if (await isAnyStopVisible(page)) {
-    return true
-  }
-  const interruptSendButton = page.getByRole('button', { name: /interrupt agent|send and interrupt/i })
-  const count = await interruptSendButton.count()
-  for (let index = 0; index < count; index += 1) {
-    if (await interruptSendButton.nth(index).isVisible().catch(() => false)) {
-      return true
-    }
-  }
-  return false
-}
-
-async function hasVisibleText(page: Page, matcher: RegExp | string): Promise<boolean> {
-  const locator = page.getByText(matcher).first()
-  const matches = page.getByText(matcher)
-  const count = await matches.count()
-  for (let index = 0; index < count; index += 1) {
-    if (await matches.nth(index).isVisible().catch(() => false)) {
-      return true
-    }
-  }
-  return await locator.isVisible().catch(() => false)
-}
-
-test('global draft create in existing workspace redirects to that workspace with created agent tab and settles to idle', async ({
-  page,
-}) => {
-  test.setTimeout(360_000)
-  const serverId = process.env.E2E_SERVER_ID
-  if (!serverId) {
-    throw new Error('E2E_SERVER_ID is not set.')
-  }
-
-  const repo = await createTempGitRepo('paseo-e2e-global-existing-workspace-')
-  const id = `${Date.now()}`
-  const seedPrompt = `hello seed-${id}. please reply exactly: ACK_SEED_${id}`
-  const secondPrompt = `hello second-${id}. please reply exactly: ACK_SECOND_${id}`
-  const lifecyclePrompt = `hello lifecycle-${id}. wait 8 seconds, then reply exactly: ACK_LIFECYCLE_${id}`
-  const seedToken = `ACK_SEED_${id}`
-  const secondToken = `ACK_SECOND_${id}`
-  expect(seedPrompt).not.toBe(secondPrompt)
-  expect(seedToken).not.toBe(secondToken)
-
-  try {
-    await gotoHome(page)
-    await ensureHostSelected(page)
-    await page.goto(`/h/${serverId}/new-agent`)
-    await expect(page.locator('[data-testid="working-directory-select"]:visible').first()).toBeVisible({
-      timeout: 30_000,
-    })
-
-    // 1) Seed workspace with an existing agent.
-    await setWorkingDirectory(page, repo.path)
-    const seedComposer = page.getByRole('textbox', { name: 'Message agent...' }).first()
-    await seedComposer.fill(seedPrompt)
-    await seedComposer.press('Enter')
-
-    await expect(page.getByText(seedPrompt, { exact: true }).first()).toBeVisible({ timeout: 30_000 })
-    await expect(page).toHaveURL(/\/workspace\//, { timeout: 60_000 })
-    const seededContext = parseWorkspaceContextFromUrl(page.url())
-    if (!seededContext.workspaceToken || !seededContext.agentId) {
-      throw new Error(`Expected seeded workspace token and agent id in URL, got: ${page.url()}`)
-    }
-    const seededAgentId = seededContext.agentId
-    expect(getOpenAgentIdFromUrl(page.url())).toBe(seededAgentId)
-    await expect(page.getByText(new RegExp(seedToken)).first()).toBeVisible({ timeout: 180_000 })
-
-    // 2) Use global New Agent entry and 3) select same workspace.
-    await page.getByText('New agent', { exact: true }).first().click()
-    await expect(page).toHaveURL(new RegExp(`/h/${serverId}/new-agent`), { timeout: 30_000 })
-    await expect(page.locator('[data-testid="working-directory-select"]:visible').first()).toBeVisible({
-      timeout: 30_000,
-    })
-
-    await setWorkingDirectory(page, repo.path)
-    await preferSlowerThinkingOption(page)
-    await page.keyboard.press('Escape').catch(() => undefined)
-
-    // 4) Create second agent from global draft.
-    const composer = page.getByRole('textbox', { name: 'Message agent...' }).first()
-    await composer.fill(secondPrompt)
-    await composer.press('Enter')
-    await page.waitForTimeout(500)
-    if (page.url().includes('/new-agent')) {
-      const sendButton = page.getByRole('button', { name: /send message/i }).first()
-      await expect(sendButton).toBeVisible({ timeout: 10_000 })
-      await expect(sendButton).toBeEnabled({ timeout: 10_000 })
-      await sendButton.click({ force: true })
-    }
-
-    // 5) Assert redirect workspace context + created agent tab is active/open.
-    await expect(page).toHaveURL(/\/workspace\//, { timeout: 60_000 })
-    const createdContext = parseWorkspaceContextFromUrl(page.url())
-    expect(createdContext.workspaceToken).toBe(seededContext.workspaceToken)
-    if (!createdContext.agentId) {
-      throw new Error(`Expected created agent id in URL, got: ${page.url()}`)
-    }
-    const createdAgentId = createdContext.agentId
-    expect(createdAgentId).toBeTruthy()
-    expect(createdAgentId).not.toBe(seededAgentId)
-    expect(getOpenAgentIdFromUrl(page.url())).toBe(createdAgentId)
-    expect(getOpenAgentIdFromUrl(page.url())).not.toBe(seededAgentId)
-
-    const createdTabTestId = `workspace-tab-agent_${createdAgentId}`
-    await expect
-      .poll(async () => {
-        const finalTabIds = await getWorkspaceTabTestIds(page)
-        return finalTabIds.includes(createdTabTestId)
-      })
-      .toBe(true)
-    expect(getOpenAgentIdFromUrl(page.url())).toBe(createdAgentId)
-
-    // 6) Response assertions are scoped to created-agent active context.
-    await expect
-      .poll(async () => hasVisibleText(page, new RegExp(secondToken)), { timeout: 240_000 })
-      .toBe(true)
-    await expect(page.getByText(seedToken, { exact: true }).first()).not.toBeVisible({ timeout: 30_000 })
-
-    // 7) Lifecycle assertions on created agent only: running first, then idle.
-    const lifecycleComposer = page.getByRole('textbox', { name: 'Message agent...' }).first()
-    await lifecycleComposer.fill(lifecyclePrompt)
-    await lifecycleComposer.press('Enter')
-    await expect
-      .poll(async () => hasVisibleText(page, lifecyclePrompt), { timeout: 30_000 })
-      .toBe(true)
-
-    let sawRunning = false
-    for (let attempt = 0; attempt < 1200; attempt += 1) {
-      if (await isRunningControlVisible(page)) {
-        sawRunning = true
-        break
-      }
-      await page.waitForTimeout(50)
-    }
-    expect(sawRunning).toBe(true)
-
-    await expect
-      .poll(async () => (await isRunningControlVisible(page)) === false, { timeout: 240_000 })
-      .toBe(true)
-    const idleComposer = page.getByRole('textbox', { name: 'Message agent...' }).first()
-    await expect(idleComposer).toBeVisible({ timeout: 30_000 })
-    await expect(idleComposer).toBeEditable({ timeout: 30_000 })
-  } finally {
-    await repo.cleanup()
-  }
-})

packages/app/e2e/global-setup.ts

@@ -1,28 +1,29 @@
-import { spawn, type ChildProcess, execSync } from 'node:child_process';
-import { existsSync } from 'node:fs';
-import { mkdtemp, rm } 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 { spawn, type ChildProcess, execFileSync, execSync } from "node:child_process";
+import { existsSync } from "node:fs";
+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) => {
     const server = net.createServer();
-    server.once('error', reject);
+    server.once("error", reject);
     server.listen(0, () => {
       const address = server.address();
-      if (!address || typeof address === 'string') {
-        server.close(() => reject(new Error('Failed to acquire port')));
+      if (!address || typeof address === "string") {
+        server.close(() => reject(new Error("Failed to acquire port")));
         return;
       }
       server.close(() => resolve(address.port));
@@ -40,18 +41,18 @@ function createLineBuffer(maxLines = 120): { add: (line: string) => void; dump:
       }
     },
     dump() {
-      return lines.join('\n');
+      return lines.join("\n");
     },
   };
 }
 
 function formatRecentOutput(getRecentOutput?: () => string): string {
   if (!getRecentOutput) {
-    return '';
+    return "";
   }
   const output = getRecentOutput().trim();
   if (!output) {
-    return '';
+    return "";
   }
   return `\nRecent output:\n${output}`;
 }
@@ -61,21 +62,15 @@ function sleep(ms: number): Promise<void> {
 }
 
 async function waitForServer(port: number, options: WaitForServerOptions): Promise<void> {
-  const {
-    host = '127.0.0.1',
-    timeoutMs = 15000,
-    label,
-    childProcess,
-    getRecentOutput,
-  } = options;
+  const { host = "127.0.0.1", timeoutMs = 15000, label, childProcess, getRecentOutput } = options;
   const start = Date.now();
   let lastConnectionError: unknown = null;
 
   while (Date.now() - start < timeoutMs) {
     if (childProcess && childProcess.exitCode !== null) {
-      const signal = childProcess.signalCode ? `, signal ${childProcess.signalCode}` : '';
+      const signal = childProcess.signalCode ? `, signal ${childProcess.signalCode}` : "";
       throw new Error(
-        `${label} exited before listening on ${host}:${port} (exit code ${childProcess.exitCode}${signal}).${formatRecentOutput(getRecentOutput)}`
+        `${label} exited before listening on ${host}:${port} (exit code ${childProcess.exitCode}${signal}).${formatRecentOutput(getRecentOutput)}`,
       );
     }
 
@@ -89,7 +84,7 @@ async function waitForServer(port: number, options: WaitForServerOptions): Promi
           socket.destroy();
           reject(new Error(`Connection timed out to ${host}:${port}`));
         });
-        socket.on('error', reject);
+        socket.on("error", reject);
       });
       return;
     } catch (error) {
@@ -99,9 +94,11 @@ async function waitForServer(port: number, options: WaitForServerOptions): Promi
   }
 
   const reason =
-    lastConnectionError instanceof Error ? ` Last connection error: ${lastConnectionError.message}` : '';
+    lastConnectionError instanceof Error
+      ? ` Last connection error: ${lastConnectionError.message}`
+      : "";
   throw new Error(
-    `${label} did not start on ${host}:${port} within ${timeoutMs}ms.${reason}${formatRecentOutput(getRecentOutput)}`
+    `${label} did not start on ${host}:${port} within ${timeoutMs}ms.${reason}${formatRecentOutput(getRecentOutput)}`,
   );
 }
 
@@ -126,25 +123,30 @@ async function stopProcess(child: ChildProcess | null): Promise<void> {
   if (child.exitCode !== null || child.signalCode !== null) {
     return;
   }
-  child.kill('SIGTERM');
+  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');
+        child.kill("SIGKILL");
       }
-      resolve();
+      settle();
     }, 5000);
-    child.once('exit', () => {
-      clearTimeout(timeout);
-      resolve();
-    });
+    child.once("exit", settle);
   });
 }
 
 function summarizeOpenAiErrorBody(body: string): string {
   const trimmed = body.trim();
   if (!trimmed) {
-    return 'empty response body';
+    return "empty response body";
   }
   if (trimmed.length <= 240) {
     return trimmed;
@@ -159,8 +161,8 @@ async function isOpenAiApiKeyUsable(apiKey: string | undefined): Promise<boolean
   }
 
   try {
-    const response = await fetch('https://api.openai.com/v1/models?limit=1', {
-      method: 'GET',
+    const response = await fetch("https://api.openai.com/v1/models?limit=1", {
+      method: "GET",
       headers: {
         Authorization: `Bearer ${key}`,
       },
@@ -170,14 +172,14 @@ async function isOpenAiApiKeyUsable(apiKey: string | undefined): Promise<boolean
     }
     const body = await response.text();
     console.warn(
-      `[e2e] OPENAI_API_KEY probe failed (${response.status}): ${summarizeOpenAiErrorBody(body)}`
+      `[e2e] OPENAI_API_KEY probe failed (${response.status}): ${summarizeOpenAiErrorBody(body)}`,
     );
     return false;
   } catch (error) {
     console.warn(
       `[e2e] OPENAI_API_KEY probe request failed: ${
         error instanceof Error ? error.message : String(error)
-      }`
+      }`,
     );
     return false;
   }
@@ -186,327 +188,530 @@ 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 args = process.argv.slice(2);
+
+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]) {
+  console.error("no pull requests found for branch");
+  process.exit(1);
+}
+
+if (args[0] === "issue" && args[1] === "list") {
+  console.log("[]");
+  process.exit(0);
+}
+
+console.error("Unsupported fake gh invocation: " + args.join(" "));
+process.exit(1);
+`,
+  );
+  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 {
+  const relayDistEntry = path.join(repoRoot, "packages/relay/dist/e2ee.js");
+  if (existsSync(relayDistEntry)) {
+    return;
+  }
+
+  console.log("[e2e] Building @getpaseo/relay for daemon startup");
+  execSync("npm run build --workspace=@getpaseo/relay", {
+    cwd: repoRoot,
+    stdio: "inherit",
+  });
 }
 
 function decodeOfferFromFragmentUrl(url: string): OfferPayload {
-  const marker = '#offer=';
+  const marker = "#offer=";
   const idx = url.indexOf(marker);
   if (idx === -1) {
     throw new Error(`missing ${marker} fragment: ${url}`);
   }
   const encoded = url.slice(idx + marker.length);
-  const json = Buffer.from(encoded, 'base64url').toString('utf8');
+  const json = Buffer.from(encoded, "base64url").toString("utf8");
   const offer = JSON.parse(json) as Partial<OfferPayload>;
-  if (offer.v !== 2) throw new Error('offer.v missing/invalid');
-  if (!offer.serverId) throw new Error('offer.serverId missing');
-  if (!offer.daemonPublicKeyB64) throw new Error('offer.daemonPublicKeyB64 missing');
-  if (!offer.relay?.endpoint) throw new Error('offer.relay.endpoint missing');
+  if (offer.v !== 2) throw new Error("offer.v missing/invalid");
+  if (!offer.serverId) throw new Error("offer.serverId missing");
+  if (!offer.daemonPublicKeyB64) throw new Error("offer.daemonPublicKeyB64 missing");
+  if (!offer.relay?.endpoint) throw new Error("offer.relay.endpoint missing");
   return offer as OfferPayload;
 }
 
-export default async function globalSetup() {
-  const repoRoot = path.resolve(__dirname, '../../..');
-  const envTestPath = path.join(repoRoot, '.env.test');
+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 ?? '', '.paseo', 'models', 'local-speech');
-  const hasDefaultLocalModelsDir = defaultLocalModelsDir.trim().length > 0 && existsSync(defaultLocalModelsDir);
-  const dictationProvider = openAiUsable ? 'openai' : 'local';
+  const defaultLocalModelsDir = path.join(
+    process.env.HOME ?? "",
+    ".paseo",
+    "models",
+    "local-speech",
+  );
+  const hasDefaultLocalModelsDir =
+    defaultLocalModelsDir.trim().length > 0 && existsSync(defaultLocalModelsDir);
+  const dictationProvider = openAiUsable ? "openai" : "local";
 
-  if (dictationProvider === 'local' && !hasDefaultLocalModelsDir) {
+  if (dictationProvider === "local" && !hasDefaultLocalModelsDir) {
     throw new Error(
-      'OpenAI key is not usable and local speech models are unavailable at ~/.paseo/models/local-speech. ' +
-        'Either provide a valid OPENAI_API_KEY or install local speech models before running app e2e tests.'
+      "OpenAI key is not usable and local speech models are unavailable at ~/.paseo/models/local-speech. " +
+        "Either provide a valid OPENAI_API_KEY or install local speech models before running app e2e tests.",
     );
   }
 
-  const localModelsDir = dictationProvider === 'local' ? defaultLocalModelsDir : null;
+  const localModelsDir = dictationProvider === "local" ? defaultLocalModelsDir : null;
   console.log(
-    `[e2e] Dictation STT provider: ${dictationProvider}${openAiUsable ? '' : ' (OpenAI probe failed)'}`
+    `[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}`,
-        // Use OpenAI speech providers in e2e to avoid local model bootstrapping delays.
-        PASEO_DICTATION_ENABLED: '1',
-        PASEO_VOICE_MODE_ENABLED: '1',
-        PASEO_DICTATION_STT_PROVIDER: dictationProvider,
-        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',
+        label: "Paseo daemon",
         childProcess: daemonProcess,
         getRecentOutput: daemonLineBuffer.dump,
       }),
       waitForServer(metroPort, {
-        label: 'Metro web server',
-        timeoutMs: 120000, // Metro can take longer to start
+        label: "Metro web server",
+        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);
-    console.log(`[e2e] Test daemon started on port ${port}, Metro on port ${metroPort}, home: ${paseoHome}`);
+    console.log(
+      `[e2e] Test daemon started on port ${port}, Metro on port ${metroPort}, home: ${paseoHome}`,
+    );
 
     return async () => {
       await cleanup();
-      console.log('[e2e] Test daemon stopped');
+      console.log("[e2e] Test daemon stopped");
     };
   } catch (error) {
     await cleanup();

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

@@ -2,29 +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 {
-  buildHostWorkspaceAgentRoute,
-  buildHostWorkspaceRoute,
-} from "../../src/utils/host-routes";
+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: {
@@ -37,11 +35,8 @@ export type DaemonClientInstance = {
     initialPrompt: string;
   }): Promise<{ id: string }>;
   sendAgentMessage(agentId: string, text: string): Promise<void>;
-  waitForFinish(
-    agentId: string,
-    timeout?: number
-  ): Promise<{ status: string }>;
-};
+  waitForFinish(agentId: string, timeout?: number): Promise<{ status: string }>;
+}
 
 function getDaemonWsUrl(): string {
   const daemonPort = process.env.E2E_DAEMON_PORT;
@@ -66,17 +61,17 @@ function buildReplyBlock(label: string, lineCount = 14): string {
   }).join("\n");
 }
 
-function buildProtocolMessage(label: string): string {
+function buildProtocolMessage(label: string, lineCount = 14): string {
   return [
     "For every message in this chat, reply with exactly the text after the final line `REPLY:`.",
     "Do not add extra words, bullets, markdown fences, or tool calls.",
     "REPLY:",
-    buildReplyBlock(label),
+    buildReplyBlock(label, lineCount),
   ].join("\n");
 }
 
-function buildReplyMessage(label: string): string {
-  return ["REPLY:", buildReplyBlock(label)].join("\n");
+function buildReplyMessage(label: string, lineCount = 14): string {
+  return ["REPLY:", buildReplyBlock(label, lineCount)].join("\n");
 }
 
 export function createReplyTurn(label: string): {
@@ -89,31 +84,34 @@ export function createReplyTurn(label: string): {
   };
 }
 
-async function loadDaemonClientConstructor(): Promise<new (config: {
+interface DaemonClientConfig {
   url: string;
   clientId: string;
   clientType: "cli";
-}) => DaemonClientInstance> {
-  const repoRoot = path.resolve(process.cwd(), "../..");
+  webSocketFactory?: NodeWebSocketFactory;
+}
+
+async function loadDaemonClientConstructor(): Promise<
+  new (config: DaemonClientConfig) => DaemonClientInstance
+> {
+  const repoRoot = path.resolve(__dirname, "../../../../");
   const moduleUrl = pathToFileURL(
-    path.join(repoRoot, "packages/server/dist/server/server/exports.js")
+    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-${randomUUID()}`,
     clientType: "cli",
+    webSocketFactory,
   });
   await client.connect();
   return client;
@@ -124,34 +122,36 @@ export async function seedBottomAnchorAgent(input: {
   cwd: string;
   title?: string;
   turnCount?: number;
+  lineCount?: number;
 }): Promise<SeededAgent> {
   const title = input.title ?? `bottom-anchor-${Date.now()}`;
   const turnCount = Math.max(3, input.turnCount ?? 5);
+  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,
     title,
-    initialPrompt: buildProtocolMessage(`${title}-turn-00`),
+    initialPrompt: buildProtocolMessage(`${title}-turn-00`, lineCount),
   });
   const initialFinish = await input.client.waitForFinish(created.id, 120000);
   if (initialFinish.status !== "idle") {
     throw new Error(
-      `Expected seeded agent ${created.id} to become idle after initial prompt, got ${initialFinish.status}.`
+      `Expected seeded agent ${created.id} to become idle after initial prompt, got ${initialFinish.status}.`,
     );
   }
 
-  let expectedTailText = buildReplyBlock(`${title}-turn-00`);
+  let expectedTailText = buildReplyBlock(`${title}-turn-00`, lineCount);
   for (let index = 1; index < turnCount; index += 1) {
     const label = `${title}-turn-${index.toString().padStart(2, "0")}`;
-    expectedTailText = buildReplyBlock(label);
-    await input.client.sendAgentMessage(created.id, buildReplyMessage(label));
+    expectedTailText = buildReplyBlock(label, lineCount);
+    await input.client.sendAgentMessage(created.id, buildReplyMessage(label, lineCount));
     const finish = await input.client.waitForFinish(created.id, 120000);
     if (finish.status !== "idle") {
       throw new Error(
-        `Expected seeded agent ${created.id} to become idle after turn ${index}, got ${finish.status}.`
+        `Expected seeded agent ${created.id} to become idle after turn ${index}, got ${finish.status}.`,
       );
     }
   }
@@ -160,29 +160,37 @@ export async function seedBottomAnchorAgent(input: {
     id: created.id,
     title,
     expectedTailText,
-    url: buildHostWorkspaceAgentRoute(getServerId(), input.cwd, created.id),
+    url: `${buildHostWorkspaceRoute(getServerId(), input.cwd)}?open=${encodeURIComponent(`agent:${created.id}`)}`,
     workspaceUrl: buildHostWorkspaceRoute(getServerId(), input.cwd),
   };
 }
 
+function getVisibleChatScroll(page: Page) {
+  return page.locator('[data-testid="agent-chat-scroll"]:visible').first();
+}
+
 export async function readScrollMetrics(page: Page): Promise<ScrollMetrics> {
-  return page.getByTestId("agent-chat-scroll").evaluate((root: Element) => {
-    const rootElement = root as HTMLElement;
-    const candidates = [rootElement, ...Array.from(rootElement.querySelectorAll("*"))];
+  return getVisibleChatScroll(page).evaluate((root: Element) => {
+    const candidates = [root, ...Array.from(root.querySelectorAll("*"))]
+      .filter((element): element is HTMLElement => element instanceof HTMLElement)
+      .filter((element) => {
+        const tagName = element.tagName.toLowerCase();
+        const isEditable =
+          tagName === "textarea" ||
+          tagName === "input" ||
+          element.getAttribute("contenteditable") === "true";
+        return !isEditable && element.scrollHeight - element.clientHeight > 1;
+      });
     const scrollElement =
-      candidates.find(
-        (element) =>
-          element instanceof HTMLElement &&
-          element.scrollHeight - element.clientHeight > 1
-      ) ?? rootElement;
+      candidates.sort(
+        (left, right) =>
+          right.scrollHeight - right.clientHeight - (left.scrollHeight - left.clientHeight),
+      )[0] ?? (root as HTMLElement);
 
     const offsetY = Math.max(0, scrollElement.scrollTop);
     const contentHeight = Math.max(0, scrollElement.scrollHeight);
     const viewportHeight = Math.max(0, scrollElement.clientHeight);
-    const distanceFromBottom = Math.max(
-      0,
-      contentHeight - (offsetY + viewportHeight)
-    );
+    const distanceFromBottom = Math.max(0, contentHeight - (offsetY + viewportHeight));
 
     return {
       offsetY,
@@ -194,29 +202,33 @@ export async function readScrollMetrics(page: Page): Promise<ScrollMetrics> {
 }
 
 export async function scrollUpFromBottom(page: Page, pixels: number): Promise<void> {
-  await page.getByTestId("agent-chat-scroll").evaluate(
-    (root: Element, amount: number) => {
-      const rootElement = root as HTMLElement;
-      const candidates = [rootElement, ...Array.from(rootElement.querySelectorAll("*"))];
-      const scrollElement =
-        candidates.find(
-          (element) =>
-            element instanceof HTMLElement &&
-            element.scrollHeight - element.clientHeight > 1
-        ) ?? rootElement;
-
-      const bottomOffset = Math.max(
-        0,
-        scrollElement.scrollHeight - scrollElement.clientHeight
+  const scrollViewport = getVisibleChatScroll(page);
+  await expect(scrollViewport).toHaveCount(1, { timeout: 30000 });
+  let remaining = Math.max(0, pixels);
+  while (remaining > 0) {
+    const delta = Math.min(240, remaining);
+    await scrollViewport.evaluate((element: Element, step: number) => {
+      const scrollContainer = element as HTMLElement;
+      scrollContainer.dispatchEvent(
+        new WheelEvent("wheel", {
+          deltaY: -step,
+          bubbles: true,
+          cancelable: true,
+        }),
       );
-      scrollElement.scrollTop = Math.max(0, bottomOffset - amount);
-    },
-    pixels
-  );
+      scrollContainer.scrollTop = Math.max(0, scrollContainer.scrollTop - step);
+      scrollContainer.dispatchEvent(new Event("scroll", { bubbles: true }));
+    }, delta);
+    remaining -= delta;
+
+    if ((await readScrollMetrics(page)).distanceFromBottom > NEAR_BOTTOM_THRESHOLD_PX) {
+      return;
+    }
+  }
 }
 
 export async function waitForAgentReady(page: Page, expectedTailText?: string): Promise<void> {
-  await expect(page.getByTestId("agent-chat-scroll")).toBeVisible({ timeout: 60000 });
+  await expect(getVisibleChatScroll(page)).toBeVisible({ timeout: 60000 });
   await expect(page.getByRole("textbox", { name: "Message agent..." }).first()).toBeVisible({
     timeout: 60000,
   });
@@ -251,7 +263,7 @@ export async function expectDetachedFromBottom(page: Page): Promise<void> {
 
 export async function waitForContentGrowth(
   page: Page,
-  previousContentHeight: number
+  previousContentHeight: number,
 ): Promise<ScrollMetrics> {
   await expect
     .poll(async () => {
@@ -263,11 +275,9 @@ export async function waitForContentGrowth(
 }
 
 export async function getChatContainerKey(page: Page): Promise<string | null> {
-  return page
-    .getByTestId("agent-chat-scroll")
-    .evaluate((element) => {
-      const nativeId = (element as HTMLElement).id;
-      const prefix = "agent-chat-scroll-";
-      return nativeId.startsWith(prefix) ? nativeId.slice(prefix.length) : null;
-    });
+  return getVisibleChatScroll(page).evaluate((element) => {
+    const nativeId = (element as HTMLElement).id;
+    const prefix = "agent-chat-scroll-";
+    return nativeId.startsWith(prefix) ? nativeId.slice(prefix.length) : null;
+  });
 }

packages/app/e2e/helpers/app.ts

@@ -1,16 +1,19 @@
-import { expect, type Page } from '@playwright/test';
+import { expect, type Page } from "@playwright/test";
+import { buildCreateAgentPreferences, buildSeededHost } from "./daemon-registry";
 
 function escapeRegex(value: string): string {
-  return value.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
+  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).');
+    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.');
+  if (port === "6767") {
+    throw new Error(
+      "E2E_DAEMON_PORT is 6767. Refusing to run e2e against the default local daemon.",
+    );
   }
   return port;
 }
@@ -20,163 +23,203 @@ async function ensureE2EStorageSeeded(page: Page): Promise<void> {
   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).');
+    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 === 'direct' && typeof c?.endpoint === 'string' && /:6767\b/.test(c.endpoint))) return true;
-      return !connections.some((c: any) => c?.type === 'direct' && c?.endpoint === expectedEndpoint);
-    } catch {
-      return true;
-    }
-  }, { expectedEndpoint, expectedServerId });
+  const needsReset = await page.evaluate(
+    ({ expectedEndpoint: endpoint, expectedServerId: serverId }) => {
+      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 { serverId?: string; connections?: unknown };
+        if (entry?.serverId !== serverId) return true;
+        const connections = entry?.connections;
+        if (!Array.isArray(connections)) return true;
+        if (
+          connections.some(
+            (c: { type?: string; endpoint?: string }) =>
+              c?.type === "directTcp" &&
+              typeof c?.endpoint === "string" &&
+              /:6767\b/.test(c.endpoint),
+          )
+        )
+          return true;
+        return !connections.some(
+          (c: { type?: string; endpoint?: string }) =>
+            c?.type === "directTcp" && c?.endpoint === endpoint,
+        );
+      } 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(
-    ({ expectedEndpoint, nowIso, expectedServerId }) => {
-      localStorage.setItem('@paseo:e2e', '1');
-      localStorage.setItem(
-        '@paseo:daemon-registry',
-        JSON.stringify([
-          {
-            serverId: expectedServerId,
-            label: 'localhost',
-            connections: [
-              {
-                id: `direct:${expectedEndpoint}`,
-                type: 'direct',
-                endpoint: expectedEndpoint,
-              },
-            ],
-            preferredConnectionId: `direct:${expectedEndpoint}`,
-            createdAt: nowIso,
-            updatedAt: nowIso,
-          },
-        ])
-      );
-      localStorage.setItem(
-        '@paseo:create-agent-preferences',
-        JSON.stringify({
-          serverId: expectedServerId,
-          provider: 'codex',
-          providerPreferences: {
-            claude: { model: 'haiku' },
-            codex: { model: 'gpt-5.1-codex-mini', thinkingOptionId: 'low' },
-          },
-        })
-      );
-      localStorage.removeItem('@paseo:settings');
+    ({ daemon: seededDaemon, preferences: seededPreferences }) => {
+      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");
     },
-    { expectedEndpoint, nowIso, expectedServerId }
+    { daemon, preferences },
   );
 
   await page.reload();
 }
 
+function parseRegistryEntry(registryRaw: string): { serverId: string; connections: unknown } {
+  let registry: unknown;
+  try {
+    registry = JSON.parse(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] as { serverId?: string; connections?: unknown };
+  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)}).`,
+    );
+  }
+  return { serverId: daemon.serverId, connections: daemon.connections };
+}
+
+function assertDaemonConnections(connections: unknown, expectedEndpoint: string): void {
+  if (
+    !Array.isArray(connections) ||
+    !connections.some(
+      (c: { type?: string; endpoint?: string }) =>
+        c?.type === "directTcp" && c?.endpoint === expectedEndpoint,
+    )
+  ) {
+    throw new Error(
+      `E2E expected seeded daemon connections to include directTcp ${expectedEndpoint} (got ${JSON.stringify(connections)}).`,
+    );
+  }
+  if (
+    Array.isArray(connections) &&
+    connections.some(
+      (c: { type?: string; endpoint?: string }) =>
+        c?.type === "directTcp" && typeof c?.endpoint === "string" && /:6767\b/.test(c.endpoint),
+    )
+  ) {
+    throw new Error(
+      `E2E detected a daemon endpoint pointing at :6767 (${JSON.stringify(connections)}).`,
+    );
+  }
+}
+
+function assertPreferencesMatch(prefsRaw: string, serverId: string): void {
+  try {
+    const prefs = JSON.parse(prefsRaw) as { serverId?: string };
+    if (prefs?.serverId !== serverId) {
+      throw new Error(
+        `E2E expected create-agent-preferences.serverId to match seeded daemon serverId (${serverId}) (got ${String(prefs?.serverId)}).`,
+      );
+    }
+  } catch (error) {
+    if (error instanceof Error) throw error;
+    throw new Error("E2E expected @paseo:create-agent-preferences to be valid JSON.", {
+      cause: error,
+    });
+  }
+}
+
 async function assertE2EUsesSeededTestDaemon(page: Page): Promise<void> {
   const port = getE2EDaemonPort();
   const expectedEndpoint = `127.0.0.1:${port}`;
   const expectedServerId = process.env.E2E_SERVER_ID;
   if (!expectedServerId) {
-    throw new Error('E2E_SERVER_ID is not set (expected from Playwright globalSetup).');
+    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');
+    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.');
+    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) {
+  const { serverId, connections } = parseRegistryEntry(snapshot.registryRaw);
+  if (serverId !== expectedServerId) {
     throw new Error(
-      `E2E expected @paseo:daemon-registry to contain exactly 1 daemon (got ${Array.isArray(registry) ? registry.length : 'non-array'}).`
+      `E2E expected seeded daemon serverId to be ${expectedServerId} (got ${serverId}).`,
     );
   }
-
-  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 === 'direct' && c?.endpoint === expectedEndpoint)
-  ) {
-    throw new Error(
-      `E2E expected seeded daemon connections to include direct ${expectedEndpoint} (got ${JSON.stringify(connections)}).`
-    );
-  }
-  if (Array.isArray(connections) && connections.some((c: any) => c?.type === 'direct' && typeof c?.endpoint === 'string' && /:6767\b/.test(c.endpoint))) {
-    throw new Error(`E2E detected a daemon endpoint pointing at :6767 (${JSON.stringify(connections)}).`);
-  }
+  assertDaemonConnections(connections, expectedEndpoint);
 
   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.');
+    throw new Error("E2E expected @paseo:create-agent-preferences to be set before app load.");
   }
+  assertPreferencesMatch(snapshot.prefsRaw, serverId);
 }
 
-export const gotoHome = async (page: Page) => {
-  await page.goto('/');
+export const gotoAppShell = async (page: Page) => {
+  await page.goto("/");
   await ensureE2EStorageSeeded(page);
-  await expect(page.getByText('New agent', { exact: true }).first()).toBeVisible();
-  const composer = page.getByRole('textbox', { name: 'Message agent...' });
-  if (!(await composer.first().isVisible().catch(() => false))) {
-    const newAgentButton = page.getByText('New agent', { exact: true }).first();
-    await newAgentButton.click();
+};
+
+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();
+
+    await expect
+      .poll(
+        async () =>
+          (await addProjectCta.isVisible().catch(() => false)) ||
+          (await addProjectSidebar.isVisible().catch(() => false)) ||
+          (await newAgentButton.isVisible().catch(() => false)),
+        { timeout: 10000 },
+      )
+      .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();
+    }
   }
   await expect(composer.first()).toBeVisible({ timeout: 30000 });
 };
 
 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) => {
@@ -185,23 +228,19 @@ export const setWorkingDirectory = async (page: Page, directory: string) => {
     .first();
   await expect(workingDirectorySelect).toBeVisible({ timeout: 30000 });
 
-  const legacyInput = page.getByRole('textbox', { name: '/path/to/project' }).first();
-  const directorySearchInput = page.getByRole('textbox', { name: /search directories/i }).first();
-  const worktreePicker = page.getByTestId('worktree-attach-picker');
-  const worktreeSheetTitle = page.getByText('Select worktree', { exact: true }).first();
+  const legacyInput = page.getByRole("textbox", { name: "/path/to/project" }).first();
+  const directorySearchInput = page.getByRole("textbox", { name: /search directories/i }).first();
+  const worktreePicker = page.getByTestId("worktree-attach-picker");
+  const worktreeSheetTitle = page.getByText("Select worktree", { exact: true }).first();
   const closeBottomSheet = async () => {
-    const bottomSheetBackdrop = page
-      .getByRole('button', { name: 'Bottom sheet backdrop' })
-      .first();
-    const bottomSheetHandle = page
-      .getByRole('slider', { name: 'Bottom sheet handle' })
-      .first();
+    const bottomSheetBackdrop = page.getByRole("button", { name: "Bottom sheet backdrop" }).first();
+    const bottomSheetHandle = page.getByRole("slider", { name: "Bottom sheet handle" }).first();
     for (let attempt = 0; attempt < 3; attempt += 1) {
       if (!(await bottomSheetBackdrop.isVisible())) {
         return;
       }
       await bottomSheetBackdrop.click({ force: true });
-      await page.keyboard.press('Escape').catch(() => undefined);
+      await page.keyboard.press("Escape").catch(() => undefined);
       await page.waitForTimeout(200);
     }
     if (await bottomSheetBackdrop.isVisible()) {
@@ -221,7 +260,7 @@ export const setWorkingDirectory = async (page: Page, directory: string) => {
     if (!(await worktreeSheetTitle.isVisible()) && !(await worktreePicker.isVisible())) {
       return;
     }
-    const attachToggle = page.getByTestId('worktree-attach-toggle');
+    const attachToggle = page.getByTestId("worktree-attach-toggle");
     if (await attachToggle.isVisible()) {
       await attachToggle.click({ force: true });
       await page.waitForTimeout(200);
@@ -242,26 +281,23 @@ export const setWorkingDirectory = async (page: Page, directory: string) => {
       await closeBottomSheet();
       await workingDirectorySelect.click({ force: true });
     }
-    await expect
-      .poll(async () => pickerInputVisible(), { timeout: 10000 })
-      .toBe(true);
+    await expect.poll(async () => pickerInputVisible(), { timeout: 10000 }).toBe(true);
   }
 
-  const trimmedDirectory = directory.replace(/\/+$/, '');
-  const activeInput =
-    (await directorySearchInput.isVisible().catch(() => false))
-      ? directorySearchInput
-      : legacyInput;
+  const trimmedDirectory = directory.replace(/\/+$/, "");
+  const activeInput = (await directorySearchInput.isVisible().catch(() => false))
+    ? directorySearchInput
+    : legacyInput;
 
   await activeInput.fill(trimmedDirectory);
 
   if (activeInput === directorySearchInput) {
     // Combobox custom rows can be either plain path labels or prefixed labels.
     const plainOption = page
-      .getByText(new RegExp(`^${escapeRegex(trimmedDirectory)}$`, 'i'))
+      .getByText(new RegExp(`^${escapeRegex(trimmedDirectory)}$`, "i"))
       .first();
     const prefixedUseOption = page
-      .getByText(new RegExp(`^Use "${escapeRegex(trimmedDirectory)}"$`, 'i'))
+      .getByText(new RegExp(`^Use "${escapeRegex(trimmedDirectory)}"$`, "i"))
       .first();
 
     if (await plainOption.isVisible().catch(() => false)) {
@@ -270,33 +306,38 @@ export const setWorkingDirectory = async (page: Page, directory: string) => {
       await prefixedUseOption.click({ force: true });
     } else {
       // Fallback: accept highlighted option (directory suggestion).
-      await activeInput.press('Enter');
+      await activeInput.press("Enter");
     }
   } else {
     // Legacy path picker fallback.
-    await activeInput.press('Enter');
+    await activeInput.press("Enter");
   }
 
   // Wait for picker to close.
   await expect(activeInput).not.toBeVisible({ timeout: 10000 });
 
   const directoryCandidates = new Set<string>([trimmedDirectory]);
-  if (trimmedDirectory.startsWith('/var/')) {
+  if (trimmedDirectory.startsWith("/var/")) {
     directoryCandidates.add(`/private${trimmedDirectory}`);
   }
-  if (trimmedDirectory.startsWith('/private/var/')) {
-    directoryCandidates.add(trimmedDirectory.replace(/^\/private/, ''));
+  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(async () => {
-    const text = await workingDirectorySelect.innerText().catch(() => '');
-    if (text.includes(basename)) return true;
-    for (const candidate of directoryCandidates) {
-      if (text.includes(candidate)) return true;
-    }
-    return false;
-  }, { timeout: 30000 }).toBe(true);
+  await expect
+    .poll(
+      async () => {
+        const text = await workingDirectorySelect.innerText().catch(() => "");
+        if (text.includes(basename)) return true;
+        for (const candidate of directoryCandidates) {
+          if (text.includes(candidate)) return true;
+        }
+        return false;
+      },
+      { timeout: 30000 },
+    )
+    .toBe(true);
 };
 
 export const ensureHostSelected = async (page: Page) => {
@@ -314,19 +355,21 @@ export const ensureHostSelected = async (page: Page) => {
     }
 
     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 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 Array<{ serverId?: string }>;
+      const prefs = JSON.parse(prefsRaw) as { serverId?: string };
+      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;
+      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));
+      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);
+      const nonce = localStorage.getItem("@paseo:e2e-seed-nonce") ?? "1";
+      localStorage.setItem("@paseo:e2e-disable-default-seed-once", nonce);
       return { ok: true } as const;
     });
 
@@ -338,20 +381,22 @@ export const ensureHostSelected = async (page: Page) => {
     await assertE2EUsesSeededTestDaemon(page);
   }
 
-  const input = page.getByRole('textbox', { name: 'Message agent...' });
+  const input = page.getByRole("textbox", { name: "Message agent..." });
   await expect(input).toBeVisible();
 
   if (await input.isEditable()) {
     return;
   }
 
-  const selectHostLabel = page.getByText('Select host', { exact: true });
+  const selectHostLabel = page.getByText("Select host", { exact: true });
   if (await selectHostLabel.isVisible()) {
     await selectHostLabel.click();
 
     // E2E safety: 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 }).first();
+    const localhostOption = page.getByText("localhost", { exact: true }).first();
+    const daemonIdOption = page
+      .getByText(process.env.E2E_SERVER_ID ?? "srv_e2e_test_daemon", { exact: true })
+      .first();
 
     if (await localhostOption.isVisible()) {
       await localhostOption.click();
@@ -365,11 +410,11 @@ export const ensureHostSelected = async (page: Page) => {
 };
 
 export const createAgent = async (page: Page, message: string) => {
-  const input = page.getByRole('textbox', { name: 'Message agent...' });
+  const input = page.getByRole("textbox", { name: "Message agent..." });
   await expect(input).toBeEditable();
   await preferFastThinkingOption(page);
   await input.fill(message);
-  await input.press('Enter');
+  await input.press("Enter");
 
   // The composer may remain on the draft screen briefly while the initial run starts,
   // so assert the user-visible result instead of forcing one route shape here.
@@ -381,21 +426,23 @@ export const createAgent = async (page: Page, message: string) => {
 
 async function preferFastThinkingOption(page: Page): Promise<void> {
   const providerTrigger = page
-    .locator('[data-testid="agent-provider-selector"]:visible, [data-testid="draft-provider-select"]:visible')
+    .locator(
+      '[data-testid="agent-provider-selector"]:visible, [data-testid="draft-provider-select"]:visible',
+    )
     .first();
   if (await providerTrigger.isVisible().catch(() => false)) {
-    const providerText = ((await providerTrigger.innerText().catch(() => '')) ?? '').trim();
+    const providerText = ((await providerTrigger.innerText().catch(() => "")) ?? "").trim();
     if (!/codex/i.test(providerText)) {
       return;
     }
   }
 
-  const thinkingTrigger = page.getByTestId('agent-thinking-selector').first();
+  const thinkingTrigger = page.getByTestId("agent-thinking-selector").first();
   if (!(await thinkingTrigger.isVisible().catch(() => false))) {
     return;
   }
 
-  const currentThinkingLabel = ((await thinkingTrigger.innerText().catch(() => '')) ?? '')
+  const currentThinkingLabel = ((await thinkingTrigger.innerText().catch(() => "")) ?? "")
     .trim()
     .toLowerCase();
   if (/\b(low|minimal|off)\b/.test(currentThinkingLabel)) {
@@ -403,16 +450,16 @@ async function preferFastThinkingOption(page: Page): Promise<void> {
   }
 
   await thinkingTrigger.click();
-  const menu = page.getByTestId('agent-thinking-menu').first();
+  const menu = page.getByTestId("agent-thinking-menu").first();
   if (!(await menu.isVisible().catch(() => false))) {
     return;
   }
 
-  const preferredLabels = ['low', 'minimal', 'off', 'medium'];
+  const preferredLabels = ["low", "minimal", "off", "medium"];
   let selected = false;
   for (const label of preferredLabels) {
     const option = menu
-      .getByRole('button', { name: new RegExp(`^${escapeRegex(label)}$`, 'i') })
+      .getByRole("button", { name: new RegExp(`^${escapeRegex(label)}$`, "i") })
       .first();
     if (await option.isVisible().catch(() => false)) {
       await option.click({ force: true });
@@ -422,11 +469,11 @@ async function preferFastThinkingOption(page: Page): Promise<void> {
   }
 
   if (!selected) {
-    const options = menu.getByRole('button');
+    const options = menu.getByRole("button");
     const count = await options.count();
     for (let index = 0; index < count; index += 1) {
       const option = options.nth(index);
-      const label = ((await option.innerText().catch(() => '')) ?? '').trim();
+      const label = ((await option.innerText().catch(() => "")) ?? "").trim();
       if (!label) {
         continue;
       }
@@ -440,7 +487,7 @@ async function preferFastThinkingOption(page: Page): Promise<void> {
   }
 
   if (!selected) {
-    await page.keyboard.press('Escape').catch(() => undefined);
+    await page.keyboard.press("Escape").catch(() => undefined);
     return;
   }
 
@@ -458,15 +505,17 @@ export interface AgentConfig {
 export const selectProvider = async (page: Page, provider: string) => {
   const normalizedProvider = provider.trim();
   if (!normalizedProvider) {
-    throw new Error('Provider must be a non-empty string.');
+    throw new Error("Provider must be a non-empty string.");
   }
 
   const providerTrigger = page
-    .locator('[data-testid="agent-provider-selector"]:visible, [data-testid="draft-provider-select"]:visible')
+    .locator(
+      '[data-testid="agent-provider-selector"]:visible, [data-testid="draft-provider-select"]:visible',
+    )
     .first();
   if (
     await providerTrigger
-      .getByText(new RegExp(`^${escapeRegex(normalizedProvider)}$`, 'i'))
+      .getByText(new RegExp(`^${escapeRegex(normalizedProvider)}$`, "i"))
       .first()
       .isVisible()
       .catch(() => false)
@@ -477,20 +526,18 @@ export const selectProvider = async (page: Page, provider: string) => {
   if (await providerTrigger.isVisible().catch(() => false)) {
     await providerTrigger.click();
   } else {
-    const providerLabel = page.getByText('PROVIDER', { exact: true }).first();
+    const providerLabel = page.getByText("PROVIDER", { exact: true }).first();
     await expect(providerLabel).toBeVisible();
     await providerLabel.click();
   }
 
-  const dialog = page.getByRole('dialog').last();
-  const searchInput = dialog.getByRole('textbox', { name: /search provider/i }).first();
+  const dialog = page.getByRole("dialog").last();
+  const searchInput = dialog.getByRole("textbox", { name: /search provider/i }).first();
   if (await searchInput.isVisible().catch(() => false)) {
     await searchInput.fill(normalizedProvider);
   }
 
-  const option = dialog
-    .getByText(new RegExp(`^${escapeRegex(normalizedProvider)}$`, 'i'))
-    .first();
+  const option = dialog.getByText(new RegExp(`^${escapeRegex(normalizedProvider)}$`, "i")).first();
   await expect(option).toBeVisible();
   await option.click();
 };
@@ -498,15 +545,17 @@ export const selectProvider = async (page: Page, provider: string) => {
 export const selectModel = async (page: Page, model: string) => {
   const normalizedModel = model.trim();
   if (!normalizedModel) {
-    throw new Error('Model must be a non-empty string.');
+    throw new Error("Model must be a non-empty string.");
   }
 
   const modelTrigger = page
-    .locator('[data-testid="agent-model-selector"]:visible, [data-testid="draft-model-select"]:visible')
+    .locator(
+      '[data-testid="agent-model-selector"]:visible, [data-testid="draft-model-select"]:visible',
+    )
     .first();
   if (
     await modelTrigger
-      .getByText(new RegExp(`^${escapeRegex(normalizedModel)}$`, 'i'))
+      .getByText(new RegExp(`^${escapeRegex(normalizedModel)}$`, "i"))
       .first()
       .isVisible()
       .catch(() => false)
@@ -517,21 +566,21 @@ export const selectModel = async (page: Page, model: string) => {
   if (await modelTrigger.isVisible().catch(() => false)) {
     await modelTrigger.click();
   } else {
-    const modelLabel = page.getByText('MODEL', { exact: true }).first();
+    const modelLabel = page.getByText("MODEL", { exact: true }).first();
     await expect(modelLabel).toBeVisible();
     await modelLabel.click();
   }
 
   // Wait for the model dropdown to open
-  const searchInput = page.getByRole('textbox', { name: /search model/i });
+  const searchInput = page.getByRole("textbox", { name: /search model/i });
   await expect(searchInput).toBeVisible({ timeout: 10000 });
 
   // Type to search/filter models
   await searchInput.fill(normalizedModel);
 
-  const dialog = page.getByRole('dialog');
+  const dialog = page.getByRole("dialog");
   const exactOption = dialog
-    .getByText(new RegExp(`^${escapeRegex(normalizedModel)}$`, 'i'))
+    .getByText(new RegExp(`^${escapeRegex(normalizedModel)}$`, "i"))
     .first();
   const exactVisible = await exactOption.isVisible().catch(() => false);
   if (exactVisible) {
@@ -539,39 +588,39 @@ export const selectModel = async (page: Page, model: string) => {
   } else {
     // Modern labels include version suffixes (for example "Haiku 4.5"), so
     // select the first filtered result using keyboard confirm.
-    await searchInput.press('Enter');
+    await searchInput.press("Enter");
   }
 
   // Wait for dropdown to close
   if (await searchInput.isVisible().catch(() => false)) {
-    await page.keyboard.press('Escape').catch(() => undefined);
+    await page.keyboard.press("Escape").catch(() => undefined);
   }
   await expect(searchInput).not.toBeVisible({ timeout: 5000 });
 };
 
 export const selectMode = async (page: Page, mode: string) => {
   const modeTrigger = page
-    .locator('[data-testid="agent-mode-selector"]:visible, [data-testid="draft-mode-select"]:visible')
+    .locator(
+      '[data-testid="agent-mode-selector"]:visible, [data-testid="draft-mode-select"]:visible',
+    )
     .first();
   if (await modeTrigger.isVisible().catch(() => false)) {
     await modeTrigger.click();
   } else {
-    const modeLabel = page.getByText('MODE', { exact: true }).first();
+    const modeLabel = page.getByText("MODE", { exact: true }).first();
     await expect(modeLabel).toBeVisible();
     await modeLabel.click();
   }
 
   // Wait for the mode dropdown to open
-  const searchInput = page.getByRole('textbox', { name: /search mode/i });
+  const searchInput = page.getByRole("textbox", { name: /search mode/i });
   await expect(searchInput).toBeVisible({ timeout: 10000 });
 
   // Type to filter modes
   await searchInput.fill(mode);
 
-  const dialog = page.getByRole('dialog');
-  const option = dialog
-    .getByText(new RegExp(`^${escapeRegex(mode)}$`, 'i'))
-    .first();
+  const dialog = page.getByRole("dialog");
+  const option = dialog.getByText(new RegExp(`^${escapeRegex(mode)}$`, "i")).first();
   await expect(option).toBeVisible();
   await option.click({ force: true });
 
@@ -601,7 +650,7 @@ export const createAgentWithConfig = async (page: Page, config: AgentConfig) =>
 
 export const createAgentInRepo = async (
   page: Page,
-  config: Pick<AgentConfig, 'directory' | 'prompt'>
+  config: Pick<AgentConfig, "directory" | "prompt">,
 ) => {
   await gotoHome(page);
   await ensureHostSelected(page);
@@ -610,25 +659,25 @@ export const createAgentInRepo = async (
 };
 
 export const waitForPermissionPrompt = async (page: Page, timeout = 30000) => {
-  const promptText = page.getByTestId('permission-request-question').first();
+  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();
+  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();
+  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 });
+  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);
@@ -645,9 +694,11 @@ export async function waitForAgentFinishUI(page: Page, timeout = 30000) {
     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
-    });
+    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 });

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

@@ -0,0 +1,312 @@
+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 }>;
+}
+
+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,
+  });
+}
+
+function getSessionRowByTitle(page: Page, title: string) {
+  return page.locator('[data-testid^="agent-row-"]').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 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/daemon-registry.ts

@@ -0,0 +1,39 @@
+export const TEST_HOST_LABEL = "localhost";
+
+export const TEST_PROVIDER_PREFERENCES = {
+  claude: { model: "haiku" },
+  codex: { model: "gpt-5.4-mini", thinkingOptionId: "low" },
+} as const;
+
+export function buildDirectTcpConnection(endpoint: string) {
+  return {
+    id: `direct:${endpoint}`,
+    type: "directTcp" as const,
+    endpoint,
+  };
+}
+
+export function buildSeededHost(input: {
+  serverId: string;
+  endpoint: string;
+  label?: string;
+  nowIso: string;
+}) {
+  const connection = buildDirectTcpConnection(input.endpoint);
+  return {
+    serverId: input.serverId,
+    label: input.label ?? TEST_HOST_LABEL,
+    connections: [connection],
+    preferredConnectionId: connection.id,
+    createdAt: input.nowIso,
+    updatedAt: input.nowIso,
+  };
+}
+
+export function buildCreateAgentPreferences(serverId: string) {
+  return {
+    serverId,
+    provider: "codex" as const,
+    providerPreferences: TEST_PROVIDER_PREFERENCES,
+  };
+}

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/launcher.ts

@@ -0,0 +1,203 @@
+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 ───────────────────────────────────────────────────────────
+
+/** Click the new agent tab button in the tab bar. Creates a draft/chat tab directly. */
+export async function clickNewTabButton(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 in the workspace tab bar. Creates a terminal tab directly. */
+export async function clickNewTerminalButton(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();
+}
+
+/** 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 ───────────────────────────────────────────────────
+
+/** @deprecated The launcher panel was removed. Actions go directly to their target. */
+export async function waitForLauncherPanel(_page: Page): Promise<void> {
+  // No-op: the launcher panel no longer exists.
+}
+
+/** 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 clickTerminal(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;
+}
+
+// ─── 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 +1,252 @@
+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 },
+): Promise<void> {
+  await openNewWorkspaceComposer(page, input);
+  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.getByTestId("new-workspace-ref-picker-trigger");
+  await expect(trigger).toContainText(`#${input.number}`);
+  await expect(trigger).toContainText(input.title);
+  await expect(trigger).not.toContainText(input.headRef);
+}
+
+export async function expectComposerGithubAttachmentPill(
+  page: Page,
+  input: { number: number; title: string },
+): Promise<void> {
+  const pills = page.getByTestId("composer-github-attachment-pill");
+  await expect(pills).toHaveCount(1);
+  await expect(pills.first()).toContainText(`#${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 };
+}

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/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

@@ -0,0 +1,274 @@
+import 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 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,
+  ): Promise<{
+    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 }>;
+  sendTerminalInput(
+    terminalId: string,
+    message: { type: "input"; data: string } | { type: "resize"; rows: number; cols: number },
+  ): void;
+  onTerminalStreamEvent(
+    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;
+  if (!daemonPort) {
+    throw new Error("E2E_DAEMON_PORT is not set.");
+  }
+  return `ws://127.0.0.1:${daemonPort}/ws`;
+}
+
+function getServerId(): string {
+  const serverId = process.env.E2E_SERVER_ID;
+  if (!serverId) {
+    throw new Error("E2E_SERVER_ID is not set.");
+  }
+  return serverId;
+}
+
+interface TerminalPerfDaemonClientConfig {
+  url: string;
+  clientId: string;
+  clientType: "cli";
+  webSocketFactory?: NodeWebSocketFactory;
+}
+
+async function loadDaemonClientConstructor(): Promise<
+  new (config: TerminalPerfDaemonClientConfig) => TerminalPerfDaemonClient
+> {
+  const repoRoot = path.resolve(__dirname, "../../../../");
+  const moduleUrl = pathToFileURL(
+    path.join(repoRoot, "packages/server/dist/server/server/exports.js"),
+  ).href;
+  const mod = (await import(moduleUrl)) as {
+    DaemonClient: new (config: TerminalPerfDaemonClientConfig) => TerminalPerfDaemonClient;
+  };
+  return mod.DaemonClient;
+}
+
+export async function connectTerminalClient(): Promise<TerminalPerfDaemonClient> {
+  const DaemonClient = await loadDaemonClientConstructor();
+  const webSocketFactory = createNodeWebSocketFactory();
+  const client = new DaemonClient({
+    url: getDaemonWsUrl(),
+    clientId: `terminal-perf-${randomUUID()}`,
+    clientType: "cli",
+    webSocketFactory,
+  });
+  await client.connect();
+  return client;
+}
+
+export function buildTerminalWorkspaceUrl(workspaceId: string, terminalId: string): string {
+  const serverId = getServerId();
+  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 Window & {
+        __paseoTerminal?: {
+          buffer: {
+            active: {
+              length: number;
+              getLine: (i: number) => { translateToString: (trim: boolean) => string } | null;
+            };
+          };
+          onWriteParsed: (cb: () => void) => { dispose: () => void };
+        };
+      }
+    ).__paseoTerminal;
+    if (!term) {
+      return "";
+    }
+    const buf = term.buffer.active;
+    const lines: string[] = [];
+    for (let i = 0; i < buf.length; i++) {
+      const line = buf.getLine(i);
+      if (line) {
+        lines.push(line.translateToString(true));
+      }
+    }
+    return lines.join("\n");
+  });
+}
+
+export async function waitForTerminalContent(
+  page: Page,
+  predicate: (text: string) => boolean,
+  timeout: number,
+): Promise<void> {
+  const deadline = Date.now() + timeout;
+  while (Date.now() < deadline) {
+    const text = await getTerminalBufferText(page);
+    if (predicate(text)) {
+      return;
+    }
+    await page.waitForTimeout(50);
+  }
+  throw new Error(`Terminal content did not match predicate within ${timeout}ms`);
+}
+
+export async function navigateToTerminal(
+  page: Page,
+  input: { workspaceId: string; terminalId: string },
+): Promise<void> {
+  // Boot the app at the workspace route directly.
+  // The fixtures.ts beforeEach addInitScript seeds localStorage on every navigation,
+  // so the daemon registry is already configured when the app starts.
+  const workspaceRoute = buildTerminalWorkspaceUrl(input.workspaceId, input.terminalId);
+  await page.goto(workspaceRoute);
+
+  // The workspace layout consumes `?open=...`, returns null during the effect,
+  // then replaces the URL with the clean workspace route after preparing the tab.
+  // On CI, Expo Router's rootNavigationState may take time to initialize,
+  // so we allow a generous timeout here.
+  const cleanWorkspaceRoute = buildWorkspaceUrl(input.workspaceId);
+  await page.waitForURL(
+    (url) => url.pathname === cleanWorkspaceRoute && !url.searchParams.has("open"),
+    { timeout: 30_000 },
+  );
+
+  // Wait for daemon connection (sidebar shows host label)
+  await page
+    .getByText("localhost", { exact: true })
+    .first()
+    .waitFor({ state: "visible", timeout: 30_000 });
+
+  // The open intent should have prepared and focused the exact pre-created terminal tab.
+  // The tab reconciliation effect also auto-creates terminal tabs once hydration completes,
+  // so we give it enough time for the full workspace hydration + tab creation cycle.
+  const terminalTab = page.locator(`[data-testid="workspace-tab-terminal_${input.terminalId}"]`);
+  await terminalTab.waitFor({ state: "visible", timeout: 30_000 });
+  await terminalTab.click();
+
+  const terminalSurface = page.locator('[data-testid="terminal-surface"]');
+  await terminalSurface.waitFor({ state: "visible", timeout: 15_000 });
+
+  // Wait for loading overlay to disappear (terminal attached)
+  await page
+    .locator('[data-testid="terminal-attach-loading"]')
+    .waitFor({ state: "hidden", timeout: 10_000 })
+    .catch(() => {
+      // overlay may never appear if attachment is instant
+    });
+
+  await terminalSurface.scrollIntoViewIfNeeded();
+  await terminalSurface.click();
+}
+
+export async function setupDeterministicPrompt(page: Page, sentinel?: string): Promise<void> {
+  const tag = sentinel ?? `READY_${Date.now()}`;
+  const terminal = page.locator('[data-testid="terminal-surface"]');
+
+  await terminal.pressSequentially(`echo ${tag}\n`, { delay: 0 });
+  await waitForTerminalContent(page, (text) => text.includes(tag), 10_000);
+
+  await terminal.pressSequentially("export PS1='$ '\n", { delay: 0 });
+  await page.waitForTimeout(300);
+}
+
+export interface LatencySample {
+  char: string;
+  latencyMs: number;
+}
+
+/**
+ * Measures keystroke echo round-trip latency.
+ *
+ * Starts a high-resolution timer on the browser keydown event (capture phase)
+ * and stops it when xterm.js finishes parsing the echoed write. This measures
+ * the full path: keydown → WebSocket → daemon PTY echo → WebSocket → xterm render.
+ */
+export async function measureKeystrokeLatency(page: Page, char: string): Promise<number> {
+  await page.evaluate(() => {
+    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 = (win.__perfKeystroke = {
+      promise: null as Promise<number> | null,
+    });
+
+    state.promise = new Promise<number>((resolve, reject) => {
+      const timeout = setTimeout(() => {
+        document.removeEventListener("keydown", onKeyDown, true);
+        reject(new Error("keystroke echo timeout (5s)"));
+      }, 5000);
+
+      function onKeyDown() {
+        document.removeEventListener("keydown", onKeyDown, true);
+        const start = performance.now();
+        const disposable = term.onWriteParsed(() => {
+          clearTimeout(timeout);
+          disposable.dispose();
+          resolve(performance.now() - start);
+        });
+      }
+
+      document.addEventListener("keydown", onKeyDown, true);
+    });
+  });
+
+  await page.keyboard.press(char);
+
+  return page.evaluate(
+    () =>
+      (window as unknown as { __perfKeystroke: { promise: Promise<number> } }).__perfKeystroke
+        .promise,
+  );
+}
+
+export function computePercentile(samples: number[], p: number): number {
+  const sorted = [...samples].sort((a, b) => a - b);
+  const index = Math.ceil((p / 100) * sorted.length) - 1;
+  return sorted[Math.max(0, index)];
+}
+
+export function round2(value: number): number {
+  return Math.round(value * 100) / 100;
+}

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

@@ -0,0 +1,713 @@
+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 as Uint8Array);
+            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 as Uint8Array);
+            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/workspace-lifecycle.ts

@@ -0,0 +1,32 @@
+import { expect, type Page } from "@playwright/test";
+import { clickNewChat, clickTerminal } 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 clickTerminal(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 +1,341 @@
+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 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;
+  } | 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: String(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: String(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) => String(entry.id) === workspaceId) ?? null;
+  if (!workspace) {
+    throw new Error(`Workspace not found: ${workspaceId}`);
+  }
+  return workspace;
+}
+
+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))) {
@@ -38,7 +72,7 @@ export async function ensureWorkspaceAgentPaneVisible(page: Page): Promise<void>
 
 export async function sampleWorkspaceTabIds(
   page: Page,
-  options: { durationMs?: number; intervalMs?: number } = {}
+  options: { durationMs?: number; intervalMs?: number } = {},
 ): Promise<string[][]> {
   const durationMs = options.durationMs ?? 2_500;
   const intervalMs = options.intervalMs ?? 50;

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

@@ -1,38 +1,72 @@
-import { expect, type Page } from '@playwright/test';
-import { buildHostWorkspaceRoute } from '@/utils/host-routes';
-import { gotoHome } from './app';
+import { expect, type Page } from "@playwright/test";
+import { buildHostWorkspaceRoute } from "@/utils/host-routes";
+import { gotoHome } from "./app";
 
 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(/\/+$/, '');
-  const parts = normalized.split('/').filter(Boolean);
+  const normalized = value.replace(/\\/g, "/").replace(/\/+$/, "");
+  const parts = normalized.split("/").filter(Boolean);
   return parts[parts.length - 1] ?? normalized;
 }
 
 function escapeRegex(value: string): string {
-  return value.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
+  return value.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
 }
 
 function candidateWorkspaceIds(inputPath: string): string[] {
-  const trimmed = inputPath.replace(/\/+$/, '');
+  const trimmed = inputPath.replace(/\/+$/, "");
   const candidates = new Set<string>([trimmed]);
-  if (trimmed.startsWith('/var/')) {
+  if (trimmed.startsWith("/var/")) {
     candidates.add(`/private${trimmed}`);
   }
-  if (trimmed.startsWith('/private/var/')) {
-    candidates.add(trimmed.replace(/^\/private/, ''));
+  if (trimmed.startsWith("/private/var/")) {
+    candidates.add(trimmed.replace(/^\/private/, ""));
   }
   return Array.from(candidates);
 }
 
 function workspaceRowLocator(page: Page, serverId: string, workspacePath: string) {
   const ids = candidateWorkspaceIds(workspacePath).map(
-    (id) => `[data-testid="sidebar-workspace-row-${serverId}:${id}"]`
+    (id) => `[data-testid="sidebar-workspace-row-${serverId}:${id}"]`,
   );
-  return page.locator(ids.join(',')).first();
+  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: {
@@ -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 }
+  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,
@@ -66,9 +115,9 @@ export async function expectWorkspaceHeader(
 }
 
 export async function seedWorkspaceActivity(page: Page, marker: string): Promise<void> {
-  const input = page.getByRole('textbox', { name: 'Message agent...' });
+  const input = page.getByRole("textbox", { name: "Message agent..." });
   await expect(input).toBeEditable({ timeout: 30_000 });
   await input.fill(marker);
-  await input.press('Enter');
+  await input.press("Enter");
   await expect(page).toHaveURL(/\/workspace\//, { timeout: 30_000 });
 }

packages/app/e2e/helpers/workspace.ts

@@ -1,43 +1,120 @@
-import { execSync } from 'node:child_process';
-import { mkdtemp, writeFile, rm, mkdir } from 'node:fs/promises';
-import { tmpdir } from 'node:os';
-import path from 'node:path';
+import { execSync } from "node:child_process";
+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>;
-};
+}
 
 export const createTempGitRepo = async (
-  prefix = 'paseo-e2e-',
-  options?: { withRemote?: boolean }
+  prefix = "paseo-e2e-",
+  options?: {
+    withRemote?: boolean;
+    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;
 
-  execSync('git init -b main', { cwd: repoPath, stdio: 'ignore' });
-  execSync('git config user.email "e2e@paseo.test"', { cwd: repoPath, stdio: 'ignore' });
-  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');
-  execSync('git add README.md', { cwd: repoPath, stdio: 'ignore' });
-  execSync('git commit -m "Initial commit"', { cwd: repoPath, stdio: 'ignore' });
+  execSync("git init -b main", { cwd: repoPath, stdio: "ignore" });
+  execSync('git config user.email "e2e@paseo.test"', { cwd: repoPath, stdio: "ignore" });
+  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" });
+
+  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" });
+  }
 
   if (withRemote) {
     // Deterministic local remote to avoid relying on external auth/network in e2e.
-    const remoteDir = path.join(repoPath, 'remote.git');
+    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' });
+    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 {
     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/host-onboarding-direct.spec.ts

@@ -1,44 +0,0 @@
-import { test, expect } from './fixtures';
-
-test('no hosts shows welcome; direct connection adds host and lands on agent create', async ({ page }) => {
-  const daemonPort = process.env.E2E_DAEMON_PORT;
-  const serverId = process.env.E2E_SERVER_ID;
-  if (!daemonPort) {
-    throw new Error('E2E_DAEMON_PORT is not set (expected from globalSetup).');
-  }
-  if (!serverId) {
-    throw new Error('E2E_SERVER_ID is not set (expected from globalSetup).');
-  }
-
-  await page.addInitScript(() => {
-    localStorage.setItem('@paseo:daemon-registry', JSON.stringify([]));
-    localStorage.removeItem('@paseo:create-agent-preferences');
-    localStorage.removeItem('@paseo:settings');
-  });
-
-  await page.goto('/');
-
-  await expect(page.getByText('Welcome to Paseo', { exact: true })).toBeVisible();
-
-  await page.getByText('Direct connection', { exact: true }).click();
-
-  await page.getByPlaceholder('host:6767').fill(`127.0.0.1:${daemonPort}`);
-
-  await page.getByText('Connect', { exact: true }).click();
-
-  // First-time connection prompts for an optional label.
-  const nameModal = page.getByTestId('name-host-modal');
-  const showedNameModal = await nameModal
-    .waitFor({ state: 'visible', timeout: 5000 })
-    .then(() => true)
-    .catch(() => false);
-  if (showedNameModal) {
-    await nameModal.getByTestId('name-host-skip').click();
-  }
-
-  await expect(page.getByTestId('sidebar-new-agent')).toBeVisible();
-  await expect(page.getByText(serverId, { exact: true })).toBeVisible();
-  await expect(page.getByRole('textbox', { name: 'Message agent...' })).toBeEditable({
-    timeout: 15000,
-  });
-});

packages/app/e2e/host-removal.spec.ts

@@ -1,119 +0,0 @@
-import { test, expect } from './fixtures';
-
-test('host removal removes the host from UI and persists after reload', async ({ page }) => {
-  const daemonPort = process.env.E2E_DAEMON_PORT;
-  const seededServerId = process.env.E2E_SERVER_ID;
-  if (!daemonPort) {
-    throw new Error('E2E_DAEMON_PORT is not set (expected from globalSetup).');
-  }
-  if (!seededServerId) {
-    throw new Error('E2E_SERVER_ID is not set (expected from globalSetup).');
-  }
-
-  const extraPort = Number(daemonPort) + 1;
-  const extraEndpoint = `127.0.0.1:${extraPort}`;
-  const nowIso = new Date().toISOString();
-
-  const extraDaemon = {
-    serverId: 'srv_e2e_extra_daemon',
-    label: 'extra',
-    connections: [
-      { id: `direct:${extraEndpoint}`, type: 'direct', endpoint: extraEndpoint },
-    ],
-    preferredConnectionId: `direct:${extraEndpoint}`,
-    createdAt: nowIso,
-    updatedAt: nowIso,
-  };
-
-  const seededTestDaemon = {
-    serverId: seededServerId,
-    label: 'localhost',
-    connections: [
-      { id: `direct:127.0.0.1:${daemonPort}`, type: 'direct', endpoint: `127.0.0.1:${daemonPort}` },
-    ],
-    preferredConnectionId: `direct:127.0.0.1:${daemonPort}`,
-    createdAt: nowIso,
-    updatedAt: nowIso,
-  };
-
-  const seedOnceKey = `@paseo:e2e-host-removal-seeded:${Math.random().toString(36).slice(2)}`;
-
-  // Add a second host once (fixtures seed the primary host on every navigation).
-  await page.addInitScript(
-    ({ daemon, seededTestDaemon, seedOnceKey }) => {
-      if (localStorage.getItem(seedOnceKey)) {
-        return;
-      }
-
-      const raw = localStorage.getItem('@paseo:daemon-registry');
-      let parsed: any[] = [];
-      if (raw) {
-        try {
-          parsed = JSON.parse(raw);
-        } catch {
-          parsed = [];
-        }
-      }
-      const list = Array.isArray(parsed) ? parsed : [];
-      localStorage.setItem(seedOnceKey, '1');
-
-      const next = [...list];
-      const hasSeeded = next.some((entry: any) => entry && entry.serverId === seededTestDaemon.serverId);
-      if (!hasSeeded) {
-        next.push(seededTestDaemon);
-      }
-
-      const alreadyPresent = next.some((entry: any) => entry && entry.serverId === daemon.serverId);
-      if (!alreadyPresent) {
-        next.push(daemon);
-      }
-
-      localStorage.setItem('@paseo:daemon-registry', JSON.stringify(next));
-    },
-    { daemon: extraDaemon, seededTestDaemon, seedOnceKey }
-  );
-
-  await page.goto('/settings');
-
-  await expect(page.getByText('extra', { exact: true }).first()).toBeVisible();
-  await expect(page.getByText(extraEndpoint, { exact: true }).first()).toBeVisible();
-
-  const hostSettingsButton = page.getByTestId(`daemon-card-settings-${extraDaemon.serverId}`).first();
-  await expect(hostSettingsButton).toBeVisible({ timeout: 10000 });
-  await hostSettingsButton.click();
-
-  const hostDetailModal = page.getByTestId('host-detail-modal');
-  await expect(hostDetailModal).toBeVisible({ timeout: 10000 });
-  await hostDetailModal.getByText('Advanced', { exact: true }).click();
-  await page.getByText('Remove host', { exact: true }).last().click();
-
-  await expect(page.getByTestId('remove-host-confirm-modal')).toBeVisible();
-  await page.getByTestId('remove-host-confirm').click();
-
-  await expect(page.getByTestId(`daemon-card-${extraDaemon.serverId}`)).toHaveCount(0, {
-    timeout: 30000,
-  });
-  await expect(page.getByText(extraEndpoint, { exact: true })).toHaveCount(0);
-  await page.waitForFunction(
-    (serverId) => {
-      const raw = localStorage.getItem('@paseo:daemon-registry');
-      if (!raw) return false;
-      try {
-        const parsed = JSON.parse(raw);
-        return Array.isArray(parsed) && !parsed.some((entry: any) => entry && entry.serverId === serverId);
-      } catch {
-        return false;
-      }
-    },
-    extraDaemon.serverId,
-    { timeout: 10000 }
-  );
-
-  // Prevent the fixture from overwriting storage on reload; verify persistence.
-  await page.evaluate(() => {
-    const nonce = localStorage.getItem('@paseo:e2e-seed-nonce') ?? '1';
-    localStorage.setItem('@paseo:e2e-disable-default-seed-once', nonce);
-  });
-  await page.reload();
-  await expect(page.getByText(extraEndpoint, { exact: true })).toHaveCount(0);
-});

packages/app/e2e/host-selection.spec.ts

@@ -1,124 +0,0 @@
-import { test, expect } from './fixtures';
-import { ensureHostSelected, gotoHome } from './helpers/app';
-
-test('new agent auto-selects the previous host', async ({ page }) => {
-  await gotoHome(page);
-  await ensureHostSelected(page);
-
-  await gotoHome(page);
-
-  // The selected host should be restored after a full reload without manual selection.
-  await expect(page.getByText('localhost', { exact: true }).first()).toBeVisible();
-  const input = page.getByRole('textbox', { name: 'Message agent...' });
-  await expect(input).toBeEditable({ timeout: 30000 });
-});
-
-test('new agent respects serverId in the URL', async ({ page }) => {
-  const daemonPort = process.env.E2E_DAEMON_PORT;
-  const serverId = process.env.E2E_SERVER_ID;
-  if (!daemonPort) {
-    throw new Error('E2E_DAEMON_PORT is not set (expected from globalSetup).');
-  }
-  if (!serverId) {
-    throw new Error('E2E_SERVER_ID is not set (expected from globalSetup).');
-  }
-
-  // Ensure this test's storage is deterministic even under parallel load.
-  const nowIso = new Date().toISOString();
-  const testDaemon = {
-    serverId,
-    label: 'localhost',
-    connections: [
-      {
-        id: `direct:127.0.0.1:${daemonPort}`,
-        type: 'direct',
-        endpoint: `127.0.0.1:${daemonPort}`,
-      },
-    ],
-    preferredConnectionId: `direct:127.0.0.1:${daemonPort}`,
-    createdAt: nowIso,
-    updatedAt: nowIso,
-  };
-  const createAgentPreferences = {
-    serverId: testDaemon.serverId,
-    provider: 'codex',
-    providerPreferences: {
-      claude: { model: 'haiku' },
-      codex: { model: 'gpt-5.1-codex-mini', thinkingOptionId: 'low' },
-    },
-  };
-
-  await page.goto('/settings');
-  await page.evaluate(
-    ({ daemon, preferences }) => {
-      const nonce = localStorage.getItem('@paseo:e2e-seed-nonce') ?? '1';
-      localStorage.setItem('@paseo:e2e-disable-default-seed-once', nonce);
-      localStorage.setItem('@paseo:daemon-registry', JSON.stringify([daemon]));
-      localStorage.setItem('@paseo:create-agent-preferences', JSON.stringify(preferences));
-      localStorage.removeItem('@paseo:settings');
-    },
-    { daemon: testDaemon, preferences: createAgentPreferences }
-  );
-  await page.reload();
-  await expect(page.getByText('New agent', { exact: true }).first()).toBeVisible({ timeout: 20000 });
-
-  await page.goto(`/?serverId=${encodeURIComponent(serverId)}`);
-  await expect(page.getByText('New agent', { exact: true }).first()).toBeVisible();
-
-  const newAgentButton = page.getByTestId('sidebar-new-agent').first();
-  if (await newAgentButton.isVisible().catch(() => false)) {
-    await newAgentButton.click();
-  } else {
-    await page.getByText('New agent', { exact: true }).first().click();
-  }
-
-  const input = page.getByRole('textbox', { name: 'Message agent...' });
-  await expect(input).toBeEditable({ timeout: 30000 });
-});
-
-test('new agent auto-selects first online host when no preference is stored', async ({ page }) => {
-  const daemonPort = process.env.E2E_DAEMON_PORT;
-  const serverId = process.env.E2E_SERVER_ID;
-  if (!daemonPort) {
-    throw new Error('E2E_DAEMON_PORT is not set (expected from globalSetup).');
-  }
-  if (!serverId) {
-    throw new Error('E2E_SERVER_ID is not set (expected from globalSetup).');
-  }
-
-  const nowIso = new Date().toISOString();
-  const testDaemon = {
-    serverId,
-    label: 'localhost',
-    connections: [
-      {
-        id: `direct:127.0.0.1:${daemonPort}`,
-        type: 'direct',
-        endpoint: `127.0.0.1:${daemonPort}`,
-      },
-    ],
-    preferredConnectionId: `direct:127.0.0.1:${daemonPort}`,
-    createdAt: nowIso,
-    updatedAt: nowIso,
-  };
-
-  await gotoHome(page);
-  await page.evaluate(
-    ({ daemon }) => {
-      const nonce = localStorage.getItem('@paseo:e2e-seed-nonce') ?? '1';
-      localStorage.setItem('@paseo:e2e-disable-default-seed-once', nonce);
-      localStorage.setItem('@paseo:daemon-registry', JSON.stringify([daemon]));
-      localStorage.removeItem('@paseo:create-agent-preferences');
-      localStorage.removeItem('@paseo:settings');
-    },
-    { daemon: testDaemon }
-  );
-
-  await page.reload();
-  await expect(page.getByText('New agent', { exact: true }).first()).toBeVisible();
-
-  // Host should be auto-selected (no manual selection required).
-  await expect(page.getByText('localhost', { exact: true }).first()).toBeVisible();
-  const input = page.getByRole('textbox', { name: 'Message agent...' });
-  await expect(input).toBeEditable({ timeout: 30000 });
-});

packages/app/e2e/keyboard-shortcuts-dialog.spec.ts

@@ -1,17 +0,0 @@
-import { test, expect } from "./fixtures";
-import { gotoHome } from "./helpers/app";
-
-test("question mark opens keyboard shortcuts dialog", async ({ page }) => {
-  await gotoHome(page);
-  await page.getByTestId("menu-button").first().focus();
-
-  await page.keyboard.press("Shift+/");
-
-  const dialog = page.getByTestId("keyboard-shortcuts-dialog");
-  const content = page.getByTestId("keyboard-shortcuts-dialog-content");
-
-  await expect(dialog).toBeVisible({ timeout: 10000 });
-  await expect(content).toBeVisible({ timeout: 10000 });
-  await expect(content).toContainText("Show keyboard shortcuts");
-  await expect(content).toContainText("Toggle left sidebar");
-});

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

@@ -0,0 +1,257 @@
+import { test, expect } from "./fixtures";
+import { createTempGitRepo } from "./helpers/workspace";
+import {
+  gotoWorkspace,
+  assertNewChatTileVisible,
+  assertTerminalTileVisible,
+  assertSingleNewTabButton,
+  clickNewTabButton,
+  pressNewTabShortcut,
+  clickNewChat,
+  clickTerminal,
+  countTabsOfKind,
+  getTabTestIds,
+  waitForTabWithTitle,
+  measureTileTransition,
+  sampleTabsDuringTransition,
+} from "./helpers/launcher";
+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 = String(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);
+
+    // Should show the composer directly (no launcher panel)
+    const composer = page.getByRole("textbox", { name: "Message agent..." });
+    await expect(composer.first()).toBeVisible({ timeout: 15_000 });
+  });
+
+  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 clickNewTabButton(page);
+
+    // Draft composer should appear (the agent message input)
+    const composer = page.getByRole("textbox", { name: "Message agent..." });
+    await expect(composer.first()).toBeVisible({ timeout: 15_000 });
+
+    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 clickTerminal(page);
+
+    // Terminal surface should appear
+    const terminal = page.locator('[data-testid="terminal-surface"]');
+    await expect(terminal.first()).toBeVisible({ timeout: 20_000 });
+
+    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: ${result.error}`);
+    const terminalId = result.terminal.id;
+
+    try {
+      // Navigate to workspace and open a terminal
+      await gotoWorkspace(page, workspaceId);
+      await clickTerminal(page);
+
+      const terminal = page.locator('[data-testid="terminal-surface"]');
+      await expect(terminal.first()).toBeVisible({ timeout: 20_000 });
+      await terminal.first().click();
+
+      await setupDeterministicPrompt(page);
+
+      // Send OSC 0 (set window title) escape sequence
+      const testTitle = `E2E-Title-${Date.now()}`;
+      await terminal
+        .first()
+        .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: ${result.error}`);
+    const terminalId = result.terminal.id;
+
+    try {
+      await gotoWorkspace(page, workspaceId);
+      await clickTerminal(page);
+
+      const terminal = page.locator('[data-testid="terminal-surface"]');
+      await expect(terminal.first()).toBeVisible({ timeout: 20_000 });
+      await terminal.first().click();
+
+      await setupDeterministicPrompt(page);
+
+      // Fire many rapid title changes — only the last should stick
+      const finalTitle = `Final-${Date.now()}`;
+      for (let i = 0; i < 5; i++) {
+        await terminal
+          .first()
+          .pressSequentially(`printf '\\033]0;Rapid-${i}\\007'\n`, { delay: 0 });
+      }
+      await terminal
+        .first()
+        .pressSequentially(`printf '\\033]0;${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 terminal = page.locator('[data-testid="terminal-surface"]');
+    const elapsed = await measureTileTransition(
+      page,
+      () => clickTerminal(page),
+      terminal.first(),
+      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 composer = page.getByRole("textbox", { name: "Message agent..." }).first();
+
+    const elapsed = await measureTileTransition(page, () => clickNewChat(page), composer, 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);
+  });
+});