Exercise the rendered SVG, dimensions, theme color, catalog artwork and fallback through the real browser app instead of inspecting hook results in JSDOM. The browser regression fails on main before the lookup fix.
Terminal profile names are registered independently of the ACP catalog. Include their shipped artwork in icon resolution and verify the current reactive hooks return its SVG.
Co-authored-by: Noah <noahguerin20@gmail.com>
* Stack workspace sheets and tighten sidebar density
- Workspace sheets stack: opening a target from a sheet covers it, Back and
drag close only the top sheet, and the header is the sheet's drag handle.
- The workspace sheet reuses the shared sheet handle, background and header.
- Fix the shared sheet handle indicator rendering black.
- Sidebar text returns to semantic sizes; Pinned and Workspaces share one
section header.
* Pause covered workspace sheets and mock the shared sheet handle in tests
* Export BottomSheetHandle from the browser test Gorhom stub
* Let a stacked sheet's backdrop catch taps on iOS
* Bring a revealed sheet to the top instead of duplicating it as a tab
Unfinished CSS strings in font preferences consumed the shared theme stylesheet. Reject incomplete syntax at the settings boundary, recover persisted malformed values through the existing default path, and restore rejected edits in the font field.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Keep a matched heading together before splitting its image and text contents. Splitting first stranded the opening heading tag in a text slice without its closing tag.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* Let readFile callers disable the request timeout
readFile resolves only when FileEnd arrives, so the default 60 s request
timeout bounds the whole transfer. A large file over a slow relay fails
before it finishes even though every chunk is arriving. Thread an optional
timeout through requestFileExplorer; 0 disables the timer, as the
workspace.create request already does, and a dead connection still rejects
on disconnect.
The relay e2e case reads a 1 MiB file through a real daemon and a local
relay with timeout 0 and compares SHA-256 digests with the source.
* Reject a binary file transfer whose byte count differs from FileBegin.size
concatByteChunks allocated FileBegin.size bytes and copied the chunks in:
a short transfer resolved as a zero-padded file, and a long one threw a
RangeError inside the transport callback and left the request pending.
Count the received bytes at FileEnd and reject the read when they differ
from the announced size.
* Download files over the session socket when no direct connection is active
The download store only knew the HTTP token route and picked the first
directTcp connection in the host profile, whatever connection the host
runtime was using. A relay-only host has none, so every download ended in
"Download host is unavailable." (#543). SSH, socket and pipe hosts failed
the same way, and a stale directTcp entry won over a working relay (#3753).
Choose the transport from the active connection: directTcp keeps the HTTP
token download unchanged; anything else reads the file over the session
socket with the binary frames previews already use, then hands the bytes
to the browser's download manager on web and Electron or writes them to
the cache directory and offers the share sheet on native. Files over
64 MiB are refused before any byte moves, using the existing fileTooLarge
string, and a daemon response without content fails instead of saving an
empty file.
* fix(app): enforce the session download size limit during the read
The token size is sampled before the file is read. Pass the download byte budget through the actual read so a changed file cannot bypass the in-memory limit.
Co-authored-by: Michael Vandesompel <mv@droggol.be>
Co-authored-by: YASICJUNWOO <joonoo3@naver.com>
Co-authored-by: JrDw0 <JrDw0@users.noreply.github.com>
* test: verify TCP downloads and own relay fixture cleanup
---------
Co-authored-by: Michael Vandesompel <mv@droggol.be>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: YASICJUNWOO <joonoo3@naver.com>
Co-authored-by: JrDw0 <JrDw0@users.noreply.github.com>
Relative navigation shared the list capped for digit shortcuts, so it wrapped at the ninth workspace. Keep all sidebar targets and cap only numbered badges. Carry forward #4128 with keyboard-routing and browser regressions.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Raisal P Wardana <IEatCodeDaily@users.noreply.github.com>
* fix(app): keep long custom models removable
Custom model labels and IDs could consume the row width and push Remove outside the sheet. Allow those text cells to shrink while preserving the button width and short-name alignment.
* test(app): wait for legacy agent creation before checking prompts
Agent-list membership includes initialization. Waiting for finish at that point can return on the initial idle transition before the initial prompt starts. Wait for the matching creation response before checking each completed turn, preserving the existing exact-once assertions.
* test(app): wait for directory search keyboard focus
The search page becomes visible before its deferred focus callback runs. Wait for the input to own keyboard focus so the journey types its query rather than dropping it during navigation.
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* Show Claude option previews in the question card
* Wrap long lines in question option previews
* Check option previews under the option that owns them
* Show the whole option preview on native so every line is reachable
On Android, a slow drag inside a preview capped at 240px did not scroll
it, so lines past the cap could not be read. The preview's nested
scroll sits inside the option's Pressable; the same drag scrolls it
once the preview is outside the Pressable. Native now renders the
preview at full height so the chat scroll reaches every line, and a tap
on it still selects the option. Web keeps the 240px cap.
The mock agent's rollout preview is now taller than the cap so the
Android replay can check the last line.
---------
Co-authored-by: Gabriel Giordano <hello@gabrielgiordano.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(app): respect provider label for xhigh thinking option
Only fall back to the localized Extra high string for bare forms.
Custom labels such as XHigh (56 | 8m) now render with their suffix.
* fix(app): show Extra high only for a bare xhigh label
A configured xhigh label with extra detail keeps its text; only a label
that is just xhigh or Extra high uses the localized string.
Co-authored-by: Elior Fureraj <efureraj@gmail.com>
---------
Co-authored-by: Elior Fureraj <efureraj@gmail.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* feat(app): open files and subagents in a closable sheet on compact
On phones, opening a file, a provider subagent, or a Paseo subagent now shows it
in a bottom sheet over the current tab instead of adding a tab, so the tab strip
stays for terminals and root agents. Paseo subagents get "Open as tab" to move
the same tab (same tabId and state) into a full tab with a composer; agents in a
sheet hide the composer.
The decision lives in the workspace layout store's openTab, driven by a
per-kind `compactSheet` rule in the panel manifest, so every open path (chat
links, Explorer, subagents track, notifications, plugins) follows it without
call-site branching, and a new kind opts in with one manifest entry. The app
shell publishes the layout mode (compact / single-pane / split-panes) once, so
callers stop passing isCompact for placement.
A sheet holding unsaved edits is never implicitly displaced; later opens go to
tabs underneath, and closing it goes through the existing discard guard. The
editor host is position: sticky because CodeMirror's scrollIntoView otherwise
scrolls the bottom sheet's container and drags the sheet to full height.
* test(app): fail the compact sheet snap check when the sheet is oversized
---------
Co-authored-by: paseo-bot[bot] <paseo-bot[bot]@users.noreply.github.com>
The New workspace screen submitted the composer's remembered mode as-is,
so a saved mode for a provider that offers no modes (or no longer offers
that one) failed creation with "Invalid mode '<mode>' for provider
'<provider>'". A new chat tab and the workspace setup dialog already
reconciled the mode against the provider's modes, each with its own copy
of the rule.
The draft composer state now exposes effectiveModeId, the mode the picker
shows, and all three create paths submit it. While the provider is still
loading, the selected mode is kept because its modes aren't known yet.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Matt Cowger <matt@cowger.us>
* Keep pairing QR code within compact tile
* Move pairing screenshot after viewport assertion
* test(pairing): verify the complete QR code on Android
The native regression failed with the fixed-size SVG because the code was clipped at both screen edges. Verify the complete square and visible margins through the Android app, and keep browser geometry assertions in the pairing helper.
Co-authored-by: fdkevin <25277753+fdkevin0@users.noreply.github.com>
---------
Co-authored-by: fdkevin <25277753+fdkevin0@users.noreply.github.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(app): run recorded letter shortcuts with non-Latin layouts
Recording unsupported letters saves the physical A-Z key, but dispatch rejected that same event. Accept its letter code while keeping ASCII character matching for layouts such as Dvorak.
* fix(app): run recorded dotted-letter shortcuts
Check the original single-letter key before case folding expands uppercase dotted I into two code points.
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(app): only offer Remove for providers defined in config.json
Plugin-registered providers (Antigravity, Muse Code) share source: "custom"
with config-defined ones, so Settings → Providers offered Remove on them.
The removeProviders patch only deletes config.json entries, so the action
succeeded silently while the plugin re-registered the provider.
Gate the Remove menu on the provider having an extends entry in config —
the actual definition custom providers live behind. Plugin providers keep
their enable/disable switch.
Fixes#6349
* Address review: named ownership check, E2E coverage for plugin providers
- Name the config-owned provider check instead of combining three
conditions inline in canRemove.
- Replace the JSDOM component tests with a Playwright case in the
existing provider-removal suite: a plugin-registered provider renders
without the Remove action and survives a removeProviders patch.
* fix(app): render local relative images in markdown file preview
File preview dropped relative images because it only accepted http/data URIs and never resolved paths against the opened markdown file.
* test(app): cover local relative images in markdown preview e2e
* fix(app): show local images in Markdown file previews
Carry forward document-relative image loading and preserve the host home
directory when resolving images beside home-relative Markdown files.
Decode file URIs once, and require decoded images in the file-pane regression.
Co-authored-by: Ouyang Xingyuan <galaxyroaming@qq.com>
* fix(app): resolve local Markdown image URL suffixes
---------
Co-authored-by: Ouyang Xingyuan <galaxyroaming@qq.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(app): install navigator polyfill before any app module loads
@xterm/headless reads navigator.userAgent at module evaluation. Since
#6455 the root layout transitively imports it, so the layout's
module-level polyfillNavigator() ran too late and native builds crashed
at launch with "Cannot read property 'includes' of undefined".
Install the crypto, navigator and screen.orientation polyfills from a
side-effect module imported first by index.ts. Calls written in index.ts
itself ran after all hoisted imports had evaluated.
Fixes#6504
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test(app): isolate native startup regression from shared globals
Exercise the real navigator polyfill and terminal dependency in fresh VM contexts, including the pre-initialization failure. Preserve a check that entry imports install navigator before the app loads.
Co-authored-by: panrafal <rl@stamina.pl>
* test(app): follow startup import lint rule
---------
Co-authored-by: panrafal <rl@stamina.pl>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(app): sync Android system bar icon style with app theme
With edge-to-edge enabled the app draws under the Android system bars,
but the bars keep their default light icon style regardless of theme,
so the status bar clock/battery/icons are invisible on light themes
(white on white). Call SystemBars.setStyle from the appearance provider
whenever the theme is applied: dark icons on light themes, light icons
on dark themes, auto following the system for adaptive themes and
plugin-contributed themes.
Refs #1272 (runtime only; the splash screen phase is not covered).
* fix(app): keep Android status icons readable as themes change
Keep system-bar contrast subscribed to the resolved app theme so automatic appearance changes update the icons as well as the app surface. Add a native regression covering explicit and automatic themes.
Co-authored-by: TeraYang <4110825+terayang@users.noreply.github.com>
* fix(app): map system-bar appearance through color scheme
Unistyles reserves style mappings for React Native style objects. Keep SystemBars style selection inside its appearance adapter and map the colorScheme prop instead.
---------
Co-authored-by: TeraYang <4110825+terayang@users.noreply.github.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(app): jump the chat outline to prompts a timeline plugin draws
A timeline transformer replaces a stream item with plugin rows whose ids
are "<pluginId>/<itemId>". The chat outline, chat find and history reveal
look a message up by the source item's id, so once a plugin drew the
prompts the outline found no row: the rail showed the prompts and
clicking one did nothing, for mounted, virtualized and unloaded prompts
alike.
Projected plugin rows now carry the id of the item they replaced, and
getStreamItemMessageId answers with it. Row wrappers, the virtualized
index lookup, the visible-message set and history reveal all read that
function, so they now find the plugin row.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test(app): verify outline jumps to virtualized plugin prompts
Replacing every prompt keeps all loaded rows mounted, so the original plugin regression never exercised a virtualized target. Retain newer native prompts and verify that a loaded unmounted plugin prompt lands without fetching.
Co-authored-by: luopeihuan <luopeihuan@roundreddot.com>
* test(app): retain outline navigation screenshots
---------
Co-authored-by: luopeihuan <luopeihuan@roundreddot.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* feat(app): add fullscreen editing to the compact composer
* fix(app): preserve read-only terminal command previews
Static command previews must render their published source; the live editor buffer is updated after render and has no mounted input to repaint read-only content.
* fix(app): import React for composer editing scope rendering
Workspace pane unit tests use the classic JSX transform, so the editing-scope provider needs React in scope when mounted.
* fix(app): preserve fullscreen composer controls and editing
Keep the fullscreen frame inside the visual viewport, gate hidden autocomplete keys, apply replacement commands to local line measurement, and return to inline before starting voice so its controls stay accessible.
* test(app): cover draft replacement through the production composer
Replace the measurement-hook fixture with profile switching and typing through the real composer. Keep measurement state inside its production surface; Android manual red/green evidence covers the native replacement flow.
Expo declares NODE_ENV readonly after web startup. Preserve the early test environment setup with Object.assign and exercise the real configuration with Expo types in a failing-first compiler regression.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(app): record shortcuts using the keyboard layout character
Shortcut capture stored the QWERTY position while dispatch matched the logical character, so a Dvorak B became Cmd+N. Use the existing logical key mappings for capture and preserve code fallback for named and rewritten Option keys.
* fix(app): keep named shortcut keys on their supported codes
Character capture must not advertise unsupported named keys. Restrict the logical lookup to printable mappings so keypad Enter keeps its existing unsupported behavior.
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
The html-ish pass feeds markdown to htmlparser2 so it can render
<details>, <img> and friends. htmlparser2 reads any `<` followed by a
letter as a tag and collects attributes up to the next `>`. A cell such
as `1<ratio<2` in a table therefore swallowed the rows after it, and the
re-serialized "tag" dropped the newlines and pipes that made them rows.
The same path broke autolinks (`<https://x/y>` became `<https: x y>`)
and cut text after a trailing `a<b`.
Only hand htmlparser2 what CommonMark reads as raw HTML: open tags,
closing tags and comments. Every other `<` stays text.
Unknown tags are rendered back as text, but they were rebuilt from the
parser's lowercased name and attributes, so `As<Integer>()` showed as
`As<integer>()`. Echo the source slice instead.
* fix(tests): run React tests from production shells
Vitest preserves an inherited NODE_ENV, selecting React without its test APIs. Pin the environment before dependency resolution and worker startup in both test entry points.
Co-authored-by: Ivan Malison <IvanMalison@gmail.com>
* test(app): expect zero edits for a pure staged rename
The rename patch correction on main preserves unchanged content. Keep the discard journey intact and update its obsolete added-line assertion.
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Ivan Malison <IvanMalison@gmail.com>
Preserve returned change request bodies as search-only option text so client filtering keeps description matches. Keep title and ID matches ranked ahead of secondary text.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(app): decode percent-encoded assistant file links
markdown-it percent-encodes non-ASCII and space characters in link
hrefs, so a link like [x](docs/reports/报告.md) arrives as
docs/reports/%E6%8A%A5%E5%91%8A.md. The file:// and bare absolute-path
branches already run safeDecodeURIComponent, but the workspace-relative,
path:line, Windows, and home-relative branches did not, so clicking the
link tried to open the still-encoded name and failed with ENOENT.
Decode the path portion before the relative-path heuristics and in the
inline absolute/Windows branches so all href forms behave the same.
Generated with [Devin](https://devin.ai)
Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(app): decode hrefs only, never raw inline-code paths
Raw inline-code and linkified tokens reach the same parser as markdown
hrefs. Decoding unconditionally would send a real file named
my%20file.md to my file.md. Gate every decode on AssistantHrefParseOptions.decodeHref — set by the resolver only when the token
came from source.href — and extract parseWindowsFileLink to keep the
entry point under the complexity limit. Adds parse-level and
resolver-level regression tests for literal %20 filenames.
Generated with [Devin](https://devin.ai)
Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(app): keep raw absolute paths out of URL parsing
new URL() percent-encodes non-ASCII, so a raw inline-code path like
/tmp/开户.md became /tmp/%E5%BC%80%E6%88%B7.md when decodeHref was
false. Hrefs still go through URL parsing (normalizes dot segments
before decode); raw paths split the # fragment manually instead.
Generated with [Devin](https://devin.ai)
Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(app): keep encoded and literal file links distinct
Preserve inline-code provenance for Windows paths and key cached file lookups by their resolved query, path, and line range. Reproduce the reported non-ASCII Markdown link through the renderer and filesystem, with browser coverage for clicking it.
Co-authored-by: Ouyang Xingyuan <galaxyroaming@qq.com>
Co-authored-by: dengjiawen <dengjiawen@hortorgames.com>
* test(app): assert file content after opening the markdown link
Co-authored-by: sunshinehomeyu-blip <267551119+sunshinehomeyu-blip@users.noreply.github.com>
* fix(app): preserve encoded line markers in filenames
---------
Co-authored-by: Ouyang Xingyuan <galaxyroaming@qq.com>
Co-authored-by: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: dengjiawen <dengjiawen@hortorgames.com>
Co-authored-by: sunshinehomeyu-blip <267551119+sunshinehomeyu-blip@users.noreply.github.com>
* feat(app): modernize compact headers and sidebar affordances
Use a shared island header with quick tab creation and keep sheet footers inside their visible layout. Opt sidebar rows into touch sizing without changing global appearance settings.
* test(app): match darker light-theme workspace selection
* test(app): align plugin journeys with compact header
* fix: prevent file mention autocomplete from triggering on emails
Skip @ when preceded by a word character or +, and reject angle
brackets in mention queries so email addresses no longer open the
file autocomplete or steal Enter.
Closes#1364
* test(app): verify emails send intact without file completion
The word-character and plus guard tests previously passed because their queries included trailing spaces. Exercise each guard with the cursor at the end of the email, and cover Enter in the browser alongside normal file completion.
Co-authored-by: Jason <mildreins@gmail.com>
---------
Co-authored-by: Jason <mildreins@gmail.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(android): declare the camera optional so camera-less devices can install
Requesting CAMERA makes Android treat android.hardware.camera as a required
feature, so Google Play marks devices without a camera as incompatible.
Pairing works without QR scanning, so declare the camera features optional.
Fixes#6303
* test(app): resolve the manifest with the Expo CLI's prebuild packages
Load @expo/prebuild-config and @expo/config-plugins from @expo/cli instead
of the older hoisted copies, and set the F-Droid flag explicitly for each
build profile case.
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* feat(review): edit compact diff comments in a sheet
* feat(review): send diff feedback to workspace agents
Send saved reviews from the diff while retaining file navigation and unrelated composer drafts. Clear acknowledged versions only so comments edited or added during submission remain available.
* fix(review): preserve drafts and coordinate feedback across surfaces
Share pending submission ownership by draft while keeping recipient menus local. Retain editable text across presentation changes and acknowledgement, and reserve safe-area clearance for floating actions.
* fix(review): center a small primary feedback button on desktop
* fix(review): restore focus to feedback button after closing menu
* feat(server): let clients create internal agents
The daemon's own branch-name and commit-message generators run as
internal agents: never persisted, listed, flagged, announced, or seen by
plugin lifecycle hooks. Nothing outside the daemon could ask for one,
because the wire config schema strips `internal` and the agent.create
transform re-pins it. A plugin that runs a one-turn helper per event
therefore left an archived session in History every time.
Add an optional `internal` flag to `create_agent_request` (inherited by
`agent.create.request`), copy it onto the config in the session's create
path, and skip provider session persistence for it like the generators
do. The client SDK forwards the flag on both creation paths and gates it
once on `server_info.features.internalAgents`.
Two daemon paths assumed every agent is public and had to follow:
`resolveAgentIdentifier` now accepts a live agent by exact id, since the
id set it matched against came from listings that exclude internal
agents and left `wait_for_finish` answering "Agent not found"; and
archiving an internal agent now closes it without writing the record
that `registry.applySnapshot` used to leave behind, so `autoArchive`
does not accumulate hidden files on disk.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* fix(server): keep internal agents whole through failure, archive, and wait
Review follow-ups for internal agents:
- A create that fails after the runtime is registered now closes the
internal agent before rethrowing, so a hidden runtime cannot outlive
the failed request. Public agents keep today's behaviour.
- Archiving an internal agent cascades to attached children like any
parent. cascadeArchiveChildren takes the parent record instead of
reading it back from storage, where an internal agent never is.
- The archived snapshot and last message of an internal agent stay
readable by exact id for ten minutes, so create-then-waitForFinish
returns the finished turn whether the wait lands before or after an
auto-archive. Storage plays that role for public agents.
- The internal-agent e2e scenarios live in the existing
create_agent_request lifecycle suite rather than a new file.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* fix(server): cascade to live internal children and keep a failed internal create's worktree until it closes
Internal children never reach storage, so the parent archive cascade scans
live agents for them as well. A failed internal create now clears the
created id only after the runtime is archived, so worktree cleanup cannot
delete the directory under a runtime that would not close. Stored agent
payloads carry lastError, so a retained failed helper reports its error
through wait_for_finish instead of the generic message.
* feat(cli,mcp): expose internal agents and document --no-wait
paseo run gains --internal and create_agent gains internal, both forwarded
to the create request. --no-wait becomes the documented spelling for
returning immediately, matching paseo send; --background and --detach stay
as hidden aliases under COMPAT tags.
* test: follow the --no-wait rename and stub getRetiredInternalAgent
* feat: list internal agents on request and add internal workspaces
Internal agents stay hidden by default, but a caller can now ask for
them: `fetch_agents_request` takes `filter.includeInternal`, snapshots
carry `internal: true`, and `paseo ls --internal` shows them in an
INTERNAL column. The app never passes the flag.
Workspaces get the same flag. `workspace.create.request` with
`internal: true` persists a hidden workspace: listings, update
subscriptions, sync reads, empty-project counting, open-by-path, MCP
`list_workspaces`, and plugin workspace hooks all skip it unless the
caller opts in with `filter.includeInternal`. Every agent created inside
an internal workspace is internal and unpersisted; AgentManager applies
that from the workspace id before the plugin hook so creators cannot
opt out. Archive and shutdown enumerate internal agents so nothing
inside an internal workspace outlives it.
Surfaces: `paseo workspace create --internal`, `paseo workspace ls
--internal`, `paseo run --internal --new-workspace`, MCP
`create_workspace({ internal })`, SDK `workspaces.create({ internal })`
and `list({ filter: { includeInternal } })`. Clients gate on
`features.internalWorkspaces`.
Co-Authored-By: Claude <noreply@anthropic.com>
* feat(workspaces): add background visibility with inherited creation defaults
* fix(cli): resolve workspace callers on the selected daemon
* test(workspaces): keep caller integration coverage in server suite
* test(cli): exercise workspace callers through CLI integration harness
* test(mcp): remove obsolete execution flags from browser fixtures
---------
Co-authored-by: marvin-ambrozio <marvin@gustavo.eng.br>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Co-authored-by: Mohamed Boudra <boudra.moha@gmail.com>
* fix(agents): preserve creation when the first prompt fails
A rejected initial prompt must leave a created agent with a turn error instead of returning the user to a draft whose creation key is already committed. Preserve receipt replay without repeating the prompt, and bound fork history to 100,000 characters before it becomes an attachment.
* test(creation): isolate user agent counters from background naming
Workspace naming uses internal provider sessions. Counting and rejecting those sessions made creation assertions depend on installed provider discovery timing. Keep discovery within the test adapters and scope observations to user-created agents.
* test(creation): retry temporary directory cleanup on Windows
Windows CI completed the rejection assertions but failed teardown with EBUSY. Match the daemon test harness bounded removal retries so transient directory locks do not fail creation tests.
* fix(agents): record rejected prompts before their errors
Startup rejection recorded only the system error, leaving the client to append its local prompt after it. Record the rejected submission at the terminal failure boundary so stream and fetched history agree on prompt-before-error ordering.
* fix(agents): require acceptance for delivery notifications
Retain rejected human submissions before their errors without projecting rejected inter-agent prompts as completed deliveries. Reproduces and fixes the notification regression exposed by merge CI.
* test(agents): retain rejected submissions across retries
The existing browser retry test assumed terminally rejected prompts disappeared. Assert the intended prompt-before-error history and a separate accepted retry while preserving draft, attachment, and ambiguous-steering checks.
* test(creation): allow integration latency in raw RPC responses
The peer inherited a one-second polling timeout despite exercising Git and provider startup. Reproduce the timeout with delayed startup and wait within an explicit integration budget while retaining all response and identity assertions.
* feat(agents): show agent messages and completion notifications with sender identity
* fix(agents): preserve human envelope examples and attachment context
* fix(agents): separate message provenance from timeline presentation
* fix(app): catch up provider subagents after the update feed reconnects
Subagent updates are not replayed across a reconnect, so a child that
finished while the app was disconnected stayed "working" until one
subagent was opened and its parent was listed again. Re-list every
parent the app holds once the update feed resubscribes.
Closes#6109
* fix(app): resync a subagent parent even when its first list failed
A parent's subagent track lists the parent once on mount. When that list
failed because the app was disconnected, the parent was never tracked,
so the reconnect resync skipped it and children that later arrived via
live updates stayed stale after the next drop. Track the parent when its
list is requested instead of when it succeeds.
* test(website): expect the plugin directory links now that they are on
1b88f1612 turned PLUGINS_LINKED on, but the e2e test still asserted the
directory was unlinked, failing the typecheck job on main and every PR.
Assert the navigation, footer, and Browse plugins links point at /plugins.
The header check now targets the navigation landmark the homepage renders;
the banner locator matched nothing, so it only passed while asserting zero.
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(claude): open autonomous turns only when Claude starts a turn
Paseo guessed from frame shape whether the main session had started work
on its own, and only a result could close the turn that guess opened. A
background subagent's task frames, its own background jobs'
task_notification and rate limit events all arrive with no parent tool
use id, so they looked like main-session work and opened a turn no result
would ever close. The agent then reported "running" while its main
session sat idle, and the app's message queue never drained.
Claude brackets every main-session turn itself: system/init opens it and
exactly one result closes it, whether a prompt, a wake-up after background
work, a slash command or an API error started it. Background subagents
never emit either on the main stream. An autonomous turn now opens only on
that init, and only when no foreground turn is open. The frame heuristic
is removed.
- Paseo launches the CLI with CLAUDE_CODE_EMIT_SESSION_STATE_EVENTS=1,
added as an overlay in buildSdkEnv so the SDK and the usage probe share
one env object. An "idle" session state closes an autonomous turn as a
safety net, never a foreground one. "running" opens nothing: Claude Code
2.1.280 stays "running" for as long as any background subagent is
alive. State frames never reach the timeline.
- An init clears a leftover pendingInterruptAbort. A new turn has begun,
so the interrupted one is over, and a stale flag would swallow the new
turn's result.
Tests that opened autonomous turns from bare frames now send the init
Claude sends. New coverage: helper frames never open a turn while the main
session is idle, a wake-up turn opens on init and closes on its result, a
folded steer keeps one turn while a turn Claude starts after the result
gets its own, an idle closes a turn no result closed, session state never
opens a turn, and the env flag reaches the SDK launch.
docs/agent-lifecycle.md records that only system/init and result bracket
a main-session turn, since the code cannot tell a reader why frame shape
fails.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(claude): interrupt Claude only while its main turn runs, and spare helpers
Every Stop, and every send that replaces a running turn (paseo send, MCP
send_agent_prompt, voice, the app's queue drain, Send now), ends in
query.interrupt(). Without the SDK's perTaskStopAffordance declaration
that call kills every running background subagent, even when the main
session is idle and there is nothing of ours to stop.
- Declare perTaskStopAffordance, so an interrupt aborts only the main
turn, like Esc in Claude Code. Claude stops a helper itself with its
TaskStop tool.
- Track whether a main-session turn is in flight from Claude's own
brackets: set by system/init and by a pushed message's
command_lifecycle "started" (/compact works for seconds before its
init), cleared by the result or an idle session state.
- Call query.interrupt() and set pendingInterruptAbort only when that
turn is in flight, or the cancelled turn's own message has already been
started. Otherwise withdraw our not-yet-started messages with
cancel_async_message instead: an idle interrupt reaches only helpers,
and with no result to follow it the flag would linger.
- Track the foreground prompt with the queued steers until Claude starts
it, so Stop withdraws a message still waiting in the queue. A turn
stopped while the Claude process is still starting never sends it.
- Accept command_lifecycle "cancelled" as terminal, so a cancelled steer
no longer waits to be withdrawn or keeps denying permission cards.
Interrupt tests now wait for the pump to read the init before stopping,
since that is when Paseo learns a turn is in flight. New coverage: a stop
before Claude starts the message withdraws it without interrupting, a
stop while only helpers run never reaches Claude, a stop mid-turn
interrupts once and drops the aborted result, a new turn clears a flag
no result will consume, the SDK option is passed, and a cancelled steer
is not withdrawn again.
docs/providers.md described the old unconditional interrupt; it now states
when the Claude adapter interrupts and why it declares the affordance.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(claude): keep a helper spawned in the background out of a turn's cancel
When a turn is canceled, Paseo terminalizes the subagents that were
running in its foreground and skips the backgrounded ones, which outlive
it. It only learned a child was backgrounded from task_updated's
patch.is_backgrounded, which Claude sends when a foreground task moves to
the background later. A child spawned in the background carries the flag
on task_started instead, and a resumed subagent always does, so the
track marked those helpers "canceled" while they kept working.
Read is_backgrounded from task_started as well. docs/agent-lifecycle.md
said task_updated was the only signal; it now names both.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(app): drain queued messages as steers
The composer queue drains when an agent goes idle. It sent each message
with no activeTurnBehavior, which the daemon treats as "interrupt": if
the agent had just started a turn of its own (a Claude wake-up after
background work finished), the drained message stopped that turn and
replaced it.
Send drained messages with activeTurnBehavior "steer". A drain that races
such a turn folds into it; with no turn running the daemon starts a
normal one, as before. "steer" is an existing protocol value, so older
daemons need nothing new.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test(claude): cover background helpers against real Claude
Every other test in this file tells Claude "Do not use a background
task", so a message sent or a Stop pressed while a backgrounded subagent
runs was never exercised. Four real-Claude scenarios cover it:
- "a background subagent keeps running after a message is sent while it
works": one helper, the main turn ends, the parent is not reported
running, a message in that window is answered, and the helper is never
read as canceled and completes.
- "a helper's own nested background job does not phantom-run the parent
or kill the helper": a helper starts its own background Bash job, then
keeps working in the foreground. The parent stays idle across the
nested job's completion (sampled for 30 s around a 5 s job), a message
sent in that window is answered, and the helper reports CHILD_DONE.
- "the app's default steer survives a backgrounded helper while the main
session works": the app's default send behavior (steer) does not kill
a helper when it interrupts the main session's own turn.
- "Stop cancels the main session's own turn but spares its background
helper": the same for Stop.
The wake-up turn Claude starts after a helper finishes lands a moment
after the helper reads completed (0.1 to 5 s in recorded traces), and the
daemon announces it as a pushed "running" agent_update, not as a
turn_started stream event. The checks poll for it and read the final idle
once the wake-up settles. waitForFinish can return a stale "idle" right
after a send, and the push stream can miss a message's final text when it
resolves, so the tests read fetchAgentTimeline like the rest of the file.
Helper status comes from agent.provider_subagents.list, since agent_stream
carries no provider_subagent events.
The scenarios need the previous commits together. Run against the frame
heuristic this branch removes, the nested-job test fails: the parent
flips to running once the nested job finishes and stays there.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(claude): interrupt a stopped prompt Claude has already taken
Stop before Claude announces a prompt started withdraws it with
cancel_async_message instead of interrupting. Claude can dequeue a
message for its next turn before announcing it, and withdrawing it is
then a no-op that reports cancelled: false, so the stopped prompt kept
running. Interrupt whenever a message could not be withdrawn.
Co-authored-by: Isak Persson <isak.a.persson@gmail.com>
* fix(claude): withdraw a stopped prompt only once
When the withdrawal of a stopped prompt failed, the interrupt path
withdrew a second time, which could cancel a prompt sent while the first
withdrawal was in flight. Withdraw once and interrupt on failure.
The real-Claude steer scenario now waits for the background helper to
exist before steering, so it proves the helper survives the steer.
Co-authored-by: Isak Persson <isak.a.persson@gmail.com>
---------
Co-authored-by: Isak Persson <isak.a.persson@gmail.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(plugins): restrict overview content to safe Markdown and HTTPS destinations
Registry overviews are author-controlled content read away from their repositories. Apply a shared presentation policy instead of the documentation renderer so relative links, non-HTTPS URLs, raw HTML, and author styles never become active overview content.
* fix(plugins): choose the first allowed screenshot for previews
* fix(app): show PR panel state, activity verbs, and thread badges in the app language
The PR panel's state label, the verb after each activity author, and the
Resolved and Outdated thread badges were English literals, so they stayed
English in every other language. The state and verb translations already
existed and were unused; the badges get new keys in every locale.
Closes#6214
* fix(app): translate every PR panel verb and state, and phrase them after the author
The zh-CN state and review labels still held English ("Open", "Draft",
"已 review"), and the Spanish and Portuguese verbs were adjectives that
do not read after an author's name. The translation test now checks
each verb and state on its own, so one English label fails it.
Co-authored-by: Dillon <dillonzq@outlook.com>
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Dillon <dillonzq@outlook.com>
* feat(plugins): enable registry installs by default
* fix(plugins): explain explicit Git source for missing registry IDs
* fix(plugins): tailor missing registry guidance to installation
* test(app): sample footer geometry in one browser frame
* fix(plugins): make registry source resolution deterministic
Directory existence must not change the meaning of a registry identifier shared by CLI, RPC, and UI installs. Classify explicit host paths in the source parser and explain recovery from registry transport and revision-selection failures.
* fix(plugins): recognize dot and Windows directory sources
* test(cli): wait for readable daemon ownership before cancellation
The cancellation test raced PID-file creation against its first write and parsed empty JSON in CI. Use the public daemon-instance reader to wait for and capture the supervisor PID before asserting cancellation cleanup.
* test(app): expect missing directory install guidance
* Open Usage as a modal over the current screen
The sidebar Usage icon and summary now open Usage in the adaptive modal
(a dialog on wide layouts, a bottom sheet on compact ones) instead of
navigating to the /usage screen, so closing it leaves you where you were.
The /usage route and UsageScreen are removed.
* Reset Usage modal cards when the host changes; keep scroll test body linear
* fix(app): preserve timeline rows across virtual layout changes
Keep recent and virtual history rows under one keyed host so entering
virtual history preserves their DOM and loaded image state. Initialize the
virtual range from the mounted viewport to avoid an empty handoff commit.
Measure committed row dimensions before accepting reading geometry,
including child layout changes flushed during resize observation. Separate
row content from position updates while retaining live revision signals.
Add a regression covering row identity through virtualization and the
position of content following a growing row. Keep the existing browser
assertions unchanged.
* test(app): verify child state survives timeline virtualization
* fix(app): show PR panel activity and add-to-chat labels in the app language
The PR panel's Activity header, its empty state, every "Add to chat"
button, "Add all to chat", "Adding...", and the comment and thread menu
labels were written in the components, so they stayed English in every
other language. The sidebar status group rows also appended an English
" group" to their accessibility label. They now come from the
translation files, with copy for all nine supported languages.
Closes#6203
* test(app): check the new PR panel labels are translated in every language
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(app): translate the PR checks summary, sidebar status labels, and "Worked for" line
These were written in the components, so they stayed English in every
other language. They now come from the translation files, with copy for
all nine supported languages.
Closes#6190
* fix(app): rebuild translated labels when the app language changes
The checks summary, sidebar status groups, and turn footer now take the
translation function, so their memoized labels rebuild on a language
switch. Russian one-check group labels use singular wording.
* test(app): compare compact footer targets at sub-pixel precision
The browser reports the 44pt footer buttons as 43.99999 wide on some
runs, which fails the finger-size check on main as well.
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(app): leave out an image when a copied selection starts below it
A drag that starts in the gap below an image in an assistant reply
anchored at the start of the image's rendered internals, ahead of the
hidden img, so the copy included the image's Markdown although only the
text after it was highlighted. A selection that starts inside an image
and ends past it now starts after the image.
Closes#6187
* fix(app): keep a failed image when a copy starts in its error text
A failed image shows its error message as text inside the image frame. A
selection that starts in that text selected part of the image, so it keeps
the image's Markdown. Only a selection that starts inside the frame before
any of its text starts after the image.
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Clicking Refresh on a usage card in the context window details closed the
card within the close grace, with the pointer still on it. Refresh disables
itself while the request is pending, which drops the click's focus to the
document, and the hover card treated any focus leaving the card as a reason
to close.
Only keyboard focus keeps a hover card open, so only losing keyboard focus
now closes it. Whether focus is keyboard focus is read when it arrives,
because the element losing focus no longer matches :focus-visible during
focusout.
Closes#6176
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
paseo-bot[bot]