* Use current Hub permissions when reconnecting a daemon
* Back off when the Hub repeatedly rejects session permissions
* Show the permission recovery commands in Hub status
(cherry picked from commit 0e67045c4b)
* Keep a finished reply visible when Claude starts a turn after a background command
When a background command finishes while Claude is idle, Claude Code sends the
task notification first and opens its own turn after it. The notification row
was stored with no turn id. A row with no turn id joins the turn before it to
the turn after it, so under Collapse turns the app showed both turns as one and
folded the finished reply behind the summary row.
The Claude provider now keeps a task notification that arrives between turns
and emits it inside the turn Claude starts for it, so the row carries that
turn's id. A notification no turn follows is emitted as before: when Claude
reports idle (a command the user stopped), before the user's next message, and
on close.
Rows already stored without a turn id are not changed.
Refs #10030.
* Show a stopped background command's notification at once
A stopped background command starts no Claude turn. While a background
subagent is alive Claude also sends no idle, so the held row stayed hidden
until the subagent finished and then took the subagent's wake-up turn id.
Only task notifications Claude reacts to (completed or failed) wait for a
turn now. A stopped one is emitted right away, as before.
Refs #10030.
* Keep a row that belongs to no turn out of the turns around it
A turn starts with the user's message or the provider's own turn start
(Claude's init). A background command's notification that arrives
between turns starts nothing, so the daemon writes it with no turn ID.
The shared turn rule read a missing ID as legacy history and let the
row continue both neighbours. The reply before it and the turn after
it then read as one turn, and Collapse turns folded the first reply.
Now a row with no turn ID never continues a row that has one, so it
cannot join two turns. History with no IDs at all keeps the trigger
rule. The app's fold and the daemon's whole-turn paging share the rule.
This replaces the earlier Claude provider hold, which had to guess from
the notification's status whether Claude would start a turn for it.
Refs #10030
* Give a background task's notification the turn it starts
A background task's notification that arrives while no turn runs starts
a turn. Claude's next run belongs to it unless Claude announces another
start first: a later notification starts the next turn, and a queued
command (the user's message, a scheduled wake-up) starts its own. When
Claude does not respond, the notification is a turn in itself.
So a reply Claude writes for a finished background command shares the
notification's turn, and the reply before it stays in its own turn and
stays visible under Collapse turns. A stopped command's row shows at
once as its own turn, and the agent stays idle.
Claude sends no frame for a Monitor event, so a run it starts for one
belongs to the last notification before it.
This replaces the shared turn-rule change, which drew the notification
as a turn between the two.
Refs #10030
* Keep a subagent's own job notification out of the main session's turns
A background job a subagent started reports to that subagent, and
Claude starts no main-session run for it. Its notification gave the
next main-session run a new turn, so when it arrived between a main
command's notification and Claude's reply to it, the reply was drawn
in a different turn from the notification it answered.
Refs #10030
(cherry picked from commit 32bdb9e7af)
When OMP changed the thinking level itself, for example on `/switch @role`
with a role that carries a level, Paseo dropped the `thinking_level_changed`
event because the OMP runtime schema did not list it. The thinking selector
and `runtimeInfo.thinkingOptionId` kept the old level while OMP ran the new
one.
Parse the event and update the session's level, preferring OMP's
`configured` selector (for example "auto") over the effective level, the
same way the Pi provider handles it (#6741).
Closes#8249
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
(cherry picked from commit a3c41ecc1d)
An in-app update and the startup version check again use the desktopManaged flag in paseo.pid, while Manage built-in daemon is on. A daemon that an earlier Desktop session launched is now replaced on update. A daemon that the user started through the CLI is still never stopped automatically.
Closes#6385.
(cherry picked from commit 59ca1f5f77)
List ignored directories as single entries and only search inside them
when a .worktreeinclude pattern matches the directory or names a path
inside it. A bare .env no longer selects node_modules/pkg/.env.
Verified with real Git and isolated daemon workspace creation on Linux.
With 300,000 ignored files, Git trace2 recorded over 6,000 directory
visits before the fix and fewer than five after. The workspace copied
the root .env. The lifecycle test also verified rollback when copying
fails.
Closes#10004
Co-authored-by: mgranderath <22001111+mgranderath@users.noreply.github.com>
(cherry picked from commit 871c714b46)
* fix(workspaces): preserve commits when archiving worktrees
A removed worktree can be the only reference to its HEAD. Keep that commit before deletion and record the checkout so detached workspaces restore at the same commit.
Co-authored-by: Okan YILDIRIM <63511519+ragokan@users.noreply.github.com>
* fix(workspaces): restore saved checkouts through one git path
Keep existing archive and deletion behavior when Git cannot read a worktree HEAD, and record every readable checkout for a common restore path.
* fix(workspaces): preserve git HEAD identity through restore
Record Git’s full HEAD ref or commit so restore can pass branch and detached checkouts to Git without guessing from a short name.
* fix(workspaces): scope saved HEAD to each archive
An archive with no readable checkout must not retain the HEAD saved by an earlier archive.
* test(workspaces): verify detached checkout after restore
Update the browser journey for detached restore and allow the daemon restart test enough time on Windows runners.
---------
Co-authored-by: Okan YILDIRIM <63511519+ragokan@users.noreply.github.com>
(cherry picked from commit 59d6018dac)
* Reload the desktop window when its renderer exits
When the renderer process of the main window exits without a clean
shutdown (killed by an out-of-memory killer, crashed, or out of memory),
Electron keeps the window frame but shows nothing until the user presses
Ctrl+R. The main process had no handler for render-process-gone, so it
neither recovered nor logged the exit.
The window now reloads itself and logs the exit reason to main.log. If
the renderer exits three times within a minute, it stops reloading and
logs that, so a page that crashes on load does not restart in a loop.
Fixes#9999
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Leave a renderer exit before the first page commits to Chromium
Before the window's first navigation commits, Chromium finishes it in a
new renderer by itself. Reloading at that point aborted the initial
loadURL with ERR_ABORTED, which fails desktop startup. Skip the reload
while the window has no committed URL.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit b24eab785b)
* Let a steer wait out a Claude compaction instead of cancelling it
While Claude compacted, the Claude session answered every steer with
"unavailable". The agent manager reads that as "replace the turn", so
it interrupted the turn and Claude Code printed "Compaction canceled.".
Claude Code already holds a `next`-priority message until compaction
finishes, so the session now accepts the steer during compaction.
Co-authored-by: NightMachinery <36224762+NightMachinery@users.noreply.github.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Steer only while the compaction is still running in the real Claude test
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: NightMachinery <36224762+NightMachinery@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 1821ac3f93)
The MCP agent tools serialized live agents without the background work
summary that the WebSocket snapshot carries, so a parent agent saw an
idle worker with a running background shell as plain idle. Both tools now
add the same summary the session adds, omitted when the agent has none.
Co-authored-by: Orcun Simsek <1612171+ortschun@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit a1088a787d)
* Create the workspace in the caller's directory when --path is relative
`paseo workspace create --path .` adopted the daemon's working directory
instead of the directory the command ran in. The CLI sent the relative
path unchanged, and the daemon resolved it against its own cwd. The CLI
now resolves a relative --path against the caller's directory before it
sends the request, as it already does when --path is omitted. Absolute
POSIX and Windows paths pass through unchanged so a remote daemon on
another operating system still gets the path as written.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Leave a home-relative --path for the daemon to expand
Resolving every non-absolute --path against the caller's directory turned
`--path=~/repo` (and a quoted "~/repo") into `<cwd>/~/repo`. Before this
change the daemon expanded the leading ~ itself. `~` and `~/…` now pass
through unchanged again.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Send a relative --path to an explicit daemon endpoint as written
The CLI already resolves caller paths only for a local daemon and sends
them unchanged to an explicit --host endpoint (resolveRepositoryDirectory,
project create). workspace create, and paseo run, which builds its
workspace source through it, now follow the same rule instead of turning
a relative path into a path on the caller's machine for a remote daemon.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 2c2385f362)
A force-cancel finalizes turn X in the daemon while the provider keeps
running it, so the next run can be bound to the same X. handleStreamEvent
dropped every terminal for a finalized turn id, including the synthetic
turn_canceled that cancelAgentRunNow dispatches for the tracked run and
the provider's own terminal. The run never settled, cancelAgentRunNow
waited on it forever, and every later cancel, reload and send queued
behind it until the daemon restarted.
A terminal for the tracked run's own turn id is no longer dropped.
Fixes#6355
Supersedes #8405
Co-authored-by: Dominic Anaeto <dominic.anaeto@gmail.com>
(cherry picked from commit 7f0fe57c48)
Tapping Load more emptied the list until the larger page arrived: the sheet
showed "Loading sessions...", unmounted the virtualized list, and brought it
back scrolled to the top. The sessions queries set keepPreviousData, but
useQueries gives a changed key a new observer with no previous data, so the
placeholder was always empty. The placeholder is now the cached page that
Load more grew from, for the same scope, search, and provider.
The compact import-session journey flaked on this in CI: its scroll helper
resolved a row, the list was replaced before the helper ran, and the walk
up from the detached row found no scroller. The helper now scrolls the list
itself, and the journey checks that Load more keeps the reader down the
list.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 94f2df6263)
On web and desktop, Download over a direct connection now fetches the file and saves it from a blob instead of navigating the window to the host URL. A rejected or unreachable download shows an error in the download toast and the app stays on the workspace. The daemon's HTTP CORS check uses the same origin rule as the WebSocket upgrade.
Closes#9895
Co-authored-by: panta82 <4291324+panta82@users.noreply.github.com>
(cherry picked from commit 02124454bc)
The host registry was parsed with a strict array schema, so one host or
connection this build couldn't read failed the whole array. The loader then
removed @paseo:daemon-registry, and every saved host and password with it.
That happens after a downgrade, or once a newer build saves a connection kind
this build doesn't know.
- Hosts and connections are read one at a time. One this build can't read
is skipped, and the other hosts load.
- Skipped entries are written back on every save, so the build that wrote
them still finds them. Removing a host also drops what this build couldn't
read for it.
- A value that isn't a JSON array is copied to
@paseo:daemon-registry:unreadable instead of being removed.
- The stored-password migration still moves a connection password to its
host when that host also has an unreadable connection.
(cherry picked from commit a8cadae0ee)
* fix(server): abort timed-out watcher setups and back off their retries
A watcher subscription that missed WORKSPACE_GIT_WATCHER_SUBSCRIBE_TIMEOUT_MS
was abandoned but its startup scan kept walking the tree to completion, and
recovery retried every five minutes forever, repeating the full scan each
time. On a CPU-starved host that kept dozens of directories rescanning.
- FileObserver.subscribe accepts an abort signal that stops the startup scan
and rejects; WorkspaceGit aborts setups at the deadline and on dispose.
- Consecutive setup timeouts extend the recovery delay past the 5-minute
ladder cap, doubling up to one hour; an accepted subscription resets it.
Degraded polling keeps its 5-60 s cadence meanwhile.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(file-observer): keep an aborted observation's backend from starting
Aborting a subscription closes its observation, which can happen while
start() is still reading the root. The backend then started after its close
had finished, and the native backend opened a root watcher that nothing
would close.
Co-authored-by: Spencer Shulem <1703265+Spshulem@users.noreply.github.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test(file-observer): abort a subscription mid-scan in the abort test
The test aborted before the backend started, so it passed with the abort
listener removed. A tree past the Linux directory limit makes a scan that
keeps walking fail on the limit instead of rejecting with the reason.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Spencer Shulem <spshulem@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>
Co-authored-by: Spencer Shulem <1703265+Spshulem@users.noreply.github.com>
(cherry picked from commit 33259746db)
A PASEO_HOME on AFS passes the socket directory checks, then refuses the
Unix socket bind with EPERM, and the terminal worker failed to start. The
worker now binds each eligible candidate in turn and moves on when the
filesystem refuses the socket (EPERM, EACCES, EOPNOTSUPP, ENOTSUP). Other
listen errors still fail the start.
Fixes#9213
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit ade3817a45)
In Bypass, Claude agents again show a permission card for asks from the user's permissions.ask rules (including managed policy), PreToolUse hooks that return ask, compound and piped commands that hit an ask rule, permissions.blockReadsOutsideWorkingDirectories, an organization's ask limit, and sandbox overrides. Claude's own checks are still approved without a card. Upgrades @anthropic-ai/claude-agent-sdk to 0.3.326, which passes the ask's source to canUseTool.
Closes#9916
(cherry picked from commit 3eedffc4ac)
* feat(desktop): size controls by the primary pointer, not any touch screen
The desktop app gave touch-sized controls to every machine with a touch
screen, including Windows laptops driven by a trackpad. In Electron, ask
whether a mouse-like primary pointer drives the UI instead: Chromium
reports a fine primary pointer for those laptops and a coarse one for a
tablet with no mouse or trackpad attached.
Browsers and native keep the touch-screen check, so an iPad keeps touch
sizes with a trackpad attached. Swipe gestures still follow the
touch-screen capability.
Refs #9892
* feat(app): size controls by the primary pointer on every platform
Replace the Electron-only check with one rule everywhere: touch-sized
when touch is the primary input, `(pointer: coarse)`, rather than when a
touch screen exists at all. WebKit keeps touch primary on an iPad with a
trackpad attached, so iPad Safari stays touch-sized; Chromium reports a
fine primary pointer for a touch-screen laptop driven by its trackpad, in
the browser and the desktop app. Native is always touch-sized.
`hasTouchScreen()` and its `maxTouchPoints` fallback had no other caller
and are removed. Swipe gestures still follow the touch-screen capability.
Refs #9892
* test(app): exercise pointer sizing through the media source
* test(app): cover touch laptops with a primary pointer
---------
Co-authored-by: Mohamed Boudra <boudra.moha@gmail.com>
(cherry picked from commit e5352c9890)
With the default send setting, a message sent while an OpenCode agent
waits on a question steers the turn, and Paseo rejects the question in
OpenCode. The OpenCode session dropped the question from its own list
without reporting it resolved, so the agent manager and the app kept
showing it. Answering or dismissing it then failed with "No pending
permission request", and the prompt could not be closed.
The session now reports permission_resolved whenever OpenCode accepts a
reply, rejection, or answer, as the other providers do.
Fixes#9879
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 038a9ca00d)
AdaptiveTextInput wraps its input with withUnistyles, and the pairing link and direct connection modals wrapped AdaptiveTextInput a second time. On web each wrapper is a display: contents element whose style lands on its direct child, so the outer style landed on the inner wrapper and the input lost its border, background and padding.
Drop the outer wrapper, AdaptiveTextInput already themes the placeholder colour. Also give the host password field a placeholder and focus it when pairing asks for a password.
(cherry picked from commit b95f1bbf31)
A Codex session opened to read history reads the thread and disposes its
app-server right away, but canStopRuntime() vetoed it because it had no
client. Idle cleanup then kept every such chat loaded in the daemon.
Closes#9804
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Spencer Shulem <1703265+Spshulem@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 5a78c1cd16)
* Let Claude model discovery finish a token refresh instead of killing it
Model discovery stopped waiting after 5 s by aborting the SDK query and
tree-killing the CLI. A CLI that is still refreshing an expired OAuth
token at that moment has already rotated the refresh token, so the kill
leaves only the revoked token on disk and logs the user out.
The 5 s limit now only bounds the wait. A CLI that has not answered gets
its input ended and exits on its own; it is tree-killed only after a
60 s grace period or when the daemon shuts down.
Fixes#9314
* fix(server): let Claude model discovery finish a token refresh after it answers
Claude Code answers the model list before its startup token refresh has
saved the rotated refresh token. Discovery killed the CLI right after the
answer, so a refresh still in flight left the revoked token on disk and
every later Claude session failed until /login. Discovery now ends the
CLI's input and lets it exit on its own whether or not it answered, with
the existing grace period and shutdown kill as the bound.
Co-authored-by: daysun-k <8309193+daysun-k@users.noreply.github.com>
---------
Co-authored-by: daysun-k <8309193+daysun-k@users.noreply.github.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
(cherry picked from commit 8522d18ed8)
After a daemon restart, an agent's timeline is rebuilt from provider
history, and those rows have no turn id. A reply the agent starts on its
own, such as Claude waking after a background subagent, follows the
previous reply with no user message between them, so the last assistant
message lookup joined the two and wait results and finish notifications
returned the old reply with the new one appended.
The timeline projection already treats a message id change as a new
message. Share that rule with the lookup, so both decide which rows are
chunks of one message the same way.
Closes#9814
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit be92e85a4e)
OMP model labels were built as `${provider}/${name}` and then cut at the
last `/`, so a free-text name such as "... ~60-70 tok/s on 0.2.10, 52/56)"
showed as "56)". Underscores were rewritten to spaces ("IQ3_S" -> "IQ3 S"),
and a colon inside the name cut everything before it.
Pi had the same bug, fixed in #9823. Move that label rule into a shared
model-label module and use it for both providers: free-text names are
kept, path-like names are shortened to their last segment, and a
"Vendor: " prefix is dropped only when it is a prefix.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Rahul Sharma <19687913+rms523@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit a302e65c85)
The private index snapshot behind index and All changes comparisons was a
fresh copy, newer than every entry, so Git skipped its same-timestamp
content check and a same-size save made right after staging read as
unchanged. The copy now keeps the source index's mtime, rounded down to
the second.
Closes#9827
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
(cherry picked from commit d49991318f)
* fix(pi): keep slashes and underscores in model display names
Pi model labels were built as `${provider}/${name}` and then cut at the
last `/`, so a free-text name such as "... ~60-70 tok/s, 52/56)" was shown
as "56)". Underscores were also rewritten to spaces ("IQ3_S" -> "IQ3 S").
The compact model trigger in the app repeated the last-`/` cut.
Build the Pi label from the model's own `name`/`id`: shorten only
path-like values with no whitespace (OpenRouter reports names like
"google/gemini-2.5-flash-lite"), and keep free-text names intact. The
compact trigger now strips only a leading provider namespace.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test(pi): cover model labels through the catalog and drop stale ACP copies
acp-agent.test.ts still imported transformPiModels, which the Pi label
fix removed, so its modelTransformer test and a stale copy of the old
Pi label tests failed. The modelTransformer test now supplies its own
transformer, and the Pi label cases run through fetchCatalog with the
reported model names instead of an exported helper.
Co-authored-by: Rahul Sharma <19687913+rms523@users.noreply.github.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(pi): keep colons inside free-text model names
The vendor-prefix strip cut a free-text Pi name at its first ": ", so
"Qwen3 IQ3_S (tail skip: on, 52/56)" showed as "on, 52/56)". Strip only
when the text before ": " has no parenthesis or comma, which keeps
OpenRouter's "Vendor: Model" names shortened.
Co-authored-by: Rahul Sharma <19687913+rms523@users.noreply.github.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Rahul Sharma <19687913+rms523@users.noreply.github.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>
(cherry picked from commit d15190803a)
Fullscreen controls retained a fixed toolbar overhang and aligned to the Send arrow inside its disc. Align their visible glyph edges with the text column independently of the toolbar geometry.
* fix(app): preserve submit scroll through existing momentum
A prompt sent during an active gesture requested the bottom, but the older gesture cancelled that request when momentum ended away. Honor the still-sticky pending request on release while allowing a newer gesture to take ownership.
* fix(app): keep delayed navigation behind user scroll intent
* fix(app): bound complete inline tool details within groups
Apply the detail height allowance to the complete input/output/error disclosure so expanding a tool cannot bury its collapse header outside the group viewport.
* test(app): verify tool disclosure geometry across painted frames
* test(app): measure disclosure bounds without evidence capture
* fix(app): give capped tool details one vertical scroll owner
* fix(app): allow inline tool details to scroll on iPad
A hover-only Pressable claimed the native responder around the detail scroll view. Keep hover reporting on a passive View so drags reach the capped detail and its later sections remain accessible.
* fix(app): preserve scroll position inside overflowing tool groups
Content-size changes also come from nested disclosures. Follow newly appended calls while preserving the reader’s position when a disclosure changes height.
* test(app): assert inline tool visibility on iPad
Offscreen Output nodes exist before the pan, so selector waits cannot prove reachability. Assert visibility before and after scrolling and verify that collapsing hides the contents before reopening.
* refactor(app): remove unused tool detail hover callbacks
* docs(app): correct inline detail test coverage note
* fix(app): apply grouped detail allowance to reasoning
A bare URL that md4c does not autolink is linked by the text-link
pattern. That pattern stopped only at ASCII whitespace, quotes and
sentence punctuation, so "https://a.com、https://b.com!" became one
link, and "看这里:https://example.com/foo,还有" linked ",还有" too.
Opening the link went to a URL ending in the encoded punctuation.
CJK, fullwidth and typographic punctuation (、。,()【】「」“”…) now
ends the URL. CJK letters in paths still stay in the link.
Refs #9743
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
A turn the provider starts on its own, such as Claude waking after a
background subagent, has no user message before its reply. The last
assistant message lookup joined every adjacent assistant row, so it
returned the previous turn's reply with the new one appended, and the
220-character notification preview showed only the old reply.
Stop the backward scan at the turn boundary so the lookup returns only
the newest turn's reply, while chunks of one turn are still joined.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
The daemon drops a dictation stream when its STT session errors and answers
every later chunk with dictation_stream_error. The app only listened for
acks, so finishing waited 30s and reported a generic timeout. The sender now
fails finish with the daemon's message, and the daemon logs the failure that
dropped the stream once instead of nothing.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
The daemon-lifecycle cleanup-failure test planted a zombie with
`sleep 0 & echo $!; exec sleep 600`. dash reaps any exited child after
a builtin, so when `sleep 0` exited before `echo` returned, the shell
reaped it and `ps -o stat= -p <pid>` failed. The child now reads fd 3
until the helper closes it, which it does only after the shell has
become `sleep`.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
* fix(cli): fail agent wait when the daemon resolved no agent
The daemon answers wait_for_finish with status "error" both when it could not
resolve the identifier to an agent (final: null) and when the agent it did
resolve finished in error (final: the agent payload). runWaitCommand handed
both to buildWaitResult, which returns a normal result row, so
`paseo agent wait <unknown id>` printed "Agent not found: ..." and exited 0
while every sibling command exits 1 with AGENT_NOT_FOUND (#9736).
Throw AGENT_NOT_FOUND when the state is error and the payload is null, and let
that command error through the inner catch instead of rewrapping it as
WAIT_FAILED. A genuine agent failure still carries its payload, so it keeps
printing its result row. The wire schema is unchanged.
* test(cli): cover the agent wait not-found mapping without module mocks
Move the not-found decision into buildWaitResult, which maps the daemon's
wait reply to the command result, and test that function directly with plain
reply objects instead of mocking the daemon client module. The e2e test in
tests/11-agent-wait.test.ts still covers the command against a real daemon.
Co-authored-by: Eapp1e <163811520+Eapp1e@users.noreply.github.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(cli): keep daemon errors during agent wait reported as WAIT_FAILED
Map the wait reply to the result after the wait's try/catch, so the
AGENT_NOT_FOUND error leaves without the inner catch passing every coded
error through. Connection and RPC errors from waitForFinish are wrapped as
WAIT_FAILED again, as on main.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Eapp1e <163811520+Eapp1e@users.noreply.github.com>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(app): show archive workspace in command center
* test(app): cover archive workspace command center entry
* fix(app): expose archive for non-git workspaces
* fix(app): archive non-Git workspaces from the command center
A plain directory workspace had no Archive workspace entry in the command
center, because the entry came only from the Git action list, which is
empty outside Git. The command center now lists archive for any workspace
that is not Git and runs the workspace's own archive controller, the same
one the Git action uses, so it works whether or not the workspace's
sidebar row is rendered.
Git workspaces are unchanged: archive stays a Git action that shows on
the default list only when it is the promoted primary action.
Co-authored-by: Andreas Heiberg <andreas@heiberg.io>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Andreas Heiberg <andreas@heiberg.io>
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
The arrow-choice test waited for /tdd to leave the list as proof that
the search had become "h". The option list is a FlatList that mounts
only the rows near its scroll position: right after the arrows wrap to
/exit at the top, /tdd at the bottom is unmounted while the search is
still "". The wait then passed early, Backspace landed before the
cursor reached the search, the /exit choice stood, and Enter archived
the agent.
The test now waits for a popover that shows /help with neither /tdd
nor /exit, in one query. With the search at "" the list always mounts
the bottom row or the chosen one; only the search "h" leaves both out.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
On web and desktop, dropping a Usage card made it jump above its new slot
and slide back down for 200 ms.
The web list cleared its drag state on drop and rendered the owner's data
again. Usage saves the order through app settings, so the new order
arrived in a later render. The render in between had the old order with
dnd-kit's transforms cleared, which started a transition back to the old
slot. The new order then moved the row's slot under that running
transition.
The list now holds the dropped key order until the owner's data changes
order, and renders the owner's current rows in that order. The drop and
the new order commit in one render, so dnd-kit skips the transition. A
drag that starts during the hold picks up the rows as displayed. Sidebar
lists already reorder in the drop's render and behave as before. The
native list has followed the same rule since #9748.
The shared icon button chrome set outlineWidth: 0, which overrides the
:focus-visible ring from public/index.html. Header and toolbar icon
buttons (sidebar toggle, workspace actions, Explorer toggle, new tab,
pane menu, diff toolbar) showed no indication of keyboard focus.
Drop the reset and draw the ring 2px inside the frame: the header title
row clips overflow, so an outset ring loses its top and bottom edges.
Pointer focus stays unringed through the existing *:focus rule.
Publish the resolved replacement selection to the composer. iOS does not emit a native selection event for this edit, leaving autocomplete at the slash until the next keystroke.
* feat(app): browse, install and manage plugins from a built-in gallery
Settings → Host → Plugins becomes a gallery with two tabs. Installed lists
the host's plugins with on/off switches and pending updates. Browse lists the
plugin directory with featured cards, categories, sorting and search. Each
plugin has its own page with its screenshots, overview, details, install or
on/off controls, and its own settings screens inline.
Every install asks first, in a sheet naming the source and version, because
installing runs the plugin's code on the host. Uninstalling also asks first;
its confirm says what is deleted.
The daemon fetches the directory from its configured registry
(plugin.registry.list/get, gated by server_info.features.pluginRegistry). The
app never fetches from the registry itself, so a host with a private registry
shows that registry.
README overviews render with the registry's overview contract (the same one
the website uses): only https links and images work, and nothing is read as a
workspace file. Before this, a word like "Node.js" became a broken relative
link.
* fix(app): refresh plugin failures and leave Settings cleanly from a plugin page
The Installed row read the app-side evaluation error without subscribing to the
registry, so a plugin fixed by Reload kept its "failed" badge. Rows and the
plugin page now subscribe through usePluginEvaluationError.
Leaving Settings with no remembered workspace replaced only the top stack entry;
with a plugin page pushed on top, the Plugins list stayed mounted underneath and
reopening Settings stacked a second copy. Dismiss the Settings entries first.
* fix(app): keep Browse working as the registry grows, and test it end to end
Registry listings now reach apps through an open wire schema
(plugin.registry.list/get carry PluginListing*), and the daemon drops a listing
it cannot read instead of failing the whole index. A registry that adds an
artifact kind or bumps schemaVersion no longer breaks Browse on older apps.
The gallery picks its opening tab once, so installing the first plugin keeps
the user on Browse. README links open only https: the parser still autolinks
bare http, www and email text, so a press checks the URL again.
Adds a Browse e2e test against a local registry and Git repository: install
from a row through the confirm sheet, stay on Browse, open the listing page,
uninstall from it.
* test(server): give the oversized-sibling diff refresh the same time budget as its siblings
expect.poll defaults to one second, which a Windows git diff refresh can exceed.
* test(server): wait as long as the other observer tests for native watcher events
expect.poll defaults to one second; macOS FSEvents can deliver later than that on CI.
* fix(app): keep a listed plugin's page usable after uninstalling it, and fetch the directory less
An installed plugin's page finds its listing through the plugin, so uninstalling
a directory plugin from the Installed tab left "not installed or listed" with no
way back. The page now becomes the listing's page, offering Install again.
Updates are re-checked when the host's plugins change and after a failed update,
whose proposal named a revision that no longer exists. The directory, listing
and update queries refetch on mount only once stale, so opening a plugin's page
no longer refetches the whole directory and every update check.
* test(app): install the plugin before opening Plugins in the busy-queue reload test
Plugins opens on Browse when the host has no plugins and stays there after the first install, so a plugin installed through the daemon after the page opened never showed on Installed.
* test(website): wait for hydration before scrolling to the hero in a short window
The router resets the scroll position when the page hydrates. Scrolling to the mockup before that left it below the window, so its start line never came on screen and the opening scene never played. It failed 7 of 10 runs on webkit.
* fix(server): show the registry setup error on Browse instead of asking to retry
The registry's request headers were built inside the try around fetch, so a setup error such as credentials over plain HTTP came back as "Could not reach plugin registry. Retry later." Install already built them first.
* fix(app): show plugin errors without the request details meant for logs
A daemon RPC error's message ends with requestType=… code=…; the daemon's own text is in userMessage. Gallery alerts and toasts now show that text.
* fix(app): show why an update failed on the plugin's page
The update operation's failure was recorded under the plugin ID, but the page only read it to show the spinner, so a failed update looked as if nothing happened. The page now shows the error in its own alert, without the plugin's Failed badge, because the installed version keeps running. While an update is still available the button reads Retry.
* fix(app): show a plugin page's update error only while the update is offered
The failed update stayed in the operations store after the plugin was updated elsewhere or uninstalled, so the page kept saying the update failed. The page now shows it with the Update button it belongs to, as the Installed tab's Updates row does, and uninstalling forgets it.
On Android, a slow drag on a Usage card handle moved the sheet instead of the
card. A drop in the same slot left the sheet ignoring body swipes. A drop in a
new slot painted the original order for one or two frames.
The sheet gesture now registers a content drag as both a failure requirement
and a simultaneous relation, and the Gorhom patch spreads relation arrays into
RNGH, so the waiting sheet recognizer no longer cancels the card drag.
The old-order frames came from the list clearing its drag state before the
reordered data arrived. The sheet portal forwards changed children from an
effect, so the list committed old data with fresh transform hosts and a cleared
active key, and the new data about 200 ms later. The draggable-flatlist patch
now renews moved hosts and clears the active key in the render that receives
the reordered data. A same-slot drop resets explicitly because its data does
not change. Every native list that uses the patch gets the drop change.
Repro tooling for both device checks is in packages/app/e2e/mobile/usage-drag.
Approve Claude execution permission callbacks before creating or emitting permission events, preventing approval cards and transient waiting states in Bypass. Keep AskUserQuestion and ExitPlanMode interactive. Add regression coverage for silent approval and live mode changes.
timeline.append accepted any string as the row kind, but the app registers
renderer kinds with /^[a-z][a-z0-9-]*$/. A plugin that appended a kind such as
repro.card got a successful append, then its client setup threw in
addTimelineRenderer and the whole plugin client failed to load.
The daemon now rejects an append whose kind breaks that rule and names the rule
in the error. The wire schema stays a plain string. Document the rule in the
plugin docs.
Fixes#7326
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>