* feat(claude): add Mythos 5.1 with its model capabilities
* feat(claude): gate Mythos on advertised model identity
* fix(claude): retain capabilities for explicitly configured Mythos
* test(usage): load SQLite before timed store tests
The first Windows OMP store case exceeded its five-second budget while subsequent equivalent cases completed under a second. Load the builtin during test collection so its cold initialization does not consume the first test deadline.
* test(workspaces): publish complete service environment fixtures
Waiting for file existence races with writeFileSync creating the file before writing its contents. Rename completed JSON snapshots into place so the existing readiness check cannot expose an empty file.
* 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(providers): launch plugin CLIs on Windows and tolerate Muse startup
Plugin providers bypassed the core Windows launcher and the plugin host exposed an empty server SDK module. Share the launcher through the SDK and bind it in the host. Give Muse initialization its own deadline because native startup traverses session history before answering MSP.
* fix(providers): terminate Windows launcher process trees
A command-script launcher leaves the CLI behind cmd.exe. Expose forced process-tree termination in the server SDK, use it for Muse and SDK ACP teardown, and bound Muse closing so inherited streams cannot leave callers waiting forever.
* test(muse): keep platform setup outside the teardown assertion
* refactor(plugin): expose process capabilities without quoting helpers
* fix(providers): close process trees on uncorrelated protocol errors
* fix(usage): derive window names and unify pin selection
Quota slots do not identify a period. Derive numeric windows and IDs from reported duration and quota scope; keep unknown lengths neutral. Resolve source defaults and explicit pins through one model, with a persisted empty selection distinct from defaults.
* fix(plugins): declare the duration helper host requirement
* fix(usage): verify persisted selections at the storage boundary
* fix(usage): keep existing logins visible and explain unavailable quota
Credential discovery must preserve accounts whose login exists even when it has expired. Keep vendor failures on the discovered account and isolate discovery failures without inventing provider cards.
* test(usage): align credential route tests with account discovery
* Revert "Remove the unreleased Antigravity provider plugin (#5781)"
This reverts commit 9d9d9af993.
* fix(providers): commit plugin notices emitted during session opening
* feat(providers): restore Antigravity with full access and session warnings
* fix(providers): warn once when opening Antigravity sessions
Driver respawn retains the live transcript and unchanged permission posture. Emit the warning only when opening a session, including resume after a daemon restart.
* refactor(providers): remove the unreleased Antigravity plugin
Remove the provider at the owner’s request because headless command approvals are unsupported. Preserve Muse, shared plugin infrastructure, and earlier editor and terminal integrations.
* test(plugins): keep builtin startup coverage generic
* fix(providers): apply configured options at session startup
Treat provider options as one opaque record, with provider-owned validation and per-agent overrides over configured defaults. Keep agent overrides separate from defaults so configuration changes apply on the next launch or resume.
* test(providers): verify provider-owned validation through Hub
* fix(providers): preserve the complete wrapped session interface
Pi 0.99 ships MCP as a built-in extension. Paseo only recognized
pi-mcp-adapter's /mcp command, so a Pi agent without the adapter started
without Paseo's MCP servers. Pi 0.99 also rejects the adapter's
--mcp-config flag, so the servers are now registered with
pi.registerMcpServer from Paseo's per-session extension. Sessions that
load pi-mcp-adapter keep the --mcp-config path.
Co-authored-by: paseo-bot[bot] <266920839+paseo-bot[bot]@users.noreply.github.com>
Co-authored-by: ArietidsZ <69024461+ArietidsZ@users.noreply.github.com>
* test(muse): record real host protocol fixtures
* feat(muse): add built-in MSP provider
Co-authored-by: millerben95 <millerben95@users.noreply.github.com>
* fix(muse): expose provider settings in the app
The daemon registered host settings without a client contribution, leaving sandbox, network, and workspace trust unreachable. Share the definition with a settings screen and expose it through Command Center even when installed plugins are disabled.
* docs(muse): publish setup and provider support
* test: correct runtime fixtures and restart readiness
* fix(muse): hide reminder housekeeping from the timeline
* fix(muse): use the official Muse Code mark
Source: https://muse.ai/images/favicon/app-squiggle.svg
Preserve the official silhouette path verbatim. Remove its gradient definition and use currentColor, scaling the original 100-unit coordinates into the 24-unit viewBox. Black silhouette renders match pixel-for-pixel.
* fix(muse): adapt to current plugin APIs
Use the screen API introduced with sidebar plugin items. Usage reports now come from source account discovery; remove the retired agent usage-reference request and headline marker, and exercise discovery through the existing provider fixtures.
---------
Co-authored-by: millerben95 <millerben95@users.noreply.github.com>
* test(antigravity): record agy stream-json fixtures
* feat(antigravity): add Antigravity provider plugin
* fix(antigravity): address Phase 1 review findings
* test(antigravity): record parity discovery blockers
* feat(antigravity): support image prompts and daemon e2e
* docs(antigravity): document provider setup and limitations
* test(antigravity): respect Windows interrupt semantics
* docs(plugins): avoid duplicate provider override guidance
* fix(antigravity): report fresh denials and correct mode metadata
Native denial results accumulate within a driver process. Track reported occurrences until respawn so later text-only turns do not repeat notices, while new denials of the same action remain visible.
* fix(antigravity): use the official Antigravity mark
* refactor(usage): remove agent-to-usage wiring
* feat(plugins): screens, sidebar header and footer items
Plugins register screens with addScreen and contribute component-based
sidebar items with addSidebarHeaderItem and addSidebarFooterItem. Items get
currentScreen, openScreen and openPopover, and render the SidebarRow and
SidebarButton kit. The footer shares the header's ordering model through the
new sidebarFooterItems preference, and Settings > Sidebar gains a Footer card.
addSurface, openSurface and addSidebarItem stay as aliases with their
settings keys and routes.
* refactor(plugins): share resolvePluginPlatform and keep header remembered-host key
Consolidate the five resolvePlatform copies into plugins/platform.ts.
Header sidebar items keep the <pluginId>/sidebar/<id> remembered-host key
so existing users keep their remembered host; footer items use
<pluginId>/sidebar-footer/<id>.
* feat(usage): global usage summary, pins and used/remaining display
- Device-local usage preferences (displayAs, pinned windows) in app settings.
- Sidebar footer usage summary: pinned windows in pin order, one per account;
opens the expanded view as a popover on desktop and a sheet on compact.
- Pin checkboxes on window rows and a used/remaining toggle shared by the
summary, popover, Usage screen and Settings usage section.
- Usage screen shows one host with a selector when several are connected.
- Extract HostSwitcher and the sidebar popover host for reuse.
* fix(usage): render each usage report as its own card
* feat(usage): row pinning, one title row for usage controls, even sidebar footer
- Window rows are the pin toggle (checkbox role, checked state); pinned rows
get a surface2 highlight, unpinned rows an interactionHighlight hover.
- Usage screen, Settings section, popover and sheet put the title left and
the host selector, used/remaining toggle and Refresh right (UsageControls).
PageLayout gains an actions slot; the menu sheet header a trailing slot.
- Sidebar footer rows share 28px height, spacing[1] gaps and one leading rail.
* fix(usage): keep row hover lighter than the pinned highlight
Hovering an unpinned window row shows the pinned surface2 at half opacity on
a separate layer, so it never reads stronger than a pinned row. Pinned rows
do not react to hover.
* fix(app): defer sidebar styles and preserve layout exports in timeline test
* fix(usage): open the Usage screen on desktop, drop the Usage footer icon, match git panel refresh
- Desktop presses on the usage summary open the Usage screen; only compact opens the sheet.
- The Usage footer icon is gone; with nothing pinned the summary renders a plain "Usage" row
in the footer row style. A stored `usage` footer preference is ignored.
- Refresh controls use the Changes panel's icon-only ToolbarButton with RotateCw.
* fix(sidebar): keep the footer icon row fixed; footer items are rows only
The footer's icon row (Hosts, Import session, Help and support, Settings) is back to main's
fixed implementation: not in Settings, not reorderable, and plugins can't add to it.
- Footer items are full-width rows between Add project and the icon row: the usage summary
and plugin `addSidebarFooterItem` rows. The Footer card in Settings lists only those rows.
- Stored `sidebarFooterItems` keys for hosts, import, help and usage are ignored.
- `SidebarButton` is removed from the plugin UI kit; a footer item renders `SidebarRow`.
* refactor(usage): call the sidebar footer item Usage
One concept, one word: the footer item is the Usage item, with key `usage`.
- Settings > Sidebar > Footer lists it as "Usage"; its preference key is `usage`.
- `usage/summary-item.tsx` becomes `usage/sidebar-item.tsx` (`UsageSidebarItem`); both states
use testID `sidebar-usage`.
- `usage/summary.ts` becomes `usage/pinned.ts` (`resolvePinnedUsage`, `PinnedUsageWindow`):
it resolves the pinned windows the item shows.
* feat(plugins): screen params, several rows per sidebar item, SidebarSeparator
openScreen takes { screenId, params }; params live in the screen route's
query so reload, history and links keep them. currentScreen carries the
params. A sidebar item may render a fragment or array of rows; a row press
anchors the popover to that row. SidebarSeparator joins the plugin UI kit.
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(plugins): any screen param key; trailing presses anchor the row's popover
Screen params go in the query under a param. prefix, so keys like serverId
no longer collide with the route segments Expo Router merges into search
params. A press anywhere inside a SidebarRow, trailing content included,
makes that row the popover anchor.
Co-Authored-By: Claude <noreply@anthropic.com>
* feat(plugins): plugins title their screens
addScreen takes { id, title, Component }. title is a string or a function
of the screen's params; the screen header shows it and follows param
changes. A title function that throws or returns empty falls back to the
screen id. Legacy addSidebarItem screens keep their item's title and icon,
and the addSurface alias titles its screen with the id.
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(usage): require discovery and remove unreleased agent resolver
* fix(plugins): restore legacy sidebar rows, share the item runtime, header hint presses its row
- Legacy addSidebarItem rows are their own group kind again: main's row, registered icon in
Settings, /sidebar/<id> route, host remembered per row.
- New-API items follow the host last used on any of the plugin's screens.
- /sidebar/<id> falls back to a same-id screen so pre-migration links resolve.
- params moves to PluginScreenProps; settings screens no longer receive it.
- Sidebar items, buttons and popovers share one runtime per installation, so no empty first frame.
- A header row's hint slot presses the row; labels wrap as on main.
- Skill and docs corrected.
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(usage): one host resolver, sheet reuses the screen's controls, name the host
- The Usage row and screen default to the same host.
- The compact sheet carries the screen's controls, host selector included.
- The host selector shows with one host, and unavailable messages name the host.
- The row, sheet title and pinned label use the sidebar.footer.usage translation.
- Window rows read their percent and reset after the pin label.
- Drop the unread per-report cache writes and headline fixture fields.
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(plugins): popovers, header buttons and composer pills own their runtime again
Only sidebar item rows use the installation's shared runtime. Popover content
disposes its runtime on unmount, so observations close with the popover.
* fix(plugins): a sidebar item releases its subscriptions when it unmounts
Each mounted sidebar item gets its own API scope over the host's client, created
during render so its first frame has content. The scope closes a microtask after
unmount (StrictMode's remount keeps it), when the item moves to another host, or
when the installation stops. Replaces the installation-wide shared runtime.
* refactor(plugins): one Paseo client per plugin; authors own their subscriptions
usePaseo() in every plugin surface, and client.paseo in setup, is the
installation's client, created with the installation in the registry. Authors
release their subscriptions in their own cleanup; plugin teardown disposes the
client and ends the rest. Nothing is created or disposed per mounted surface:
the per-surface runtimes, their boundaries, and the per-run Command Center and
slash command APIs are gone.
* refactor(plugins): one Paseo client per plugin; authors own their subscriptions
usePaseo() in every plugin surface, and client.paseo in setup, is the
installation's client, created with the installation in the registry. Authors
release their subscriptions in their own cleanup; plugin teardown disposes the
client and ends the rest. Nothing is created or disposed per mounted surface:
the per-surface runtimes, their boundaries, and the per-run Command Center and
slash command APIs are gone.
* fix(usage): show usage from released hosts
Normalize released-host usage responses at the client boundary so existing usage and pins remain available across app and daemon versions. Include account labels in legacy responses for older apps.
* fix(usage): keep the usage capability check out of the public SDK
* test(server): retry transient locks during runtime test cleanup
* test(usage): model hosts without either usage capability
---------
Co-authored-by: Claude <noreply@anthropic.com>
The long-lived session wait hit Node’s headers timeout while OpenCode kept working. Reconcile execution events against active state and the durable log so transport observation failures do not fail healthy turns, including across event gaps and reconnects.
* feat(server): run internal plugins in-process through the plugin runtime
* fix(server): harden internal plugin startup and boundary checks
* feat(usage): add plugin usage sources and Codex discovery
* fix(usage): share SDK helpers and bound source cache
* feat(usage): migrate usage fetchers to plugin sources
Serve the legacy usage RPC from discovered source reports so older apps retain the same host usage list while new clients use the source registry. Keep credential reads and normalization inside each source.
* refactor(usage): shape sources as fetch functions
Keep the plugin usage API limited to stable report and presentation helpers. Source-local parsers and timeout handling preserve credential and response parity while the cases move alongside each source.
* feat(usage): resolve reports from active session accounts
Resolve source inputs at fetch time so model and credential changes select the account each session spends. Keep the legacy discovered-account RPC for existing clients. Sync OpenCode model events outside foreground turns so usage follows model switches.
* fix usage source discovery and session report parity
* test adapter usage references through daemon RPC
* name usage reference contract across server
* refactor(usage): own usage RPCs and validate session auth
* feat(omp): resolve session usage from active OAuth credential
* fix(omp): match usage source input and valid credential fallback
* test(omp): use upstream credential database schema
* test(omp): verify live session usage references
* fix(usage): declare YAML parser and consolidate credential validation
* feat(app): show usage sources in the composer and a Usage screen
Replace the provider-matched usage in the context ring tooltip with a
composer pill driven by agent.get_usage_report, add a sidebar Usage
screen over usage.list_reports grouped by host, and move the host
settings Usage page to usage.list_reports. Hosts without
features.usageSources show an update message. Source icons come from
the report's catalog SVG. The app no longer calls provider.usage.list.
* fix(usage): declare the Paseo requirement on internal usage-source plugins
A manifest without requirements.paseo targets Paseo before 0.8, so the daemon
refused to start eight of the nine usage sources.
* fix(app): keep host connection statuses live under the React Compiler
The hook read a store version only as a memo trigger (void version); the
compiler dropped it, so pages that mounted before a host came online kept
showing it as connecting.
* fix(app): place the context ring slot before the usage pill
The ring reserves its slot before it has data. With the pill first, the empty
slot sat between the pill and the mic; now it falls in the flexible space.
* refactor(app): extract the settings page layout into PageLayout
Desktop puts the title in the page above a centered column; compact moves it
into a back header. Settings renders the same pixels through it.
* feat(app): lay out the Usage screen like a settings page
The title moves from the header bar into the page on desktop; compact gets the
settings back header, which returns home when the screen was opened by URL.
* refactor(app): drop the unused actions slot from PageLayout
* test(plugin): classify the usage source SDK entry
* fix(usage): preserve platform test contracts
* fix(settings): preserve page title test contract
* feat(usage): identify and cache reports by account ID
Keep report identity stable across input routes and token rotation, and make selective refresh operate on known account reports.
* feat(app): read usage reports through report IDs
Share report data between agent pills and host views while preserving the existing popover interaction for the next UI phase.
* fix(usage): keep account identity aligned with credentials
Read labels from local token claims and use one credential lookup for usage fetches and account identification. Bound cached profile identities across token rotation.
* fix(usage): isolate failure IDs and prune expired reports
Failure reports live outside the account-key namespace, and cache writes discard entries past the five-minute TTL.
* refactor(app): share the compact time-ago prose formatter
Move the tab tooltip's "3m" -> "3m ago" wording into utils/time so the
usage card can reuse it.
* feat(usage): show when each report was fetched
The usage card reads fetchedAt as "Updated 3m ago": a tooltip on the
card header where hover exists (web, desktop), printed in the card
footer on native and compact. Shared by the pill popover and the Usage
screen.
* feat(usage): open the pill popover from cache and refresh on demand
Opening the popover no longer forces a vendor fetch; it shows the cached
report. A Refresh control force-refreshes only this report
(usage.list_reports with its single ID and forceRefresh), shows
Refreshing... while it runs, keeps the previous report visible, and says
"Unable to refresh usage" if the request fails.
* feat(app): let ToolbarButton show a tooltip other than its label
An icon action can now say a status in its tooltip ("Updated 3m ago")
while keeping its label for accessibility.
* feat(usage): refresh a report from its card header
Refresh is an icon button at the right edge of every usage card header,
in the pill popover and on the Usage screen. It force-refreshes only that
report, writes the result to the report's cache and its host list, shows
a spinner and disables itself while pending, and prints "Unable to
refresh usage" in the card on failure. Its tooltip carries the
freshness; without hover the freshness stays in the card footer. The
status indicator sits just left of the button, and the popover's
separate refresh row is gone.
* fix(server): include internal plugin assets in built output
* feat(server): built-in plugins in /plugins (#5544)
* feat(server): load compiled built-in Codex plugin
* test(server): guard built-in plugin loading and installs
* test(server): verify built-in plugins from dist in CI
* refactor(server): migrate remaining usage sources to built-in plugins
* test(server): guard all packaged usage source registrations
* docs: describe usage validation across plugin hosts
* test(ci): include built-in plugins in routing contract
* test(server): import moved Codex usage source
* fix(server): retain ACP updates received with session creation
* refactor(server): give Pi extension handling a registry
* feat(server): map Pi task extensions to Paseo task lists
* feat(server): map Pi subagent extensions to the Subagents track
* refactor(server): hydrate Pi child sessions asynchronously through the extension host
Child session JSONL reads blocked the daemon during live events and replay. The extension host now prepares immediate upserts and bounded async timeline hydration; replay awaits it, and closed sessions discard it. Preserve claimed notification text so completion details remain visible in the parent conversation.
* feat(server): present Pi question extensions as one question form
Keep the extension questionnaire in one Paseo permission while Pi walks its RPC dialogs. The host defers an early dialog until the form is answered, then replies to each dialog using the collected choices.
* fix(server): satisfy Pi extension start hook contract
The server build checks the session interface more strictly than the fast typecheck. Return the optional permission explicitly so the adapter satisfies the published extension contract.
* docs: document Pi extension adapters
* chore(server): format Pi subagent captures
These real RPC fixtures were added before the formatter ran over the captured JSON. Normalize whitespace so repository format checks pass without changing the captured data.
* test(server): resolve Pi child fixtures as local paths on Windows
File URL pathnames retain a leading slash before a Windows drive letter. Convert fixture URLs to filesystem paths and escape those paths when inserting them into captured JSON, so child timeline assertions exercise the real file on every CI platform.
* refactor(server): remove Oh My Pi handling from the Pi provider
* fix(server): keep Pi extension adapter failures out of core session handling
* fix(server): bound Pi child-session hydration during history replay
* fix(opencode): support v2 while preserving v1 provider behavior
* fix(opencode): target v2.0.7, keeping v1 support
OpenCode 2.0.4 removed the /api/health endpoint and the plugin
await-activation endpoint, and moved form reply/cancel under the session
namespace. The v2 adapter was pinned to the 2.0.3 SDK, so any 2.0.4+
binary failed at helper-server startup with an opaque HTTP 404.
- Pin @opencode/client to 2.0.7 and require v2 >= 2.0.4 at runtime
selection, with a clear error for older binaries.
- Use server.info for helper-server readiness.
- Move form reply/cancel/list to session.form and send the permission
decision field instead of reply.
- Replace plugin.awaitActivation with a plugin.list readiness wait. A cold
location registers its config-derived commands, skills, and providers
asynchronously; an empty inventory means it is still warming.
- Send session.command.name, which replaced command, and drop the removed
SkillInfo.slash field.
- Update the v2 test harness, unit tests, dependency lockfile, CI matrix,
and provider docs to 2.0.7.
* fix(opencode): target 2.0.10 and address v2 review findings
* test(opencode): colocate v2 suites with their modules
Greptile's review flagged that the v2 tests lived in top-level
opencode-agent.*.test.ts files and imported the readiness helper, v2
client, harness, and API types from a distance. Move each v2 suite beside
the module it describes and leave the top-level files to the v1 public
path:
- opencode/v2/agent.test.ts: session lifecycle, structured output,
resume, reconciliation, interrupt/replacement, helper isolation.
- opencode/v2/readiness.test.ts: location readiness deadline behavior.
- opencode/v2/permissions.test.ts: question and approval routing.
- opencode/v2/configuration.test.ts: permissionRules, feature flags, and
helper-isolation decisions, exercised directly through the module.
- opencode/v2/mapping.test.ts: model and mode normalization.
- opencode/runtime-client.test.ts: runtime version selection.
The v1 recovery, permission-action, and catalog-deadline suites stay in
place. Test counts are preserved, with deeper per-module coverage for
configuration.
* test(opencode): split v2 local E2E by behavior and add harness helpers
The re-review flagged the v2 scenarios in opencode-bridge.local.e2e.test.ts
for inline event filtering, permission-kind branching, and polling in test
bodies. Move stream, permission, and subagent coordination behind helpers
in opencode/test-utils/v2-local-e2e-helpers.ts, and split the session
scenarios into opencode-v2.local.e2e.test.ts so each file covers one
behavior:
- opencode-bridge.local.e2e.test.ts: process sharing, shell.env scoping,
the native tool bridge, and runtime version selection.
- opencode-v2.local.e2e.test.ts: session lifecycle, structured output,
streaming, approvals, MCP preapproval, stop/replacement, rewind and
subagents, commands, and skills/attachments.
Each helper waits on or answers the provider directly, so the tests state
domain intent and assert the collected result instead of polling or
branching inline. The CI-selected "versioned runtime" case stays in the
bridge file, and the live approvals and stop/replacement scenarios were
re-run green through the new helpers.
* fix(ci): preserve required server checks across OpenCode versions
Run v1 and v2 integration coverage sequentially so branch protection keeps its existing static check names.
* fix(agents): load startup history before registering sessions
A failed history stream while adding a runtime notice must close the unregistered session instead of leaving an initializing agent behind.
* fix(agents): limit startup history before buffering it
Apply the existing timeline content limits to each incoming history event so raw cumulative tool output is not retained until registration completes.
* fix(opencode): resolve Windows executables from platform packages
Recent v1 npm releases place opencode.exe in the architecture dependency rather than the CLI bin directory. Resolve that dependency from the CLI package while retaining support for the older layout.
* refactor(opencode): name public runtime option types
* test(opencode): let the spec own both runtime versions
Install each version in an isolated temporary directory and select it through provider configuration. Restore the CI workflow unchanged so version coverage is owned by the spec rather than workflow orchestration.
* test(opencode): launch native Windows binaries in isolated specs
Avoid npm command wrappers retaining locked executable files after helper shutdown during temporary installation cleanup.
* fix(opencode): preserve the legacy path when adding v2 support
Fall back to the existing client when version probing is inconclusive. Load v2 code and its bridge plugin only after positive v2 selection, keep v1 sessions undecorated, and restore original v1 discovery errors.
* fix(opencode): retry inconclusive version selection
A failed or unrecognized version probe previously cached the legacy fallback, pinning an upgraded binary to the v1 API until client recreation. Retry detection on the next operation while keeping the legacy path available.
* fix(opencode): release legacy runtime after inconclusive probe
Retryable version detection can clear the selected client after falling back to v1. Track whether the legacy client was selected so shutdown still releases its server without touching an unused v1 manager during a v2-only run.
* test(opencode): assert full runtime notice history
* fix(opencode): restore notice and retry behavior after rebase
---------
Co-authored-by: Jonathan karras <6307688+karrots@users.noreply.github.com>
Co-authored-by: Dhruv Singal <dhruvsingalabc@gmail.com>
Expose Claude Agent SDK's documented extraArgs map through providerOptions so Claude-specific flags such as --chrome reach the CLI.
Document boolean and valued flags, link the upstream contract, and validate through schema tests, the public session interface, and a real Claude launch.
* fix(server): probe Cursor thinking options per model
Cursor ACP only reports thought_level for the current model, so the catalog was copying Haiku's Off/On onto Grok 4.6 until the next probe happened to land on Grok.
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(server): drop inherited thinking options when an ACP catalog probe fails
A rejected model switch was leaving the probe session's thinking list on the failed model, so Grok could still advertise Haiku's Off/On.
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(server): probe per-model thinking when session/new has no thought_level
Composer/Auto as the Cursor ACP default omits thought_level, so the catalog helper skipped every model switch and Grok kept an empty thinking list.
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(server): discover model options without changing preferences
Use the provider model catalog extension instead of probing model switches that persist user preferences. Keep provider-specific discovery in each shim and validate extension responses at the boundary.
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Mohamed Boudra <boudra.moha@gmail.com>
* fix(providers): remember model selections across workspace creation
Keep saved model intent independent of catalogue readiness and use the caller-owned host and directory from the first render. Share discovery through provider-owned keys while preserving execution context.
Give each target one committed snapshot, publish actual transitions atomically, and share catalogue content across client directory associations. Regression coverage exercises successful repeated creation and provider lifecycle boundaries.
* fix(providers): preserve shared cache bodies during eviction
Admit directory references with their bodies, retain reference recency during eviction and recovery, and remove the unused persisted index. Refresh known workspace catalogues concurrently.
Make repeated-creation picker states explicit and seed mode tests from the installed provider catalogue, preserving assertions across Codex versions.
* fix(providers): bound discovery independently per provider
Settings refresh repairs all remembered targets, which otherwise starts an unbounded number of discoveries. Admit work after catalogue-key deduplication and keep each provider’s budget independent so a stalled provider cannot block another. Preserve active accounting across configuration changes and settle obsolete queued requests during replacement and shutdown.
Add backend-only exact-provider controls for the Paseo tool catalog across native and MCP-backed agents, including persisted configuration, session snapshots, public documentation, and regression coverage.\n\nCo-authored-by: Matt Cowger <matt@cowger.us>
* fix(opencode): reuse servers across agent sessions
Paseo's always-present per-agent environment forced every OpenCode session onto a dedicated helper, multiplying cold starts and memory use. Route ordinary sessions through one shared helper while a daemon-owned session bridge preserves exact environment and caller-scoped tools; keep custom process environment and MCP configuration isolated.
* fix(build): bundle the agent bridge from existing dependencies
The helper package duplicated Zod and changed the fixed-output desktop dependency graph even though its tool wrapper was identity-only. Use the server's existing Zod runtime directly so packaged daemons can materialize the bridge without extra traced dependencies.
* refactor(server): isolate provider runtime composition
Keep provider-specific lifecycle and tool-catalog wiring behind an agent-owned runtime boundary so the generic daemon bootstrap remains provider-agnostic. The runtime now owns bridge startup, snapshot injection, and idempotent cleanup.
* fix(desktop): prebundle the agent bridge plugin
Packaged daemons run from an application archive where esbuild cannot spawn its binary. Produce the self-contained plugin during the server build and make compiled runtimes read that artifact; source-mode tests retain a gated compiler fallback.
MCP and extension initialization can keep provider control-plane requests pending beyond 30 seconds even when the response will arrive. Use a configurable 60-second provider default and include the active RPC phase in timeout errors.
* feat(opencode): add active-turn steering via native V2 prompt
Add OpenCode support for active-turn steering (PR #3394) using the
native V2 session prompt with delivery: steer. Implements
steerActiveTurn, pending steer echo correlation for multiple steers,
prompted event handling, and definitive vs ambiguous error handling.
Updates docs and SDK to 1.18.18 for typed V2 access.
* fix(opencode): use V1 prompt for steer so model sees it
V2 prompt was admitted to a separate inbox not visible to the V1
processor, so "How old is bigbird?" was shown inline but never in
the LLM context. Switch steer to session/prompt_async (V1) which
queues while busy and is included in the next LLM call's history.
Update tests to expect V1 and keep the timeline grouping.
* fix(opencode): honor steering permission contract
Admit the steer before denying permissions that block its delivery, matching the provider steering contract. Remove the abandoned V2 path and SDK upgrade because active-turn delivery uses the existing prompt endpoint.
* chore(opencode): preserve SDK patch metadata
---------
Co-authored-by: Mohamed Boudra <boudra.moha@gmail.com>
* fix(server): stop a steered prompt from deadlocking on a permission
Claude parks its turn inside the blocking canUseTool callback while a
plan approval or a tool prompt waits for an answer. The SDK stops to
read its input stream there, so a message sent at the open card is
queued and never read. The user must reject the card or interrupt the
agent by hand, and then send the message a second time.
Deny the pending permissions after the steer is admitted. The prompt is
already in the SDK input queue when the permission callback returns, so
Claude reads it in the same turn. Plan mode stays on and no turn
restarts.
One pass is not enough. A turn that made several tool calls opens its
next request 10 to 300 milliseconds after the previous one is answered,
so a single snapshot advances the batch by one and parks the turn again.
The Claude adapter already tracks the steers the SDK has not read, so
expose that as an optional AgentSession.hasUnreadSteer and answer each
new request while that method reports the prompt as unread. The release
stops the moment the prompt is read, so a permission the agent asks for
after it acts on the message still reaches the user.
Only the two entrypoints that carry a message from a person set the
flag. A scheduled prompt, a child-finished notification, and an
agent-to-agent prompt must not answer a card the user is looking at.
Denial left no timeline row, so a denied plan disappeared from the
transcript. Claude and Codex now push a plan_approval row with
approved: false. For Codex the row comes from resolvePlanPermission,
which is the path that dismissPendingPlanApprovals also uses.
In the app, queue mode held the message until the turn ended, and a
parked turn never ends. The composer now sends while a permission is
pending, and the button label agrees with the submit path.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix(server): keep a post-prompt permission out of the release
The release loop answered the pending requests before it asked whether
the agent had read the prompt. The agent can read the prompt during a
poll interval and then open a permission request because of it. The next
pass denied that request, so a decision that belongs to the user was
made for them.
The loop now asks first and answers second. The first pass stays
unconditional, because a provider that does not report unread steers
must still get one release.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix(server): make steer permission clearing provider-owned
Keep the daemon policy explicit while letting each provider queue the steer and release permissions that block it. This removes timing-based polling and preserves rejected plans as canonical timeline rows.
* fix(server): preserve omitted steer options
Only construct provider steer options when permission clearing is requested. System prompts otherwise keep the existing undefined options contract.
---------
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
Co-authored-by: Mohamed Boudra <boudra.moha@gmail.com>
A steer pushed into the live SDK query input survived query.interrupt():
the SDK dequeued it afterwards and resumed the turn the user had just
stopped. Codex clears pending turn input when it aborts a turn; Claude
does not, so its adapter now cancels the SDK messages it queued before
interrupting.
Interrupt-discards is the deliberate ceiling for steer cancellation:
Codex exposes no un-steer operation, so a cross-provider take-back
cannot exist. A steer the model never read keeps its transcript row.
Headless hosts only have ~/.config/cursor/auth.json from a CLI login,
not the desktop state.vscdb token. Without that fallback Settings usage
stayed unavailable on VPS-style machines.
Co-authored-by: Cursor <cursoragent@cursor.com>
* feat(agents): add active-turn steering
Let users choose whether an immediate prompt interrupts or joins the active turn. The daemon owns exact-turn admission, provider delivery, safe fallback, and canonical timeline identity while queueing remains client-side.
* fix(agents): stabilize steering history and replay QA
* fix(agents): preserve steering timeline order
* fix(agents): reconcile resumed provider history
* fix(agents): serialize steering with cancellation
* test(agents): prove Claude multi-tool steering
* test(agents): force explicit history hydration
* fix(agents): flush output before steer admission
* fix(providers): make catalog refresh deadlines configurable
Provider startup can legitimately exceed a fixed short timeout when catalog discovery has substantial initialization work. Give each provider refresh one configurable deadline covering availability and the complete catalog probe, and cancel named child operations before allowing retries so timed-out work cannot accumulate in the background.
* fix(opencode): abort catalog server acquisition
A provider refresh could expire while the helper server was still starting because acquisition only tracked the activity without observing its abort signal. Reject acquisition before transferring a server reference, while leaving the manager-owned startup reusable by the next caller.
* test(file-observer): allow native burst recovery
* feat(providers): add native options and exact MCP grants
Keep provider policy in provider-owned schemas so unattended Hub runs can grant only requested tools from injected MCP servers. Preserve the native resolved sandbox policy when a workflow override is present, including configured writable roots.
* test(hub): verify real provider tool policy
* test(app): avoid pinning transient restart status
Make provider catalogs compact and content-addressed so repeat launches can reuse the persisted snapshot instead of transferring it again. Remove eager global catalog loading and the stopped-agent defensive tail fetch, while keeping sequence-gap and reconnect recovery as the correctness paths. Add an opt-in production-ID Android profiling build for local startup traces.
* fix(acp): preserve permission choices
ACP question bridges encode answers as repeated permission options. Preserve each option through the UI and return the exact selected ID, while keeping auto-accept from choosing on the user's behalf.
* fix(acp): validate selected permission behavior
Reject inconsistent action IDs instead of allowing the action ID to override the response behavior. Move the misplaced browser spec so the current base CI contract can run.
* fix(acp): preserve invalid permission requests
Validate selected ACP actions before consuming the pending request so inconsistent client responses can be corrected and retried.
* test(acp): verify real permission questions
* fix(acp): reject empty permission action ids
* test(e2e): scope Kimi config to ACP spec
* feat(acp): auto-accept shared permission requests
Some ACP agents expose permission requests without a native mode that can bypass them. Keep the opt-in at the shared adapter boundary so every ACP implementation gets the same per-session behavior.
* test(acp): include the shared feature in provider expectations
* fix(acp): honor unattended auto-accept intent
* fix(agent): list client features without a model
* Reapply: Stop completed turns from appearing stuck (#2484)
* Reapply: Keep older chat history and image previews stable (#2490)
* Lock timeline layout stability with user-driven specs
Two reds reproduce the reported jank:
- new agent, first prompt: the row paints at top=100 then jumps to 144 through
authoritative hydration, because the draft tab and the agent panel each render
their own AgentStreamView for the same logical row.
- long loaded timeline, submit: scrollHeight collapses 10793 -> 9111 mid-turn,
because splitOrderedTail reclassifies measured rows into the virtualized segment
where the virtualizer falls back to estimateStreamItemHeight.
Also restores centered timeline rows: the pagination anchor wrapper introduced in
#2490 was a plain block div, which broke alignSelf: center.
* Measure the turn footer blink from when the footer appears
Bounding the window at the prompt row conflated two things: the footer arriving a
frame after the row, and the footer leaving the layout and coming back. Bounded from
the footer's first frame, the blink still reproduces, so the footer-ownership race is
real rather than a sampling artifact.
* Rename the shadowed prompt row binding in the frame recorder
* fix(desktop): allow concurrent worktree dev instances
Desktop terminals can inherit another instance's profile path and debugging flags. Scope the profile to the current dev root and select an available debugging port so instances do not share global runtime state.
* fix(desktop): expose development providers in dev builds
Desktop-managed daemons inherited no Paseo runtime mode, so dev builds constructed the production provider registry. Mark unpackaged internal launches as development while continuing to force packaged launches to production.
* fix(timeline): stabilize optimistic turn layout
Keep initial history loading dormant until user intent and preserve active-turn layout across delayed transport updates. Add painted-frame coverage for optimistic timeline transitions.
* fix(timeline): preserve optimistic prompts through hydration
Canonical replacement relied on a transport submission record that can settle before a provider echoes the user row. Keep presentation continuity owned by the unreconciled row while destructive history replacement remains governed by active transactions.
* fix(timeline): make submitted prompts canonical
Submission settlement previously depended on status updates that could arrive after the optimistic row. Persist every accepted prompt in the canonical timeline and keep the footer active until RPC and canonical acknowledgement converge.
* fix(timeline): stop older history pages repeating entries
Backward pagination counted source sequence rows, so page boundaries inside a merged projected item could return that item repeatedly. Count projected entries and advance from their display anchors instead, with server and browser regressions covering the contract.
* fix(timeline): resume agents from the latest history
Use a bounded tail probe on resume so long background runs appear atomically instead of replaying every missed page. Preserve unchanged views, reconcile local/live rows, and keep skipped history reachable through backward pagination.
* fix(timeline): reconcile turn state after missed updates
Turn completion events can be missed while an agent timeline is unsubscribed, leaving footer activity stale even after the directory reports idle. Reconcile the single-owned turn lifecycle from authoritative resume snapshots while preserving newer live evidence.
* test(timeline): cover running state across refresh
Protect the hydration boundary where directory state must restore consistent tab, composer, and footer activity. Expose the running indicator semantically so the regression checks user-visible state instead of spinner implementation details.
* fix(timeline): keep history loading geometry stable
Reserve a fixed history-start gutter regardless of pagination state so showing or hiding the loader cannot shift timeline content.
* fix(timeline): keep turn activity coherent
Derive footer, tab, and composer activity from one lifecycle and submission projection. Idle directory transitions now request an authoritative resume so missed terminal events converge without treating directory delivery order as turn authority.
* fix(tests): align checks with runtime boundaries
* fix(desktop): stop dev app surviving its terminal
Electron was detached from the desktop runner, allowing it to outlive a stopped workspace terminal with broken output pipes and retain the single-instance lock. Keep Electron owned by the runner and preserve explicit process-group cleanup for detached child trees.
* fix(timeline): preserve activity through turn handoff
Correlate lifecycle events and snapshots with daemon-owned turn identity, and keep optimistic submissions active until their canonical row arrives. This prevents stale updates and RPC timing from exposing a false idle state.
* refactor(timeline): remove submission timing state
* fix(timeline): restore prompt identity on resume
* fix(timeline): exclude attachments from canonical prompt text
Structured provider prompts can include hidden history attachments before the user's text. Persist only plain prompt blocks so authoritative user messages match what the composer displayed while provider input remains unchanged.
* refactor(agent): remove unused pending-run message identity
* fix(desktop): let Chromium allocate the debugging port
* refactor(agent): remove impossible finalized-turn tracking
* fix(timeline): load history after keyboard scrolling
Treat observed upward displacement as user intent so keyboard and scrollbar navigation can wake dormant pagination.
* test(server): preserve accepted turn stream order
The manager-owned accepted turn boundary intentionally precedes coalesced output. Assert that lifecycle event explicitly while retaining the timeline-before-terminal ordering contract.
* test(app): scope sidebar fixtures by identity
Shared E2E daemon state may contain unrelated projects. Keep stale-route coverage tied to its named missing project and drag only the two rows seeded by the reorder spec.
* fix(timeline): gate prompt revisions by client capability
* fix(timeline): finalize canonical prompt delivery
Treat provider identity as a single stored-row transition so legacy clients receive one deliverable canonical prompt. Release identity-pending rows before later foreground output or the current terminal while preserving stale-turn isolation.
* fix(server): preserve legacy prompt pagination
* fix(timeline): keep assistant block identity stable
Assign canonical block IDs during live promotion so terminal finalization cannot remount image-bearing rows.
* fix(timeline): apply covered revisions during resume append
Run the exact canonical revision pass before overlapping resume pages append newer rows, while leaving cursor advancement to the incremental path.
* fix(server): preserve queued agent transitions
Project each queued live snapshot before awaiting stored metadata so later lifecycle state cannot overwrite an earlier transition.
* fix(server): snapshot queued agent updates
Capture each live agent payload before it enters the per-agent queue so shared nested state cannot change while an earlier storage read is pending.
* docs(app): date agent turn compatibility cleanup
* refactor(timeline): mint submitted prompt identity at acceptance
Canonical prompt rows now keep one immutable wire identity while provider identity remains internal for rewind translation. This removes the unreleased same-sequence revision and legacy projection contract.
* refactor(app): remove prompt revision reconciliation
Submitted prompt rows are immutable on the wire, so resume and live reducers no longer need to apply already-covered same-sequence updates.
* test(app): cover submitted prompt rewind roundtrip
* docs(timeline): describe immutable submitted prompt identity
Provider echoes now enrich daemon-internal identity only; wire rows keep their acceptance-time identity and rewind performs the provider translation.
* test(server): validate mock rewind targets
Reject unknown provider message IDs so app rewind coverage exercises wire-to-provider identity translation.
* fix(server): validate rewind before interrupting
* fix(server): omit rewind identity for daemon prompts
* test(app): guard rewind eligibility and rejection
* fix(server): preserve prompt identity across hydration
* refactor(timeline): remove stale revision remnants
* Revert "fix(server): preserve prompt identity across hydration"
This reverts commit 6adcf7e5b0.
* fix(timeline): settle missed submission acknowledgements
* fix(opencode): prevent late aborts from canceling replacement turns
OpenCode aborts are session-scoped, so provider idle alone does not make a session safe to reuse. Keep abort retries owned by the stopping turn and require both abort settlement and provider idle before dispatching the next runner operation.
* fix(opencode): retain the stop barrier through provider idle
Keep the stopping operation authoritative until both its session-scoped abort and provider-idle evidence settle, including stale idle-session interrupts. Also keep the real drift fixture credential-allowlisted and wait for command output pipes to drain.
* test(opencode): launch Windows fixture shims safely
* fix(opencode): recover stop observation safely
* fix(opencode): preserve autonomous wake through stop
* fix(opencode): retain deferred wake completion
* fix(opencode): serialize stop boundary replay
* test(opencode): use shared real e2e lane
* fix(opencode): own stops started during replay
* fix(opencode): retry stop observation on demand
* fix(opencode): follow replacement stop boundaries
* fix(opencode): retain wakes during stop recovery
* fix(opencode): hold replay turns behind boundary
* fix(opencode): filter canceled reconnect output
* fix(opencode): bind status evidence to stop
* fix(opencode): acknowledge owned cancellation
* fix(opencode): reconcile stale busy evidence
* fix(opencode): bind replay terminals to wakes
* refactor(opencode): own the stop boundary in one explicit state
The run boundary had grown into a mutable flag bag plus a raw-event
buffer that was replayed and re-classified after the fact. Each new
interleaving needed another predicate, and the classification kept
losing autonomous wakes.
Restore stopping as a turn-state variant that owns exactly two things:
the session-scoped abort it issued (with its single owned retry) and the
canceled run's terminal. Runner operations wait on both; a failed abort
fails closed. Provider status reconciliation now ends the stop through
the same transition an SSE terminal takes.
The canceled run's residue is suppressed only up to that terminal.
Events published afterwards are a new run by construction and take the
normal live path, so a wake no longer waits for abort settlement and
cannot be dropped by a retention heuristic.
* fix(opencode): emit cancellation before adopting the next run
The acknowledgement ran off a promise combinator, so a wake adopted from
the event right after the terminal could reach subscribers before the
canceled turn's turn_canceled. Acknowledge synchronously from whichever
proof lands first instead.
* docs(opencode): attach the stop doc comment to its interface
* fix(opencode): retry the stop in progress instead of opening a second
Pressing Stop again after an abort failed built a detached boundary: its
fresh abort could succeed while the canceled turn kept waiting for an
acknowledgement, so the daemon sat out the full cancel timeout before
force-canceling. A repeated Stop now replaces the abort owned by the
stop already in progress, which is what carries the acknowledgement and
what lifts the fail-closed gate.
* fix(opencode): accumulate every abort a stop has issued
A repeated Stop replaced the stop's abort settlement instead of adding
to it, so a first request still in flight — the slow-teardown case the
cancel cap exists for — no longer held the gate. The replacement could
start while that request was live and get cancelled by it, which is the
race this boundary exists to close. Reuse now waits for every abort the
stop issued, while only the newest can hold the gate closed.
* fix(opencode): scope abort settlement to the session, not the stop
Abort is session-scoped, so its settlement has to be too. Tracking it on
the stop meant a new stop boundary silently dropped an older request
that was still in flight, and that request could still cancel whatever
run started next. The session now accumulates every abort it has issued;
the stop is left owning only the run it is stopping.
* fix(opencode): recheck the abort gate after waiting on it
A waiting prompt read the settlement once, so a Stop issued while it
waited was invisible to it: the older abort would settle, the prompt
would start, and the newer one could still cancel it. Drain until the
settlement observed is the current one.
Canonical provider message IDs and optimistic client IDs occupy different namespaces. Preserve both identities for deterministic reconciliation while retaining a dated content fallback for older daemon timelines.
Pi and OMP share only the JSONL child-process transport while retaining provider-owned launch, RPC, runtime, session, history, and permission behavior. Removes captured fixtures in favor of typed harnesses and real-provider coverage.
Closes#2006Closes#2060
* fix(pi): avoid raw images for text-only models
Pi text-only models reject image content, and Pi persists the rejected raw image in session history. Materialize those image prompts to local files and send path hints while preserving raw images for vision-capable models.
* fix(pi): require explicit vision support
Default unknown Pi model capabilities to the text-hint path so raw image forwarding only happens for models that declare image input. Also keep materialized image files in a private per-user temp directory with restrictive file modes.
* fix(pi): harden materialized image temp dirs
* test(server): accept per-user image attachment dirs
* fix(server): use private temp dirs for materialized images
* fix(server): refresh missing image temp cache
* test(server): clean image temp dirs
* fix(providers): show Claude Ultra Code by capability
Expose Ultra Code for xhigh-capable Claude models and keep first-party Claude model metadata in one manifest so future model updates do not need scattered feature lists.
* fix(providers): preserve Claude fast mode aliases
Normalize versioned Claude model IDs through the manifest before checking fast-mode support so dated Opus IDs keep the same feature surface as their base model.
* fix(providers): keep Claude fast mode first-party
Anchor Claude model normalization so custom settings models that merely contain first-party IDs do not inherit manifest-only fast mode.
* fix(providers): preserve Claude runtime aliases
Keep Fast Mode capability checks strict to manifest IDs while allowing provider-form Claude runtime model strings to resolve known model metadata.
* fix(opencode): prevent indexing the entire home directory
Paseo launches opencode serve with cwd=os.homedir() and refreshes the
global provider snapshot with directory=/Users/admin. OpenCode treats
that as a workspace and starts location services + bigram indexing for
the entire home tree, causing ~466% CPU and ~4GB RAM usage.
- Use a neutral scratch directory as the opencode serve cwd.
- Use a separate scratch directory for global provider catalog refresh
so model/mode discovery no longer triggers home directory indexing.
Fixes high CPU/RAM when Paseo starts opencode with no explicit project.
* fix(opencode): use realpath-aware matcher for home detection in catalog refresh
Switch the home-directory check in fetchCatalog from a string-based
path.resolve() comparison to createRealpathAwarePathMatcher, so we
catch macOS /private/var/... aliases, symlinks, trailing separators,
and Windows casing — consistent with the rest of opencode-agent.ts.
Also:
- Hoist the matcher to module scope so each fetchCatalog call doesn't
rebuild it (the matcher runs realpathSync twice on construction).
- Log a debug line when we rewrite the cwd to the scratch path, so
it's easy to diagnose missing per-directory config in catalog scope.
- Update opencode-agent.test.ts to expect the scratch directory when
cwd === os.homedir(), with a comment pointing to the rationale.
* fix(opencode): isolate helper server home
* fix(opencode): pass semantic global catalog scope
* fix(opencode): release catalog acquisition on home resolution failure
---------
Co-authored-by: rex-chang <rex-chang@users.noreply.github.com>
Co-authored-by: Mohamed Boudra <boudra.moha@gmail.com>
* refactor(agent): decouple Paseo tools from MCP
Keep MCP as the fallback adapter while exposing the same runtime catalog for providers that can register tools directly.
* fix(protocol): keep native tool support server-internal
* fix(i18n): sync Japanese shortcut labels
* fix(agent): preserve schedule provider error
Mohamed Boudra
