Inner sidebar content was pinned to a fixed openWidth (derived from store state) so it kept its full size during open/close animation. But store state only updates on pointer-release during a drag, which meant inner content didn't reflow while the user was actively dragging — only after they let go.
Keep openWidth as the inner width when the sidebar is idle (so slide-out animation still shows full-width content under the clipping aside), but during an active resize bind the inner content to a CSS variable that applyLiveWidth updates on every pointermove. The aside continues to use direct px width for transition reliability; the variable is just a side channel for the inner element.
OpenInAppButton and ProjectActionsButton now render icon-only triggers in the header instead of icon + text. Padding is tightened accordingly and the now-unused selectedButtonLabel / formatActionButtonLabel helpers are gone.
With narrower triggers the previous center-aligned dropdown with translate -30px started overflowing the viewport on the left edge for ProjectActions. Switched ProjectActions dropdown to align="start" and OpenInApp dropdown to align="end" (matching their positions in the header) and removed the manual translate offset.
Also fixed a long-standing inconsistency where clicking an already-running action in the ProjectActions dropdown re-ran it instead of stopping; the non-compact branch was missing toggleStopIfRunning=true that the compact branch already passed.
MiniChat header now uses bg-sidebar to match the main desktop header palette.
Sidebars now slide open/closed with a 200ms cubic-bezier(0.22, 1, 0.36, 1) transition on width. Inline transitionProperty/Duration/TimingFunction are used directly on the aside so the change is browser-driven rather than relying on Tailwind arbitrary classes that didn't reliably trigger the width transition.
Inner sidebar content is pinned to the open width with shrink-0 so it stays full-size while the outer aside collapses around it, eliminating the vertical jitter that came from content reflowing when its container shrank. The chat-frame keeps a constant 1px border and 10px radius on all four sides — only the left/right border color toggles between border/50 and transparent — so no layout reflow happens when sidebars open. Mask-corner overlays are always mounted and animate their left/right offset and opacity in lockstep with the sidebars instead of mounting/unmounting.
Header now spans the full window width above the [sidebar | chat | right-sidebar] row instead of nesting inside the central column. The chat area becomes a self-contained framed window with its own border and rounded corners on all four sides, and sidebars sit flush against the header sharing its bg-sidebar so the seam is invisible.
Removed the duplicated shell controls the old layout needed to fake header-height inside sidebars: portal host on RightSidebar, paddingTop reservation, top drag overlay, duplicated layout-left / chat-new buttons in SidebarHeader, the showDesktopSidebarChrome block in SessionSidebar, and the conditional traffic-lights inset on the desktop header. Mac WCO inset now lives only on the header.
Moved the new-session action into the SessionSwitcher dropdown as its first item, removed the standalone chat-new button from the header, relocated scheduled-tasks into the left action group of the sidebar header, and bumped ContextPanel tab strip to h-10 to balance the more prominent header.
Restart OpenCode after successful updates so the new version is active
Open native About menu into the app About dialog
Update desktop View menu actions for the new layout
* fix(ui): associate save plan title label
* test(ui): make save plan label test locale-independent
---------
Co-authored-by: Isaac Sanchez <isanchez-hawkins@arize.com>
* feat(quota): add Wafer.ai quota provider
- New provider: wafer.js fetches from https://pass.wafer.ai/v1/inference/quota
- Auth: reads wafer/wafer-ai/wafer_ai keys from auth file
- Timeout: AbortSignal.timeout(15_000) with timeoutSignal.aborted detection
- Response: parses remaining/limit/overage/usedPercent/window_end/plan_tier
- valueLabel: planTier + remaining/limit + overage suffix
- Window: 5h (18000s) via resolveWindowLabel
- Cross-runtime: added to web registry, UI types, VS Code dispatcher
* fix(quota): wafer provider fixes — auth alias, decompression, logo
- Add 'wafer.ai' auth alias to match actual auth key format
- Use 'Accept-Encoding: identity' header to fix Bun fetch
decompression issue with Cloudflare-backed responses
- Match copilot valueLabel format: 'planTier · X / Y left'
- Add wafer logo alias so Providers and Usage pages resolve
to the same wafer.ai logo from models.dev
* style(quota): fix indentation of timeoutSignal declaration
* fix(quota): derive window duration from API instead of hardcoding
Compute windowSeconds from window_end - window_start timestamps,
with WAFER_WINDOW_SECONDS (5h) as fallback if timestamps are missing.
---------
Co-authored-by: Bohdan Triapitsyn <artmore@protonmail.com>
* fix(ui): update viewport height without visualViewport
* fix(ui): initialize visual viewport state from helper
---------
Co-authored-by: Isaac Sanchez <isanchez-hawkins@arize.com>
* fix: make PWA install and OpenCode update toasts dismissible
Both 'Install OpenChamber' and 'OpenCode update available' toasts use
duration: Infinity with no close affordance, so they persist on screen
until the user accepts (install/update) or reloads the tab. For users
who do not want to install the PWA or upgrade right now, this is
intrusive and there is no opt-out.
Add a Dismiss button (sonner cancel action) to both toasts. When the
user dismisses:
- PWA: persist a flag in localStorage so the prompt does not reappear
on future sessions. Accepting Install still works as before.
- OpenCode update: persist the dismissed version in localStorage. The
toast will appear again only when a newer version becomes available.
Adds new i18n keys pwa.installPrompt.dismiss and
opencodeUpdate.toast.actions.dismiss across all seven locales (en,
es, ko, pl, pt-BR, uk, zh-CN).
* test: extract toast dedup helpers and cover with 28 unit tests
Lift the dismissal-decision logic out of usePwaInstallPrompt and
OpenCodeUpdateToast into a React-free sibling module so it can be
unit-tested directly. The React surfaces remain sole owners of side
effects (storage writes, toast.info, event listeners); the new module
only answers 'should we show?'.
New module openCodeUpdateDedup.ts exposes four helpers:
- shouldShowPwaInstallToast(input) - three gates: dismissed,
sessionShown, hasActiveToast.
- shouldShowOpenCodeUpdateToast(input) - empty version, seen set,
dismissed===version gates; a different dismissed version lets the
toast resurface for the new release.
- resolveOpenCodeUpdateVersion(detail) - parses CustomEvent payloads
defensively (null/non-object/non-string -> '').
- resolveOpenCodeUpgradeStatusVersion(status) - parses upgrade status
payloads (status falsy / available!==true / latestVersion non-string
-> '').
Consumers now call the helpers and only run the side-effect when the
decision is true. Behaviour is unchanged.
Coverage: 28 tests via bun:test, 33 expects, all pass first try.
Prevents duplicate native notifications when a session is open in both sidebar and editor
Uses an extension-host claim step so only one VS Code webview displays each notification
Keeps notification behavior unchanged for single-panel sessions
Targets active editor chat before falling back to sidebar
Adds session open-in-editor action in the VS Code sidebar
Uses attachment badges for VS Code file and selection context
Desktop shells no longer subscribe to the browser notification stream
Electron keeps using the native desktop notification path only
Web browser notifications continue to use the notification stream
Web notifications now listen to the notification stream directly
Native notifications no longer depend on PWA push being enabled
Window focus is respected when deciding whether to send background push notifications
Change the chat text-selection Add to notes action to use the existing markdown selection payload, matching Add to chat formatting behavior while still falling back to plain text when markdown extraction is unavailable.
Stop distilling or summarizing selected text before saving it to project notes. The action now appends the selected content directly so code, lists, emphasis, and other markdown survive.
Increase the project notes text limit from 1000 to 3000 characters to account for saving raw selected text instead of a short generated summary.
Update user-facing copy across locales from distilled insight language to selected text language.
Disable the active Zen summarization flow because the unauthenticated/free Zen provider is no longer available and now returns usage-limit errors for this feature.
Keep /api/text/summarize as an API-compatible stub that returns local sanitized or distilled fallback text with summarized=false, rather than attempting external model calls.
Remove notification and voice playback summary behavior from runtime paths. Notification {last_message} now always uses normalized truncated text, and TTS playback ignores historical summarize request fields.
Hide the notification summary settings and voice summarize-before-playback controls while preserving legacy persisted settings for compatibility. Also disable Zen model startup validation and make Zen model list routes return empty results.
Update module documentation and tests to describe the retired provider behavior and the remaining compatibility stubs.
Move VS Code/Cursor desktop notifications onto the webview Notification API instead of the extension-host watcher path, which could not reliably produce native OS notifications.
Route OpenCode runtime events from the shared sync pipeline into the VS Code webview so completion, error, question, and permission notifications use the same live event stream as the UI.
Respect the OpenChamber notification settings in VS Code, including template rendering, completion cooldowns, permission auto-accept suppression, and the notify-while-focused mode.
Use VS Code's window focus signal from the extension host instead of document.hasFocus() inside the webview, so hidden-only notifications are suppressed while Cursor or VS Code is focused across platforms.
Avoids restarting OpenCode after transient health probe failures
Coalesces concurrent health checks and briefly caches probe results
Adds configurable health timeout, retry threshold, interval, and cache settings
Shows loaded skills in slash autocomplete with clear type badges
Stabilizes keyboard navigation across autocomplete menus
Fixes skill link rendering and skill autocomplete scrolling
* fix(status): pick latest assistant message in single pass
Sync reconciliation can splice messages out of array order, so the
previous "filter + sort + last" pipeline could pick the wrong record
as the latest assistant message and surface a stale status.
Replace it with a single-pass scan that compares created-time (tiebreak
by id) — same selection, but resilient to non-sorted message arrays.
* refactor: use comparator + running max for latest assistant pick
Address Greptile review: collapse the dual-exit-point single-pass max
finder (separate continue + isNewer branches, repeated null checks)
into a single isLater(a, b) comparator with one running-max comparison
per iteration. Same selection rule, easier to verify.
---------
Co-authored-by: vhqtvn <8930337+vhqtvn@users.noreply.github.com>
* fix(sessions): align archive-cascade count with executed list
When archiving a parent session, the dialog showed the total descendant
count but the actual archive request skipped already-archived
descendants — so the toast/result didn't match what the user was told.
Collect descendants once, then filter per-action: hard-delete cascades
to every descendant; archive skips already-archived ones. Dialog count
and the ID list passed to archiveSessions/deleteSessions now agree.
* fix(sessions): snapshot descendant IDs at dialog open
Address Greptile review: handleDeleteSession populated the dialog
count from a fresh collectDescendants() call, then confirmDeleteSession
re-collected at confirm time — so a background sync that mutated
childrenMap while the dialog was open could re-introduce the same
count-vs-executed mismatch this PR fixed.
Store the computed descendant IDs in deleteSessionConfirm and pass
them through to executeDeleteSession, so the executed list is always
the snapshot the user was shown. The no-dialog direct-execute path
also routes through the snapshot for consistency.
---------
Co-authored-by: vhqtvn <8930337+vhqtvn@users.noreply.github.com>
The repo's .nvmrc contained the literal string `lts`, which isn't a
valid nvm version specifier — running `nvm use` (or any tool that
respects .nvmrc) fails with "Version 'lts' not found". When that
happens, the shell falls back to whatever Node is on PATH, which
often doesn't meet vite's >= 20.19 / 22.12 requirement and silently
breaks the build.
`lts/*` is the canonical "latest LTS" specifier in nvm and resolves
correctly. Same intent, working syntax.
Co-authored-by: vhqtvn <8930337+vhqtvn@users.noreply.github.com>
The archived bucket is the only sidebar group that routinely grows
into the hundreds or thousands of rows (projects accumulate archived
sessions over time). On main, expanding the bucket mounts every row
at once — for a project with ~1400 archived sessions that's
thousands of subscriptions, layouts, and event handlers committed
on a single render.
Wrap the archived bucket's row map in `@tanstack/react-virtual`
(already in deps) when the visible count crosses 50. DOM stays
bounded to the visible window + overscan regardless of total size.
Implementation notes:
- Threshold + estimate are local constants; only the archived
bucket is gated, every other group keeps its eager render.
- The virtualizer needs a scroll element. The sidebar's scroll
lives on an ancestor `ScrollableOverlay`; rather than threading
a ref through every intermediate component for this one use
case, we walk up the DOM from the row container.
- `measureElement` handles variable row heights (subagent
expansion, status indicators, etc.).
- ResizeObserver re-measures on sidebar resize, mirroring the
pattern in ChangesSection.
Co-authored-by: vhqtvn <8930337+vhqtvn@users.noreply.github.com>
* fix: prevent mobile keyboard from occluding terminal viewport
On mobile, 100dvh does not shrink when the soft keyboard opens, causing
the terminal and other absolute-inset-0 views to extend behind the keyboard.
Add useVisualViewport hook that tracks window.visualViewport.resize/scroll
via RAF. On mobile, use the visual viewport height (actual visible area)
for the root container instead of 100dvh, so content resizes above the
keyboard when it opens.
* fix: skip unchanged visual viewport updates
---------
Co-authored-by: Bohdan Triapitsyn <artmore@protonmail.com>
Aligns tab behavior with browsers and editors: pressing the middle mouse
button on any closable tab in SortableTabsStrip closes it without going
through the visible close icon.
- onAuxClick handler triggers onClose when event.button === 1.
- onMouseDown preventDefault on middle button suppresses the browser's
autoscroll affordance.
- Both handlers no-op when the tab is not closable, keeping non-editor
tab strips (skills catalog, settings sections, etc.) untouched.
The same session can appear in two render contexts at once — most
visibly under "Recent" and under its project's root group. Expansion
state was keyed by bare session id, so toggling the chevron in one
context flipped it in the other; the user clicked one row and a
duplicate elsewhere expanded/collapsed in sync.
Switch the expansion set to composite keys of the form
`<renderContext>:<active|archived>:<sessionId>`. SessionNodeItem
already computes this exact shape for its menu instance key
(menuInstanceKey); reuse it as expansionKey so the two render-time
instances of one session keep independent expand state. Updated:
- isExpanded reads `expandedParents.has(expansionKey)`.
- Chevron click / Enter / Space pass `expansionKey` to toggleParent
instead of session.id.
- The memo-equality check compares the same composite key on both
sides so it correctly re-renders when only one context's state
changes.
- toggleParent's parameter is renamed `expansionKey` to reflect the
new shape; the implementation is unchanged.
- The auto-expand-parent-on-subagent-navigation effect doesn't know
which context the user will look at the parent in, so it fans the
parentID out to all four (project|recent) × (active|archived)
combinations.
Storage is bumped to `oc.sessions.expandedParents.v2`. Existing v1
data (bare session ids) is one-shot migrated by fanning each id
across all four contexts and rewritten under the v2 key; the v1 key
is then removed so a downgrade-then-reupgrade doesn't re-migrate.
Co-authored-by: vhqtvn <8930337+vhqtvn@users.noreply.github.com>
When a new session draft is open on mobile and the user switches
projects via the ProjectBar, the draft target was not updated,
causing a mismatch with the desktop project/branch selectors.
handleProjectSwitch now mirrors the desktop handleDraftProjectChange:
when a draft is open, it uses setActiveProjectIdOnly +
setNewSessionDraftTarget instead of setActiveProject. The draft is
also guarded against preserveDirectoryOverride to prevent
overwriting a worktree-locked directory.
When no draft is open, the original setActiveProject behavior is
preserved.
Co-authored-by: Jovines <jovines@qq.com>
* feat: add copy-as-markdown and copy-as-json buttons to question card
Adds two small icon buttons in the QuestionCard header so users can copy
the full question payload (text, options, descriptions) to the clipboard
in either Markdown or JSON form. Useful for forwarding interactive
questions to an external LLM or note.
- Markdown serialization preserves question headers, multi-select hints
and option descriptions.
- JSON serialization mirrors the on-wire QuestionRequest shape for
programmatic reuse.
- Toast feedback on success/failure using the shared @/components/ui
wrapper and the existing copyTextToClipboard helper (clipboard API
with execCommand fallback).
- New i18n keys added across all 7 locales.
* test: extract question serializers and cover with 15 unit tests
The two serializers backing the copy-as-markdown and copy-as-json
buttons on QuestionCard were defined as top-level consts inside
QuestionCard.tsx. Exposing them for unit tests would trip the
react-refresh/only-export-components ESLint rule because that file
also exports the QuestionCard component.
Move serializeQuestionAsMarkdown and serializeQuestionAsJson into a
React-free sibling module, questionSerializers.ts, and re-import them
from QuestionCard.tsx. Behavior is byte-identical.
Add 15 unit tests via bun:test covering header fallback, multi-select
hint gating, blank-description elision, ordering across multiple
questions, JSON canonical shape (no transient id/sessionID), boolean
normalisation, and the empty-questions edge case.
* fix: preserve state when reconnect-time fetches fail
Several client API methods swallowed fetch/SDK errors and returned an
empty value (`[]`, `{}`), which was indistinguishable from a successful
"server says nothing here" response. Reconnect resync paths trusted that
empty result as authoritative and deleted local state — so after a
network blip (sleep/wake, wifi reconnect, tunnel switch), the UI could
show:
- sessions stuck on the "running" indicator (status never cleared)
- pending permission prompts disappearing from the UI
- pending question prompts disappearing from the UI
and only a page reload would recover. A related case: `listAgents`
silently returning `[]` defeated the 3-attempt retry loop in
`useAgentsStore` because the loop never saw an error.
The systematic fix:
- `getSessionStatusForDirectory` now returns `null` on fetch failure
(vs the previous `{}`); the reconnect resync treats only a non-null
response as authoritative — candidates missing from the response are
written as `{type: "idle"}`, candidates after a failure are left
untouched.
- `listPendingPermissions`, `listPendingQuestions`, and `listAgents`
now throw on SDK/network failure. The pre-existing outer try/catch
blocks in `resyncBlockingRequestsForDirectory` and the retry loop in
`useAgentsStore` were already in the right shape — they just never
fired because no exception was thrown. A small `formatSdkError`
helper renders the SDK `{data, error}` shape into the thrown message.
- `permissionStore.setSessionAutoAccept` catches the new throw and
falls back to whatever sync-store snapshots provide; the next SSE
event or reconnect resync will catch up anything missed.
AGENTS.md gets a new "Distinguish fetch failure from empty success"
subsection documenting the principle (throw vs `T | null` patterns,
when to pick which, the retry-loop trap) so this doesn't regress.
Adds 3 regression tests covering the resync paths: existing
questions/permissions are preserved when the corresponding `list*`
method throws, and a permission-fetch failure does not block the
question block from running (verifies per-block try/catch isolation).
* fix: pause reconnect loop when offline or hidden
The SSE/WebSocket reconnect loop retried indefinitely with no awareness
of whether the browser was online or whether the tab was even visible.
Three issues compounded:
- No `online`/`offline` event handling. With a foreground tab on a dead
network, we'd hit the server every ~5s forever, and on network
recovery we'd wait up to ~5s for the next probe instead of reacting
to the `online` event.
- No visibility awareness. A backgrounded PWA on a flaky link kept
probing at the same rate as a foreground tab. The browser does
throttle hidden-tab timers, but the intent wasn't expressed in code.
- The "exponential backoff" math
`min(5000, max(retryDelayMs, 250) * (failures <= 1 ? 1 : 2))`
re-initialized `retryDelayMs` to 250 every iteration, so the cap of
5s was never reached — we waited 500ms forever after the second
failure. Not actually exponential.
Now:
- `online` event aborts the current attempt (if disconnected) and
cuts inter-attempt waits short. `offline` event aborts so the loop
enters the slow-probe path immediately.
- `computeRetryDelay` returns the long cap (60s) when `navigator.onLine`
is false or the tab is hidden; the short cap (5s) when foreground +
online. The `online` event is the expected recovery path; the 60s cap
is a fallback for browsers that miss the event.
- Real exponential growth: `BASE * 2^min(failures-1, 8)`, clamped.
- New `waitForRetry` helper interrupts on `online`,
visibility-becomes-visible, and abort signal — so visibility/network
recovery doesn't wait out the rest of the current sleep.
AGENTS.md gets a "Reconnect-loop pacing" subsection alongside the
fetch-failure rule, since they're the same family of resilience
concerns.
One regression test: simulates offline + failed first attempt + `online`
event after the failure; verifies the next attempt fires within seconds
instead of waiting the full 60s offline cap.
* fix: long-cap backoff for permanent 4xx server errors
Before this commit the reconnect loop didn't distinguish HTTP error
types. A stuck-path client (wrong URL after server upgrade) or an
expired-auth client (stale token) would hit the server at the normal
5-second cap forever — ~12 reqs/min, indefinitely, with no path to
recovery besides the user reloading.
Now the catch block extracts an HTTP status (looking on `error.status`
and `error.response.status` — the SDK exposes both depending on the
code path) and overrides the backoff:
- 4xx other than 408/429 → use the long cap (60s) immediately.
Blind retries won't fix wrong path / bad auth / forbidden, so don't
pound the server. waitForRetry's `online` / visibility-visible
interrupters still apply — when an operator fixes the server-side
config and the client comes back to foreground, recovery is prompt.
- 408 (Request Timeout) and 429 (Too Many Requests) → normal
exponential path. Those are retryable in spirit.
- 5xx / network / unknown → normal exponential path. Unchanged.
AGENTS.md gets a new bullet under "Reconnect-loop pacing" covering
this — the rule fits naturally alongside the existing `navigator.onLine`
and visibility signals.
Two regression tests:
- A 404-throwing SDK doesn't fire a second attempt within 250ms (proves
we left the exponential path). After `online` interrupts the wait,
subsequent attempts fire promptly — proves the override doesn't break
recovery once the underlying problem is fixed.
- A 429-throwing SDK recovers within 2s — proves 429 still hits the
fast exponential path and isn't caught by the permanent-error branch.
---------
Co-authored-by: vhqtvn <8930337+vhqtvn@users.noreply.github.com>
Bohdan Triapitsyn