* refactor(worktrees): fetch source once during creation
* fix(worktrees): remove worktrees in background
* fix(worktrees): show background removal progress
* fix(worktrees): name the worktree in removal toasts
* feat(worktrees): fetch remote source branch before worktree creation
New worktrees based on a local branch that is behind its upstream now
fetch first and branch from the remote-tracking ref, so they are not
born stale. A global setting (on by default) in Settings > Behavior
controls this, and fetch failures toast a warning and fall back to
local state instead of blocking creation.
* fix(worktrees): wire fetch-source toggle to store and honor failed runtime fetches
The Behavior toggle only persisted the setting; the consumer reads the
config store at creation time, so a just-toggled-off setting kept
fetching until the next hydration. Update the store optimistically on
toggle and on page load, and roll it back when the save fails.
The VS Code runtime bridge resolves git fetches with { success: false }
instead of throwing, which the consumer read as success and silently
based the worktree on the stale remote ref. Treat any non-success
result as a failed fetch: warn and fall back to local state, matching
the web/desktop/mobile path.
* fix(worktrees): stop new remote-based worktrees from tracking the base branch
Creating a worktree with a remote start ref made git auto-track the
base branch (branch.autoSetupMerge), so with the new remote fetch every
behind-root worktree was born with upstream origin/<base> and plain
git push refused under push.default=simple.
The new branch's own upstream does not exist until its first push, and
the bootstrap deliberately refuses to write tracking config for refs
that were never fetched, so --set-upstream-to cannot re-point it.
Suppress the auto-track with --no-track on new-mode creation from a
remote ref: the branch ships with no upstream, matching the behavior
before the remote fetch until the first push sets it. Explicit
upstream keys now also win over the remote start ref inference,
aligning the create path with the validate path and the VS Code
runtime.
* fix(worktrees): keep the pre-create remote ref refresh soft
The client fetch and the server's pre-create fetchRemoteBranchRef both
refresh the same branch, and the second fetch throws on failure — so a
connection dropped between the two turned the promised soft fallback
into a rejected creation even though the remote-tracking ref was
already available locally.
The refresh is now best-effort when the ref exists locally (creation
proceeds from it) and still mandatory when the ref was never fetched,
preserving the materialization behavior for remote-only branches.
Applied to both the web server and the VS Code runtime.
* chore: ignore the .openchamber app runtime state directory
The panel stored notes, todos and plans inside one shared JSON file that
six unrelated domains also wrote to, synchronised itself through window
CustomEvents, and could only read plans. It is now Project knowledge:
server-owned storage with explicit routes, a store with rollback, a
section sidebar, plans that open and edit in place, and search across
all of it.
Notes and plans the user pins travel with every message sent in that
project. Pinning is project state, not an attachment to one message, so
it holds until unpinned and the work status panel names what is riding
along and can detach it.
Agent memory is added alongside, in two scopes: what is true about the
user, and what is true about this codebase. The split is not cosmetic —
a wrong project fact costs one project and is noticed, while a wrong
global fact quietly shapes every session everywhere and the user has no
code to check it against. It stays separate from notes so an agent
mistake cannot land in what the user wrote. Sessions receive an index of
titles only; bodies are read on demand, because an index carrying full
text grows until it crowds out the conversation.
Deciding what a session must be told, and whether it has been told, now
lives on the server. The client owned it before, which meant sessions
started without a UI — scheduled tasks, sessions the agent dispatches —
received nothing at all, and a tab's record of what it had sent outlived
the conversation: after compaction the agent no longer held the block
while the tab went on believing it did. What was delivered is recorded
in the session's own metadata, and compaction restores it through the
runtime that already restores pinned messages, in the same turn.
Agent memory ships dark behind OPENCHAMBER_MEMORY_ENABLE: unset, there
is no tool, no routes, no session index, no settings row and no panel
tab. Absent rather than switched off, so nothing invites turning on a
feature that has not been announced. Pinned notes and plans are
unaffected and ship as normal.
The preview panel worked by proxying a dev server through OpenChamber's own
origin and rewriting the HTML that came back. Anything the rewriter did not
anticipate broke, and pages that refuse to be embedded never loaded at all.
This deletes the proxy (-1604 lines and its tests) and merges the preview and
browser panels into one surface backed by a real Chromium view.
What the panel is now
- A `<webview>` in its own session partition: logins and cookies persist, hot
reload works because nothing is rewritten, DevTools are one click away.
- Annotation: pick one element, drag a region, or draw freehand, write a note,
and it reaches chat with a screenshot of the visible page with the marks on it.
- Toolbar: hard reload, page zoom, device sizes, a light/dark switch that
applies to the page rather than the app, and cookie/cache clearing scoped to
the panel alone.
- Several pages at once, each tab showing the page's own favicon, and an address
bar that suggests pages already visited in this project.
- Dev servers are listed from what is actually listening on the machine, checked
against what a project announced, so a server is offered no matter how it was
started. One that is still starting is waited for instead of failing.
Remote dev servers
The desktop app binds a local port and pipes raw bytes to the OpenChamber host
over the existing authenticated connection, so the page keeps its own origin at
the root of its own host. The reachable set is exactly what discovery reports
and is re-checked per connection, so an authenticated client cannot dial
arbitrary local services on the host. Links and redirects to another loopback
port stay on the machine that served the page. A tunnel that cannot be opened is
reported; it is never replaced by the plain loopback URL, which would answer
from the user's own machine under a remote address.
Agent control
Browser actions are a separate `openchamber_web` tool: open, snapshot, click,
type, scroll, inspect computed styles, resize between mobile/tablet/desktop, and
capture a screenshot into `.openchamber/screenshots/` in the project. The
existing `openchamber` tool keeps sessions, worktrees and scheduled tasks. Each
has its own setting in the new Settings -> General -> OpenChamber Tools section,
and the plugin is not injected at all when both are off.
Capability belongs to the connected client, not to configuration: a client
declares on its event stream that it can drive a page, which only a Chromium
host does. Exactly one client performs each request — it claims the request
before acting, and the first claim wins — because deciding by whose result
arrives first would be too late for a click that already happened. No client
listening is answered immediately with an explanation rather than a timeout.
Runtime boundaries
Web tabs get a plain iframe that can display a page but not inspect one. The
VS Code extension no longer offers the surface at all, since nothing that makes
the panel worth having works there. Mobile is unaffected.
Native boundary
Camera, microphone, location and device-picker requests from panel pages are
denied — Electron grants them by default when no handler is set, and the panel
loads whatever address the user types. Page capture, appearance emulation and
storage clearing verify that their target belongs to the panel's own session
instead of trusting a web-contents id from the renderer.
Persisted state
Stored `preview` tabs migrate to `browser` (v13 -> v14). Context panel tab
limits are now per surface, so filling one surface no longer evicts another's
tabs. Address history is stored per project and per runtime.
Documentation
`preview.mdx` and `desktop-browser.mdx` rewritten across all locales, the agent
tool settings path corrected, new `DOCUMENTATION.md` for the browser-control
broker and the dev tunnel, and the `ui-api-decoupling` skill updated where it
still described the deleted proxy.
Adds a work-status panel beside the transcript. Context fill, model and
cost, todos, running subagents and the permission requests blocking
them, branch and working-tree state, MCP servers, pinned messages and
context sources were scattered across the header, the composer and the
context panel — a blocked subagent was reported nowhere at all. The
panel reads them from live channels rather than persisted history, and
becomes an overlay where the chat is too narrow to seat a column.
It is on by default, including for existing installs. Because it now
carries these readouts, the desktop header and composer drop the ones it
duplicates: todo and changed-files chips, usage and MCP tabs. VS Code
and mobile keep theirs — neither hosts the panel.
Fixes MCP authorization, which was broken from the panel, invalidated by
a directory switch through a redirect URI that encoded the working
directory, and left the desktop app in the background because browsers
will not follow a custom-protocol link without a user gesture. The
settings page no longer asks the user to understand the MCP spec before
adding a server: one field takes the command or the link, with the kind
inferred and a visible override, and client-registration fields appear
only when a server actually asks for its own credentials.
Also: skills load from the panel instead of only when the composer's
slash autocomplete opens; the header button names the current instance
rather than falling through to the word "Instance" for relay hosts.
Three new optional UI settings keys, all migrated. No change to stored
MCP server configuration.
A diff is ordered by file path, which is almost never the order in which a
change makes sense. This adds a Walkthrough surface that reorders it: the model
groups related hunks into stops, explains what each group changes about
behavior, and orders the stops so each builds on the last. It explains and
orders; judging code stays with the existing Review action.
Reviews uncommitted work (all, staged, unstaged), a branch against its base, or
a pull request. Generation is always user-initiated — nothing runs on a timer,
on a file change, or as a side effect of opening a panel.
Invariants worth preserving:
- Hunk identity is derived on the server and only there. Ids are content
hashes, so an anchor that no longer resolves is proof the code it described
changed, and staleness needs no heuristics. The client matches ids to ids and
never recomputes them; two implementations would have to agree forever.
- The digest is never truncated. A diff that does not fit the model's context
is refused with an actionable reason, because a walkthrough written against
half a diff reads as confident and is wrong.
- Nothing disappears. Lockfiles and other generated output are excluded from
the model's input by name — never by size — and everything no stop covers is
listed at the end, so "have I seen all of it" stays answerable.
- Cost is explicit. Results are content-addressed, so returning the working
tree to an earlier state costs nothing; generation outlives its request, so a
refresh detaches the client rather than discarding paid-for work, and only an
explicit cancel stops it.
Supporting changes to shared modules:
- git: expose the existing getRangeDiff as GET /api/git
listUntrackedPaths and getUntrackedDiffs. The latter resolve the repository
once for a batch instead of per file, taking a panel
~340ms on an 80-file working tree.
- small-model: structured output across four wire forma
and abort signal, and an onOverflow policy so an oversized prompt fails
loudly instead of being silently clipped. A provider
remembered so the prompt-side fallback goes first next time.
- models.dev metadata: surface structured_output as tri
false blocks a model, a missing field does not, because the catalog omits it
for roughly half of all models.
Desktop and tablet only: VS Code serves Git through its
these routes, and the mobile shell does not consume the surface registry.
Docs: packages/docs walkthrough page in English and all eight locales.
Add an opt-in OpenCode plugin that replaces the built-in provider behavioral prompt with a minimal identity while preserving environment, project, MCP, skill, history, and tool context.
Track the active agent per session and apply the transform only to build and plan. Keep plan/build mode reminders and permission enforcement owned by OpenCode, leave all other agents untouched, and fail safely when the expected prompt boundary is absent.
Expose the feature in Behavior settings with localized guidance, explicit Save + Reload application, settings search integration, persisted boolean validation, and managed-runtime lifecycle composition that does not load the plugin while disabled or on external OpenCode servers.
Document the runtime contract and cover plugin materialization, config preservation, build/plan selection, agent switching, unknown prompt formats, and settings sanitization.
Left-side controls now use macOS traffic-light order (close, minimize,
maximize). Remove the unused auto option and default window controls to
the right for Windows and Linux frameless chrome.
Authored-by: Serhii Dziupin <makeittech@users.noreply.github.com>
Persists draft starter visibility across desktop and web settings
Adds a new OpenChamber visual setting with localization and search support
Hides the draft starter chips when the setting is off
Add a shared OpenChamber control service with two thin adapters — a native
`openchamber` tool injected into managed OpenCode, and new CLI commands — so
users can manage parallel sessions, worktrees, and scheduled tasks
conversationally through agents or from the terminal.
Control plane:
- New openchamber-control service owning a fixed action contract:
projects.list, models.list, session list/create/send/fork/status/messages,
and schedule list/create/run/delete/toggle. Session and worktree deletion
and project registration are deliberately not exposed.
- New openchamber-sessions module owning create/worktree/prompt orchestration,
Goal Mode dispatch, wait semantics (initial idle never counts as completion;
timeout and cancellation are failures), and explicit partial-failure results.
- Scheduled-task logic extracted into a service shared by routes, CLI, and the
agent tool.
Agent tool:
- Managed OpenCode gets a materialized plugin registering one typed tool with
a loopback-only callback, per-child ephemeral bearer (timing-safe, never
persisted or logged), and abort propagation into the service.
- The ~1.5k-token schema applies progressive disclosure: short descriptions,
server-side validation returning actionable usage errors, and intent
guardrails — created sessions/tasks are user-facing work (not age
self-delegation); worktree/goal/agent/variant/wait are omit-by-default;
dispatches produce no completion notification, and later result r
to session.messages, which now returns the authoritative sessionStatus.
- session.create without a user-named model picks from favorites/re
send/fork omit the selection and the service reuses the target session's
last user-message model, agent, and variant before falling back t
- An "Agent control tool" setting (default on, Save + Reload to apply)
disables plugin injection entirely.
CLI:
- New `openchamber session`, `schedule`, `projects`, and `models` commands
with automatic instance targeting, --wait/--timeout/--last-assist
worktree flags, and Goal Mode, preserving interactive, non-TTY, --quiet,
and --json contracts. The control HTTP timeout derives from the w
instead of the 4-second default.
UI:
- New built-in "Schedule a Task" starter (/schedule-task) running a
dialogue that defines a task and offers to create it via the tool after
explicit confirmation; Craft a Goal and Feature Planning gain the
handoff offer, and guided starters reserve the question tool for concrete
option choices. Localized in all 10 locales, migrated into custom
starter lists, hidden on VS Code.
- Sidebar shows CLI/agent-created sessions live via the control eve
- openchamber tool calls render with per-action titles and metadata.
Adds a General settings switch to show or hide the macOS menu bar item.
Persists the new desktopMacMenuBarEnabled setting and passes tray state into Electron.
Disables tray sync and native tray setup when the menu bar is turned off.
Improve OpenChamber responsiveness under large session workloads while fixing
cache, synchronization, and persistence correctness across runtimes, projects,
directories, and worktrees.
- prioritize selected and visible sessions during bootstrap and defer
non-critical enrichment work
- reduce redundant message loading, event processing, store publication, and
hidden sidebar work
- prevent stale session and message requests from overwriting newer
authoritative state
- preserve existing data when authoritative fetches fail instead of treating
failures as successful empty responses
- scope session materialization, messages, drafts, queues, todos, pins,
permissions, folders, tabs, Git state, and pull request data by runtime and
directory identity
- harden runtime switching, reconnect, cleanup, mutation reconciliation, and
persisted-state ordering
- preserve live subagent Task linkage when metadata arrives after an older
message request or while streaming parts are suspended
- coalesce overlapping tail refreshes without losing newer refresh demand
- improve cold-session loading by moving deferrable work out of the critical
bootstrap path
- isolate URL authentication, mobile credentials, native secrets, and other
runtime-owned state across endpoint changes
- bound long-lived caches and remove avoidable allocations from event and
rendering hot paths
- limit virtualization to archive collections where it improves rendering
without disrupting active sidebar layout
- stabilize session folders, pin ordering, expanded state, and persisted
sidebar behavior
- open skill files through the same secure editor and outside-workspace grant
flow used by file navigation, including worktree sessions
- expand regression coverage for stale completions, runtime collisions,
reconnect behavior, persistence races, authoritative empty results, and
subagent refresh ordering
- document the updated synchronization, cache ownership, performance, and
runtime-isolation invariants
Replace the legacy terminal flow with a shared authenticated WebSocket
runtime used across web, desktop, relay, and mobile surfaces.
- introduce the v3 terminal protocol with scoped attachments, snapshots,
ordered output, bounded replay history, reconnects, and explicit lifecycle
- harden PTY creation, restart, resize, close, force-kill, idle cleanup,
shell selection, login mode, environment sanitization, and appearance sync
- add runtime-aware terminal APIs with relay authentication and Electron parity
- add a fullscreen mobile terminal workspace with touch scrolling,
long-press selection, safe-area controls, quick keys, and Ctrl/Alt input
- add terminal selection attachments, preview detection, project actions,
shell settings, and localized UI
- harden Ghostty rendering, resize recovery, Unicode handling, block
characters, line height, and stale-row behavior
- remove the obsolete terminal SSE path and update reverse-proxy guidance
- expand terminal runtime, transport, input, selection, and store coverage
- avoid duplicate web builds when preparing mobile assets in root CI builds
* feat(chat): add desktop prompt navigator rail
Add a ChatGPT-style right-center prompt marker rail for web/desktop chat
with hover/keyboard preview panel, load-more for partial history (panel only),
Chat setting, and mod+alt+p shortcut. Disabled in VS Code across rail,
shortcut, settings, help, and search surfaces.
Co-authored-by: Serhii Dziupin <makeittech@users.noreply.github.com>
* fix(ui): read promptNavigatorEnabled from getState in shortcut handler
Match the file convention used by other shortcut handlers so the toggle
does not rely on a hook-level selector closure.
Co-authored-by: Serhii Dziupin <makeittech@users.noreply.github.com>
* fix(ui): drop use-no-memo and default prompt navigator off
Remove the project-unprecedented React Compiler opt-out, and ship the
prompt navigator as opt-in to match other recent chat UI toggles.
Co-authored-by: Serhii Dziupin <makeittech@users.noreply.github.com>
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Serhii Dziupin <makeittech@users.noreply.github.com>
* feat(electron): add Linux AppImage releases
* ci: cache Linux OpenCode CLI artifacts
* fix(ci): await Linux release inventory check
* fix(electron): add frameless window controls on Linux desktop
Linux AppImages were created without native WM decorations and without
in-app controls, leaving users unable to close the window with a mouse.
Treat Linux like Windows: frameless BrowserWindow plus the existing
WindowsWindowControls header buttons and app-menu entry. macOS keeps
hidden title bar with traffic lights unchanged.
Shared usesFramelessElectronChrome() helper drives main window, mini
chat, header insets, and titlebar controls.
Co-authored-by: Serhii Dziupin <makeittech@users.noreply.github.com>
* feat(desktop): add configurable window controls position by OS
Add desktopWindowControlsPosition setting (auto/left/right) with OS-aware
defaults: Linux left, Windows right. Wire frameless chrome controls in
Header, TitlebarLeftControls, and MiniChatLayout, plus a Sessions settings
control for Windows and Linux desktop shells.
Co-authored-by: Serhii Dziupin <makeittech@users.noreply.github.com>
* fix(desktop): address Linux AppImage release review findings
Propagate updater capability errors to the UI, treat missing
latest-linux.yml feeds as no-update, stop installed-apps IPC spam on
Linux, document FUSE/AppImage limits, add CHANGELOG entry, migrate
remaining btriapitsyn URLs, and run Electron Linux unit tests on PRs.
Co-authored-by: Serhii Dziupin <makeittech@users.noreply.github.com>
---------
Co-authored-by: jibanez-staticduo <staticduo@gmail.com>
Co-authored-by: Serhii Dziupin <makeittech@users.noreply.github.com>
Move per-session permission auto-accept policy ownership from the UI to the
OpenChamber server so enabled sessions continue running when clients disconnect
or the server restarts.
- persist explicit per-session policies in OpenChamber settings
- inherit the nearest explicit policy across subagent session hierarchies
- allow child sessions to opt out of an inherited parent policy
- immediately accept matching global and directory-scoped pending requests
- process future requests without requiring a connected UI client
- reconcile pending permissions after startup and event-stream reconnects
- deduplicate concurrent requests and retry transient reply failures
- synchronize policy updates across connected clients
- migrate existing browser-persisted policies to server storage
- suppress auto-accepted permission cards before they enter UI state
- show deduplicated permission toasts for inactive sessions
- preserve foreground-only permission handling in VS Code
- integrate directory-aware notification routing from main
- add coverage for persistence, inheritance, retries, reconciliation, pending
requests, client hydration, and inactive-session toasts
Arm the target button in the composer and the next prompt becomes a goal:
the server keeps the session working toward it (idle tick -> small-model
audit -> continuation) until the objective is verifiably complete, blocked,
or out of budget — even with the UI closed.
Server (packages/web/server/lib/session-goal):
- event-driven loop on the global SSE hub; goal state lives in
session.metadata.openchamber.goal (merge-safe patches, stale-write guard
by goal id), so it survives restarts and syncs to every client for free
- the small-model audit (objective + last assistant turn only, language
pinned to the objective) is the sole termination authority; blocked needs
3 consecutive verdicts, audit outages tolerate one unaudited continuation
then stop the goal as resumable-blocked
- hard stops: optional token budget, auto-continuation cap (Resume grants a
fresh allowance), turn errors; user abort pauses the goal instead of
blocking it, and resuming over an aborted tail nudges immediately
- token accounting as a snapshot of the latest turn (input + cache.read +
output), goal-relative via a creation baseline and segmented across
compactions; a compaction summary skips the audit and continues
- continuations reuse the session's own provider/model/agent/variant
UI:
- three-mode target button (arm / disarm / manage dialog), informational
goal strip with inline pause/resume and an Evaluating indicator, sidebar
state glyph, objective length counter (2000-char server clamp),
read-only completed goals
- goal entry points: composer (sessions and drafts), start-new-session-
from-answer dialog, plan implement dialog (plan content becomes the
objective), scheduled tasks (Run as goal + budget)
- Settings -> Chat -> Goal: feature toggle + default token budget with
three-layer parity (web server, client persistence, VS Code bridge);
VS Code renders goal state but hides the entry points (the loop runs in
the web server only)
Notifications: per-turn "ready" notifications are suppressed while a goal
is active; settling sends one final notification (desktop, web-push, APNs
generic titles with the session name as body) honoring the completion
toggle. Error/question/permission notifications are untouched.
Docs: user guide (session-goals) in all 9 locales + sidebar entry,
scheduled-tasks cross-reference, server module DOCUMENTATION.md.
Add Windows launch-at-login with background startup support and extend the
native tray integration to Windows.
Add a Windows-only setting to minimize or close the main window to the
system tray, persist it through desktop settings, and expose it in Settings
search and all locale dictionaries.
Keep tray state synchronized with live sessions on both macOS and Windows,
while preserving the existing macOS behavior.
Adds a server-side "small model" capability: direct, cheap LLM calls that
reuse the user's existing OpenCode provider logins — the mechanism OpenCode
uses internally for titles and summaries but does not expose through the
SDK or plugins. Zero new dependencies; plain fetch with per-provider wire
formats, credentials never leave the server.
Core (packages/web/server/lib/small-model):
- Resolution mirrors OpenCode's session scoping: explicit settings override
→ small_model from the OpenCode config → family scan within the session's
provider → the session's own model. The global provider scan only serves
callers without a session context, and background callers forbid it
entirely (restrictToPreferredProvider), so conversation content never
reaches a provider the user didn't pick — explicit choices excepted.
- Per-provider auth replicating OpenCode's plugin loaders: GitHub Copilot
(device token as bearer, no exchange), ChatGPT plan via the codex
Responses API (single-flight OAuth refresh written back to auth.json),
Anthropic messages, Google generateContent, generic OpenAI-compatible.
- OpenCode's free models (opencode/big-pickle, *-free) are never called
directly; unauthenticated providers are skipped by design.
- Prompt clamping to the model's catalog context limit; thinking disabled
where a wire switch exists (Z.AI/GLM, MiniMax-M3, Gemini Flash); robust
content parsing with a clear error when a thinking model spends its whol
budget on reasoning.
- Settings → Sessions gains a Small Model group: use-default checkbox plus
an override picker limited to authenticated providers, persisted with
web/desktop/VS Code sanitization parity.
Consumers:
- Session assist: a server-side watcher on the global SSE hub generates a
short recap and one suggested follow-up after a session idles quietly fo
a minute, stored on session metadata (openchamber.assist). Freshness is
keyed to the last assistant message id, so new activity invalidates the
payload everywhere with no extra writes. The chat shows the recap under
the last message after five quiet minutes and the suggestion as a
dismissible chip above the composer (tap fills the input, never sends).
Gated by a new Chat setting (default on) that is a hard generation
switch. Language is anchored to the conversation itself, with a
script-mismatch guard against model/backend language hallucination.
- TTS: a third input mode, summarized — long replies are condensed to
spoken prose before playback on any TTS engine.
- Git: commit-message and PR generation moved off the active chat session
onto the small model fed with real diffs and the commit list (bodies
included), with a session-transport fallback for free-model-only setups.
- Notes: Add to notes distills long selections into 1-3 dense sentences
preserving exact identifiers, with verbatim fallback on failure.
Fixes along the way:
- The global event watcher now starts unconditionally; it was gated behind
the desktop-notify env, leaving the server-side event hub dead in
packaged apps.
- OpenCode re-emits message.updated for old user messages after idle; the
watcher no longer mistakes those for new activity.
- Session metadata merges from a fresh read right before the PATCH, so
writes made during the generation window (suggestion dismissals, review
links) are preserved; the assist runtime stops during graceful shutdown.
Complete rebuild of voice input on a server-authoritative streaming
architecture, replacing the legacy Web Speech / whole-blob / WASM engines
and the dead voice-agent layer (~4k lines removed).
Speech-to-text (dictation):
- Client streams 16 kHz mono PCM16 chunks over /api/dictation/ws with
seq/ack ordering; buffered audio is retained and replayed on reconnect
- Server transcribes and streams live partial transcripts back;
segments auto-commit every ~15s with silence suppression and adaptive
finalization timeouts
- Local provider (default, zero config): sherpa-onnx models in a forked
worker process — auto-download with progress, staged extraction with
verification, corrupt-model auto-recovery, idle shutdown after 5 min
- Model catalog with settings picker (accuracy/speed ratings, sizes,
download/delete): Parakeet TDT v2 (English) and v3 (25 European
languages, auto-detected), Whisper base and tiny (multilingual, light)
- OpenAI-compatible provider for any Whisper endpoint
- Composer overlay with live transcript, volume meter, timer, and
cancel / insert / insert-and-send actions; failed transcriptions keep
their audio for retry or accepting the partial text as-is
- Configurable keyboard shortcut (default mod+alt+v) toggles dictation;
Enter confirms and Escape cancels while recording
- Overlay is pixel-aligned with the composer (measured footer height,
matching paddings/typography/gaps) — no layout shift when toggling
Text-to-speech:
- Local Kokoro provider (English, 11 voices) synthesized in the same
worker via /api/dictation/tts/speak, managed by the shared model
pipeline; sentence-pipelined playback keeps time-to-first-audio at
~1 sentence regardless of message length, and stop cancels in-flight
synthesis
- Sanitizer keeps inline-code content (strips backticks only), reads
interword slashes aloud, and removes only absolute file paths
Settings:
- Voice page unified: a single read-aloud toggle owns all playback
options (the confusing "Enable Voice Mode" is gone); a new "Enable
voice input" toggle (default on, persisted to settings.json) hides
the composer mic entirely when disabled
Mobile and transport:
- iOS/Android microphone permissions added (dictation was previously
impossible on mobile)
- Fixed Android WebSocket upgrades: the Capacitor WebView origin
(https://localhost) was missing from the packaged-client allowlist,
403-ing every WS connection — root cause of the old mobile SSE lock,
which is now removed for all transports
Security and conventions:
- All HTTP routes sit behind the global /api auth gate; the WS upgrade
explicitly validates the UI session and origin, with oc_url_token
narrowly allowlisted and covered by tests; the dictation socket mints
a fresh URL token before connecting
- Routes register before the generic OpenCode proxy; the client goes
through runtimeFetch/getRuntimeUrlResolver, and runtime switches
reset the dictation socket
- VS Code deliberately reports dictation as unavailable (no server
process in that runtime)
CI: workflow Node bumped 20 -> 22 to match the repo engines and fix
better-sqlite3 installs broken by node-gyp@latest on Node 20.
New dependency: sherpa-onnx-node (prebuilt N-API; macOS/Linux x64+arm64,
Windows x64 — Windows-on-ARM falls back to the OpenAI-compatible provider)
* feat: support OpenCode steer delivery / follow-up behavior settings
Implements issue #1766 — steer delivery mode for mid-turn message
insertion, replacing the old boolean queue-mode toggle with a tri-state
follow-up behavior setting (Steer / Queue / Send immediately).
- Plumbing: threaded optional delivery: 'steer' through sendMessage
-> routeMessage -> opencodeClient.sendMessage -> promptAsync
- Store: messageQueueStore stores followUpBehavior; migration from
legacy queueModeEnabled persisted state
- Settings: Chat -> Follow-up behavior shows three radio options
using existing settings UI patterns
- Composer: when session is busy, a floating queue button remains;
force-sending a queued message (via chip click) uses delivery: 'steer'
during a busy session; Steer button intentionally omitted — steer is
available via the two-gesture path (Enter to queue -> chip to steer)
- Keyboard: queue mode = Enter queues, Ctrl+Enter sends; otherwise
Enter sends, Ctrl+Enter queues
- Persistence: DesktopSettings, web settings payload, and server-side
sanitizer handle the new key with legacy fallback
- i18n: follow-up behavior section and option labels in all 9 locales
plus new chat.chatInput.actions.queue label
- Search: settings registry updated from chat.queue-mode to
chat.follow-up-behavior
Validation: type-check passes (no new errors), lint clean.
* fix(#1766): make steer mode actually steer
The followUpBehavior === 'steer' branch in handlePrimaryAction and the
keyboard handler was a no-op — both fell into the else branch and sent
without the delivery: 'steer' flag, so selecting 'Steer (insert into
the running turn)' in settings produced identical behavior to 'Send
immediately'.
- handlePrimaryAction: when steer mode is selected and the session is
busy, call handleSubmit({ delivery: 'steer' }) directly
- Keyboard handler: in steer mode, Enter steers and Ctrl+Enter sends
immediately (consistent with queue mode where Ctrl+Enter bypasses
the special handling)
Also removes the unused chat.chatInput.actions.queue i18n key from all
9 locales (it was a dead key after the Steer button was removed from
the composer).
Validation: type-check clean, lint clean.
* refactor(#1766): flatten nested ternary in followUpBehavior resolution
Replace nested ternary with explicit if/else chain per project code style
(CONTRIBUTING.md). Import FollowUpBehavior type explicitly for the new
let declaration.
* feat(chat): drop redundant 'immediate' follow-up mode, keep Queue + Steer
'Immediate' was wire-identical to 'Steer' on a busy session: OpenCode only
supports delivery 'steer' | 'queue' and defaults to 'steer', so an immediate
send (no delivery flag) already steered into the running turn. The three-mode
UI therefore exposed two settings that did the same thing.
Collapse to two modes — Queue (unchanged: client-side queue with edit/reorder)
and Steer. Any persisted/legacy 'immediate' (and legacy queueModeEnabled=false)
now maps to 'steer', preserving prior behavior. Removes the immediate option,
its keyboard branch, the i18n label across all locales, and narrows the
followUpBehavior union to 'steer' | 'queue'.
---------
Co-authored-by: Leonid Skorobogatyy <bash@opencode.itc.local>
Co-authored-by: Bohdan Triapitsyn <artmore@protonmail.com>
* fix(settings): persist per-model visibility and sibling selector state
The server-side settings sanitizer only allowlisted favoriteModels and
recentModels, so hiddenModels, collapsedModelProviders, recentAgents, and
recentEfforts were stripped on every write to settings.json — per-model
visibility and collapsed-provider state silently reset on every container
redeploy or settings reload.
Add the four missing fields to sanitizeSettingsUpdate:
- hiddenModels: sanitizeModelRefs(..., 1024) — same shape as favoriteModels;
1024 covers dense multi-provider setups while bounding persistence/memory.
- collapsedModelProviders: normalizeStringArray with Array.isArray gate
(matches usageDropdownProviders).
- recentAgents: normalizeStringArray (Array<string> per ui-store).
- recentEfforts: new sanitizeRecentEfforts validating Record<string, string[]>
(shape confirmed in ui-store + addRecentEffort action); trims/dedupes keys
and variants, caps at 128 keys x 5 variants/key (5 matches client slice).
No ui-store version bump or migration: zustand's default merge spreads
persisted state over defaults, so missing fields fall back to [] / {} until
the next toggle. favoriteModels and recentModels are untouched.
Tests: 8 new cases in settings-helpers.test.js using the real
sanitizeModelRefs / normalizeStringArray — round-trips, empty-[] parity with
favoriteModels, garbage rejection, and a full-payload regression test.
* fix: sync model selector settings
---------
Co-authored-by: Bohdan Triapitsyn <artmore@protonmail.com>
Packaged desktop showed no sessions in 1.12.4. Root cause: the sanitized
session-list proxy path added in #1538 forwarded the renderer's
"authorization" header (the OpenChamber UI client token) to the managed
OpenCode upstream alongside the managed "Authorization" credential.
OpenCode does not recognize UI client tokens, so every session-list
request answered 401 — only in the packaged app, because only its
renderer (openchamber-ui:// origin) attaches a bearer token; dev web and
dev Electron run same-origin without one. The legacy http-proxy path
overwrote the header correctly, which is why everything except session
lists kept working.
Proxy fix:
- proxy-headers: filter the client "authorization" header out of
forwarded request headers; the OpenCode upstream must only ever see
its own managed credentials. Covered by tests.
Desktop cwd:
- electron: launch the managed OpenCode CLI from the user home instead
of app userData, matching upstream desktop behavior. userData-as-cwd
made OpenCode treat the app-data folder as a separate empty workspace.
Home directory poisoning loop:
- directoryPersistence: stop replaying localStorage homeDirectory
through synchronizeHomeDirectory on boot/auth resync. The persisted
value is only a boot-time cache; replaying it re-wrote stale values
(e.g. a project path) into desktop settings on every start, overriding
the authoritative /api/fs/home resolution.
- persistence: never overwrite an injected window.__OPENCHAMBER_HOME__
with a persisted value.
- useDirectoryStore: host switches happen in place (no reload), so
re-resolve home from the new runtime's /api/fs/home on endpoint
change instead of keeping the previous host's value.
- opencode client: only short-circuit to the injected desktop home when
the active runtime is local; remote runtimes ask /api/fs/home.
Settings hygiene:
- persistSettings: log field names only — change payloads can carry
credentials (UI password, client tokens, tunnel tokens) that must not
reach the log file; drop step-by-step log chatter.
- validateProjectEntries: only stat project paths when the incoming
update actually touches the projects list, not on every settings save.
- remove the write-only approvedDirectories setting everywhere and add
a migration that strips the stale key from persisted settings.
Tests:
- usePluginsStore.test: register an own runtime-fetch module mock so the
suite is independent of process-global mock.module leakage from other
files, and restore globalThis.fetch after the suite.
- persistence.test: clean up the window global created for the suite.
Let users curate the draft welcome chips: pin existing commands and skills
as starters, remove them, and drag to reorder — all inline on the draft
screen via a '+' picker dialog and per-chip remove, with no separate
settings UI.
A starter references a command or skill; its scope is inherited from the
item (user-scope -> global, project-scope -> per-project). Global starters
persist to settings.json (useUIStore + client/server sanitizers); project
starters persist to the project config alongside worktree setup commands.
The two scopes form ordered namespaces shown global-first then project,
reorderable only within each group.
The six built-in Session magic-prompt commands are the default global set
and stay available in the picker for re-pinning if removed; they keep their
bespoke icons, while user commands/skills fall back to the Commands/Skills
section icons. Chip labels are normalized (/simplify-code -> 'Simplify
code'). Missing commands/skills are skipped rather than shown broken.
Drag-to-reorder works on desktop and mobile: rectSortingStrategy for the
wrapping multi-row layout, CSS.Translate (no scale) so the lifted chip
doesn't stretch, and MouseSensor + long-press TouchSensor so taps still
submit and swipes still scroll. The '+' picker is a searchable dialog on
every surface.
* feat: add collapsible reasoning traces with animated labels
* feat(ui): redesign reasoning blocks with merged collapsible Thought view
- Replace per-part reasoning blocks with a single merged block per turn
(VSCode Copilot pattern), controlled by new `groupReasoningBlocks` store flag
- `ReasoningTimelineBlock` redesigned: chevron toggle, summary preview on
collapsed header, 'Thinking'/'Justification' label when expanded, BusyDots
while streaming, auto-scroll to bottom during live streaming
- Short texts (< 120 chars) render inline without a toggle
- Summary now strips markdown and truncates at a word boundary with ellipsis
- New `MergedReasoningPart` component merges all reasoning parts for a message
into one block at the position of the first reasoning part
- `defaultExpanded` prop lets callers override initial expand state
- Remove `.thinking-dot` CSS animation (replaced by BusyDots component)
- Fix reasoning markdown font-size: use `--text-markdown` instead of `--text-meta`
* refactor(ui): scope working phrases inside useAssistantStatus and simplify reasoning status
- Move WORKING_PHRASES array and getRandomWorkingPhrase() inside the hook
so they are no longer exported (were only consumed by ReasoningPart which
no longer needs them)
- Change the 'reasoning' activity status text from a random working phrase
to the deterministic string 'thinking' — matches the new UI label
* test(ui): expand ReasoningPart tests for new collapsible and summary behavior
- Update baseline test to use text long enough to trigger the collapsible
path (short texts now render inline) and assert on the correct aria markup
- Add test for 'Justification' label when pre-expanded via defaultExpanded
- Add test for 'Thinking' label for the thinking variant when expanded
- Add test verifying summary is a word-boundary-truncated excerpt ending with
an ellipsis character
* i18n: rename 'Reasoning Traces' to 'Thinking Blocks' and add thought key
- Rename settings label from 'Show Reasoning Traces' → 'Show Thinking Blocks'
across all supported locales (en, es, ko, pl, pt-BR, uk, zh-CN)
- Add `chat.reasoningTrace.thought` key to all locales (used by merged
reasoning block header in completed state)
* feat(ui): add collapsibleThinkingBlocks setting with full persistence wiring
- New boolean store field `collapsibleThinkingBlocks` (default true) with
`setCollapsibleThinkingBlocks` action; persisted to localStorage
- Threaded through DesktopSettings, SettingsPayload (API types), desktop
persistence (sanitize + apply), web appearance persistence, appearance
auto-save watcher, and server-side settings-helpers sanitize/format
- Server defaults to true when the field is absent in formatSettingsResponse
- MessageBody reads the flag: false → render reasoning as plain AssistantTextPart;
true → existing collapsible/merged block path
* feat(settings): expose Collapsible Reasoning Blocks toggle in visual settings
Add a checkbox under the 'Show Thinking Blocks' row (visible only when
showReasoningTraces is enabled) that toggles the collapsibleThinkingBlocks
preference. Follows the existing toggle pattern: div role=button, keyboard
handler for Enter/Space, Checkbox primitive, aria-pressed attribute.
* i18n: revert showReasoningTraces label rename and add collapsibleThinkingBlocks strings
- Revert 'Show Reasoning Traces' → 'Show Thinking Blocks' rename (the
collapsibleThinkingBlocks toggle is now a separate control, so the parent
label stays as 'Reasoning Traces' for clarity)
- Add `collapsibleThinkingBlocks` / `collapsibleThinkingBlocksAria` strings
across all seven supported locales (en, es, ko, pl, pt-BR, uk, zh-CN)
* test(server): add settings-helpers coverage for collapsibleThinkingBlocks
- Verify sanitizeSettingsUpdate accepts boolean true/false and rejects
non-boolean values (string, number)
- Verify formatSettingsResponse forwards the value correctly for both true
and false, and defaults to true when the field is absent
* fix(ui): respect defaultExpanded prop and remove dead alwaysShowActions from ReasoningTimelineBlock
The useEffect on [isStreaming] was firing on mount and immediately calling
setIsExpanded(false) (since isStreaming is false for completed blocks),
overriding any defaultExpanded={true} passed by callers. The fix uses a
prevIsStreamingRef so the effect only collapses the block on a true→false
transition and is a no-op on initial mount.
Also removes alwaysShowActions from ReasoningTimelineBlockProps — the new
header design always shows the chevron, making the prop obsolete. The prop
was already absent from the component destructuring (a dead type entry) and
was silently ignored at runtime. Removed it from ReasoningPartProps,
MergedReasoningPartProps, and the two call-sites in MessageBody as well.
* chore: remove unused reasoningpresentation module and test
* fix(ui): polish collapsible reasoning block UI
---------
Co-authored-by: Bohdan Triapitsyn <artmore@protonmail.com>
- Add local Whisper STT via Transformers.js with Web Worker (no UI freeze)
- Default sttProvider to 'local' in Electron (browser STT unavailable)
- Fix infinite toast loop: stop auto-restart on network errors
- Add retry limit with exponential backoff for transient STT errors
- Append voice transcript to input field (append-inline), not replace
- Add model catalog with download/load button in Voice Settings
* feat: add 'Open files in preview mode' setting
- Add defaultFileViewerPreview setting to persist user preference for file viewer default mode
- Add checkbox in Settings → Sessions → Session Defaults to toggle the setting
- Files now open in preview mode by default when setting is enabled
- Respects per-file-type localStorage persistence for markdown/HTML/JSON files
- Setting persists across sessions via /api/config/settings endpoint
* fix: address PR review feedback
- FilesView: respect HTML localStorage preference in file-change effect,
falling back to global setting only when nothing stored
- DefaultsSettings: read defaultFileViewerPreview directly from config store
instead of redundant local state + separate fetch
- DefaultsSettings: extract duplicated toggle logic into
handleToggleFileViewerPreview callback
* fix: honor default preview setting for markdown files
Previously mdViewMode only read localStorage on mount (deps: []), so
when settingsDefaultFileViewerPreview was enabled and no MD_VIEWER_MODE_KEY
was stored, markdown files silently opened in edit mode — ignoring the
setting for the very file type users most want to preview.
Fold md init into the per-file-change effect, mirroring the html handling:
localStorage preference wins, falling back to the setting-derived default.
The saveMdViewMode callback stays untouched so user-initiated toggles
still persist.
---------
Co-authored-by: Bohdan Triapitsyn <artmore@protonmail.com>
𝖎𝖚𝖑𝖎𝖎𝖆