* feat: tabbed right sidebar, context panel, floating diff comments
* fix: auto-close left sidebar when context panel opens
- Increase default context panel width from 520 to 600 pixels
- Increase sidebar minimum width from 200 to 300 pixels
- Replace collapsible component with custom button in diff view
* refactoring: rework sidebars, tabs, and file tree layout
- Rewrite AnimatedTabs as segment-style with sliding indicator
- Upgrade SidebarFilesTree to match FilesView features (context menus,
git status, file icons, CRUD dialogs, fuzzy search ranking)
- Restructure FilesView header: tabs row + actions row, remove breadcrumbs
- Show relative path in context panel header, track active tab
- Allow left sidebar to stay open alongside context panel
- Hide diff/files tabs from header on desktop (mobile-only)
- Move chevron after group name in session sidebar
- Compact tab heights in right sidebar and git view
- Size PreviewToggleButton to match other action buttons
- Remove directory loading spinner from folder icons
* feat: add project icon and color customization
- Enable users to assign custom icons to projects
- Allow users to choose accent colors for projects
- Stabilize repo status UI during project switching
* feat: add scroll fade indicators to editor tabs
* style: reduce spacing and icon sizes in header
* style: adjust tab component padding from uniform to vertical-horizontal
* feat: Add session state indicators to project tabs
* feat: Enhance session status handling and improve UI responsiveness
* fix: preserve upstream tracking on branch rename
* fix: improve initial remote selection for pull requests
- Uses saved remote name from previous session when available
- Selects remote based on tracking branch when possible
- Falls back to origin or first available remote
* perf(diff): faster highlight, stable stacked scroll
- split/unified Pierre worker pools; prefer shiki-wasm
- align diff CSS line-height; disable scroll anchoring; drop WebKit compositing hacks
- harden stacked pin/align (cancel on user scroll/input); prevent overscroll
- make overlay scrollbar MutationObserver optional; disable for diff container
* feat: handle binary files in diff view
* fix: adjust project tabs layout and drag regions
* style: update drag overlay visual styling
* feat: enable number keys to switch projects in the sidebar
* fix: recognize octet-stream as text-based MIME type
* feat: add keyboard navigation to context panel
* feat: add session pinning to sidebar
- Pin important sessions to keep them at the top
- Pinned sessions persist across browser sessions
* refactor: move context usage display from chat input to header
Save and restore main-window bounds/state in desktop settings so reopen behavior is consistent, while keeping new windows on defaults. Add debounce, off-screen fallback, and minimum-size guards to prevent unusable or stale geometry restores.
* refactor(quota): migrate server quota providers to modular registry architecture
* docs: add quota module documentation
Explain the quota module purpose and runtime signals.
List entrypoints, provider registry, and response shape.
Outline steps to add a new provider and testing commands.
The manifest was locked to portrait-primary, preventing landscape mode
on tablets even with device rotation enabled. Changed to "any" to follow
system rotation settings.
* feat: add NanoGPT quota provider and usage fetch
Detect NanoGPT as a configured quota provider.
Expose daily and monthly usage windows with reset times.
Return configured and ok state based on API key presence.
* feat: fetch NanoGPT quota and detect config
Detect NanoGPT in configured providers from auth.
Fetch and expose daily and monthly usage windows for NanoGPT.
* feat: add NanoGPT quota provider option
* feat: add compact price display and capability slides in ModelControls
Show compact price text on desktop in ModelControls.
Render slides for price and capability icons in a rotating metadata section.
Rotate metadata only on non-VSCode runtimes.
* feat: prefer capabilities over price in static runtimes for ModelControls
* feat: add worktree validation and deleteLocalBranch option
Add API to validate and create worktrees with new payload types
Allow deleting local branches when removing worktrees via UI and API
Introduce OpenCode style random names for worktrees when not provided
* feat: enable SSH/HTTPS transport detection for PR picker
Load remotes for the current project directory to inform PR picker options.
Determine preferred push transport from remotes and apply it.
Expose sshUrl in API for frontend to build SSH clone URLs
* feat: extend head repo with sshUrl and improve push error messages
Add sshUrl field to head repo mapping
Enhance push failure handling to display stderr or stdout details
Return push details on success
* fix: worktree path
* feat: worktree set upstream on creation
Enable pushing to upstream by default when no remote is specified
Remove per-remote dropdown for push actions and auto-use first/upstream remote
Update server and VSCode git services to support push without explicit remote and set upstream
* fix: worktree-name sanitization
* feat: rename worktree path field and branch prefix
* feat(worktrees): add git.worktree facade, validation endpoint, upstream/remote-aware creation, and non-blocking setup execution
* refactor(git): use git.worktree namespace in branch picker
* feat(worktrees): sync OpenCode sandbox metadata on create/remove
* fix(worktrees): accept new path key in workspace guard and validate remote startRef
* chore(docs): remove temporary worktree testing plan
* feat: add git worktree management API (list/create/delete/validate) for vscode
* feat: wire root tracking remote and defaults for new worktrees
Add resolveRootTrackingRemote to detect upstream remote for root branch
Apply upstream defaults when creating new worktrees to auto-set upstream
Replace validation and creation flow to use new worktreeCreate APIs
* feat(worktrees): enable root tracking remote handling
* perf: implement virtualized rendering for diff viewer
- Enables efficient rendering of large diffs by only rendering visible content
- Uses shared virtualizer cache to optimize memory across multiple diff viewers
- Configures virtual scrolling with 24px line height for consistent layout
* fix: improve diff viewer line selection and annotation rendering
* fix: normalize line ranges to prevent selection bugs
* chore: upgrade @opencode-ai/sdk to v1.1.65
* feat: message list refactoring and optimizations
* refactor: collapse 3 memory-limit settings into single messageLimit
- Replace memoryLimitHistorical/Viewport/ActiveSession with one `messageLimit` (default 200)
- Background trim derived automatically (limit * 0.6), not user-facing
- Fix stale historyLimit overriding current messageLimit on session reload
- Remove streaming cap that reduced fetch to VIEWPORT_MESSAGES during active streams
- Deduplicate detectTurns() call in TurnGroupingContext (reuse staticValue.turns)
- Zustand v3 migration: collapse old fields, preserve user-customized values
- Server: sanitize single messageLimit field (10-500)
Add probeExternalOpenCode() to detect a running OpenCode instance via
GET /global/health before attempting to spawn a new one. When an external
server is detected (or OPENCODE_SKIP_START is set), the isExternalOpenCode
flag prevents gracefulShutdown() and restartOpenCode() from killing the
user's independent OpenCode process.
- Probe configured port, then default 4096 on startup
- Use /global/health + body.healthy === true for reliable detection
- Skip kill -9 on shutdown/restart when server is externally owned
- Re-probe (instead of kill + respawn) on restart for external servers
Improve panel and search UI colors to match theme
Improve focus and hover states for search input and buttons
Highlight search matches with distinct info and selection styling
* hive(01-add-mobile-autocorrect-to-commit-message-input): Added mobile autocorrect to commit message input.
* hive(02-add-mobile-autocorrect-to-pr-title-inputs): Added mobile autocorrect to PR title inputs in Pul
* hive(03-add-mobile-autocorrect-to-pr-description-textareas): Added mobile autocorrect to PR description textare
* Fix chat last message overlap
* feat: enable local-origin desktop features and copy in terminal support
- Enable arbitrary web loads in desktop web content for TAURI apps.
- Guard directory dialog access behind local-origin origin check.
- Add copy-to-clipboard support for terminal selections via common shortcuts.
* fix: improve initial directory resolution and persistence in directory store
* feat(BottomTerminalDock): add close button next to the fullscreen toggle in the dock
* style: replace hardcoded gradient with theme value in shine text variant
* fix(header): adapt instance button for desktop only
* refactor(chat): polish sticky turn UX and message action rows for better readability
Switch to stable sticky-only turn behavior and redesign user/assistant action controls (placement, hover rules, ordering, spacing, selection-safe clamp) to reduce visual noise and improve interaction flow.
* feat(chat/message): refactor buttons in messages footer
* feat: enhance image preview functionality in chat messages
- Added a new ImagePreviewDialog component to handle image previews with navigation support.
- Updated ToolOutputDialog to utilize the new ImagePreviewDialog for displaying images.
- Modified the ToolPopupContent type to include a gallery of images and an index for the current image.
- Removed the old inline image display logic from ToolOutputDialog.
- Improved file handling in the file store, including better MIME type guessing and handling of server paths.
- Introduced a new API endpoint for handling large session message payloads, allowing for better management of multi-file attachments.
- Updated the VSCode bridge to support session message requests with appropriate headers and body handling.
* feat(proxy): implement SSE forwarding and enhance generic API request handling
* feat(chat): support submitting only queued messages
* feat: add image preview transition state
* fix: default VSCode view to draft and fixed sessions list regression
* feat(macos): add New Window menu item
Add a New Window option to the macOS menu bar
Trigger opening a separate window when the New Window action is selected
* feat: enable multiple desktop windows and new window shortcut
Enable multiple desktop windows by extending manifest to main-*
Expose desktop_new_window to create additional windows from UI
Show New Window shortcut in Help dialog (Desktop only) with Shift + Alt + <mod> + N
* feat(desktop): enable opening host URL in new window from host switcher
Add a new action to launch a URL in a separate window from the host switcher
Wire the UI to open the selected host in a new window using a dedicated control
Expose a cross-platform API to create a new window pointing at a URL
* fix(desktop): validate URL scheme before opening new window
* fix(desktop): skip unreachable hosts when opening new windows
* feat(voice): add voice input/output support with multiple providers
- Add BrowserVoiceButton component for Web Speech API voice input
- Add VoiceProvider context for managing voice state across the app
- Add TTS (Text-to-Speech) support with browser, macOS Say, and OpenAI providers
- Add message TTS buttons to read assistant messages aloud
- Add VoiceSettings page in OpenChamber settings
- Add server endpoints for TTS and summarization services
- Include slider component for voice rate/pitch/volume controls
- Add hidden session support for background voice operations
- Add Caddyfile for HTTPS support (required for microphone access)
* fix: Build errors fixed and removed outdated ElevenLabs test code.
* refactor(voice): use zen API with gpt-5-nano for TTS summarization
Replace the hidden session + OpenCode SDK approach with direct calls
to the opencode.ai zen API (same pattern used for commit message and
PR description generation).
- Rewrite summarization-service.js to call zen/v1/responses with gpt-5-nano
- Remove hidden session logic (hiddenSession.ts, sessionStore filtering)
- Remove summarizeModel setting and model selector from VoiceSettings
- Simplify client-side summarize.ts to no longer pass model params
- Clean up callers in useMessageTTS and useBrowserVoice
* fix(voice): remove false 'voice not supported' warning in settings
Mobile Safari does support voice but the isSupported check was
incorrectly flagging it. Remove the warning banner entirely.
* feat(voice): add configurable summary length limit for TTS output
Add a slider (50-2000 chars) in voice settings to control max summary
length. The limit is passed through the summarize endpoint and speak
endpoint to the zen API prompt, with token budget scaled accordingly.
* fix(voice): add diagnostic logging and sanitize TTS fallback
Add console logging throughout the summarization flow (client + server)
to trace why text may not be summarized. Fix silent error swallowing in
/api/tts/speak. Always apply sanitizeForTTS even when summarization is
disabled so raw markdown/code is never spoken verbatim.
* fix(voice): fix token budget starving model of output tokens
max_output_tokens includes both reasoning and output tokens. With
effort:'low', reasoning alone consumes ~128 tokens, so a budget of
100 left zero tokens for the actual summary text. Use a fixed 1000
token budget (matching commit message generation) and control output
length via the prompt's character limit instruction instead.
* chore(voice): remove diagnostic logging from summarization flow
* fix(voice): don't request mic permission on mobile page load
Remove the useEffect that pre-requested microphone permission when the
BrowserVoiceButton component mounted on mobile. This caused an unwanted
permission prompt immediately on page load before the user tapped the
mic icon. Permission is now only requested on explicit user interaction.
* fix(voice): remove unused BrowserVoiceButton binding
* fix(voice): desktop mic flow + non-continuous draft mode
* fix(voice): stabilize continuous loop and polish controls
* feat(settings): mark voice section experimental
---------
Co-authored-by: Bohdan Triapitsyn <artmore@protonmail.com>
Nelson Pires