Craft Agents: baixar versions anteriors

Features
- Craft Agents has a new home: thecraftagents.com — the app, downloads, auto-updates, shared session links, and documentation now live on the new domain. Updating to this version switches your app to the new update feed automatically; existing share links and sign-in flows keep working during the transition
- Documentation rebuilt as a fast static site — the docs at thecraftagents.com/docs are now a fully static site with instant search, same content and URLs as before
Improvements
- Agents now reference the product documentation via the public docs site instead of a built-in documentation source — one less background connection per session

Features
- Claude Opus 4.6 is back as a selectable model — restored by popular demand (craft-agents-oss#1007). It appears again in the model pickers for direct Anthropic connections, Amazon Bedrock, and Pi-backed catalogs. Connections that previously had it force-migrated to Opus 4.8 get Opus 4.6 re-added automatically (one-shot — your default model is left untouched, and removing it again sticks). Opus 4.8 remains the default for new installs
Bug Fixes
- Explore mode no longer goes silent when a tool call is blocked — since the v0.11.3 SDK uplift, the first blocked tool ended the turn before the agent could respond. The agent now sees the block reason again and can suggest switching modes or submit a plan instead of stopping without a word

Features
- New archive_session session tool — agents can archive (or unarchive) another session in the workspace by ID to tidy up finished or superseded sessions. Archiving only hides the session from the active list and unread counts, never deletes it; a session cannot archive itself, and archiving is refused while the target is mid-turn
Improvements
- Claude Agent SDK uplifted to 0.3.220 — the bundled SDK moves from 0.3.197 to 0.3.220 (Claude Code v2.1.220 parity). Pulls in upstream fixes for an abort-listener leak in streaming queries, richer process-exit diagnostics (CLI stderr now included in errors), correct HTTP status reporting for rate-limit errors, and session resume with dash-leading IDs. Note: upstream now caps subagent nesting depth at 1 by default (was 5) — override with CLAUDE_CODE_MAX_SUBAGENT_SPAWN_DEPTH if you rely on agents spawning nested agents
- If installing a downloaded update fails at the final handoff, the app now restarts cleanly instead of continuing in a half-shut-down state where later work could be lost
Bug Fixes
- macOS auto-update installs again — updates no longer end up downloaded-but-never-installed because quit cleanup interrupted the installer handoff
- The chat input no longer force-capitalizes the first letter, which broke pinyin and other CJK IME composition
- Local file links whose paths contain spaces (%20) open correctly again from chat messages
- New sessions no longer inherit an excluded filter — a lone "exclude Done" filter used to create new sessions as Done; only include-mode filters are inherited now
- Custom OpenAI-compatible endpoints validate again: relays that send an empty tool_calls: [] field no longer break stream completion
- Re-authenticating a ChatGPT or GitHub Copilot connection now refreshes its model list, so entitlement changes show up without a restart
- MCP tools with dots in their names (e.g. pat.batch_plan) work on OpenAI/Codex-backed connections
- Windows: image previews no longer fail by doubling the workspace path in front of absolute C:\ paths
- Windows: a gitBashPath configured in config.json is honored by the agent's Bash tool, fixing "No bash shell found" for per-user Git installs
- A stale server lock whose process ID was recycled by the OS no longer prevents the app from starting — the lock now records which executable wrote it, and startup verifies the live process actually matches

Features
- Claude- and Pi-backed agents can create unstarted Tasks on the board with the new create_task tool — including title, description, acceptance criteria, sources, skills, LLM connection/model, and working directory; Tasks inherit the current project unless another project is specified, and unknown source or skill references are reported as warnings instead of blocking creation
- Sessions can now be sent between any two workspaces, including remote-to-local and local-to-local transfers
Improvements
- Press Up arrow in a truly empty chat input to cancel the running turn and bring your last prompt back for editing
- In-app "background session finished" alerts are now off by default — enable them under Settings → Appearance
- Background task chips can be dismissed without stopping the task, and a chip with no updates stops spinning and shows an Unknown state without claiming the task stopped
Bug Fixes
- Large transfers involving a remote workspace now allow more time for bundle export and import, reducing timeouts on slower connections
- Queuing a message while the agent is still replying no longer pushes the in-progress response below your new message
- Queuing a message mid-turn no longer makes the agent treat its already-finished previous response as interrupted

Features
- Support for OpenAI GPT-5.6 (Luna, Terra, Sol) via OpenAI API key, ChatGPT account, and Azure OpenAI connections
- The Max thinking level is now sent natively to models that support it (GPT-5.6, adaptive Claude models) instead of capping at XHigh
Improvements
- New OpenAI connections default to GPT-5.6 Sol
- Updated Pi SDK to 0.80.6: request-wide input-token pricing tiers for accurate long-context cost accounting on GPT-5.4/5.5/5.6

v0.11.0 — Projects & Kanban Task Board (Beta)
The biggest release in a while: workspaces gain Projects, and work gets a visual home on a Kanban task board backed by durable Tasks that can fan out into orchestrated multi-session runs.
⭐ Projects
Workspaces now have first-class Projects: named collections of sessions, assets, and context stored at {workspaceRootPath}/projects/{slug}/. Bind a session to a project and it inherits the project's working directory and gets a <project_context> block injected into its system prompt — including the project's asset manifest and a size-capped MEMORY.md — so the agent starts every conversation already knowing what the project is about.
- Projects panel & info page — a ProjectsListPanel in the navigator and a ProjectInfoPage with Sessions / Assets / Settings tabs; assign sessions from the session menu. (f2eb9542, 95bc761e, 423ff362)
- Project-aware session list — group and filter your sessions by project. (0ad7189e)
- Per-project accent color — with an appearance-treatment toggle; workspaces got avatars and a rebuilt color picker along the way. (2b2d64bd, 481851f8)
- Live binding + context injection — binding events update the UI live; asset manifest and project memory flow into every bound session. (7d540944, 742a2761)
- Fully localized in all 7 languages. (ada9ded1)
⭐ Kanban Task Board — Beta
A new board view turns a workspace's sessions and tasks into draggable cards over per-project, fully editable columns. Under it sits a durable Tasks system: tasks live independently of chat, can be edited, re-run, and — for bigger jobs — decomposed into dependency DAGs that a Conductor orchestrator executes over real child sessions.
- The board — add, rename, recolor, and reorder columns per project; columns carry an optional drop status (settable from the board header), so moving a card can transition the underlying session's status. Cards keep stable ordering, auto-move columns as states change, and never auto-close — closing a card stays your decision. (e6f5ea5b, 5ae740b7, 82a661f4, fd9c0f04, 854a9860, aa9d581e)
- Tasks — a dedicated task editor with create/edit modes, a results tab, and live session binding; per-task labels; acceptance criteria; create-only vs. create-then-run flows; filter-aware "New Task"; run-from-tile; inherited project working directories. (9a608cc3, 7c7eb49c, 476a0c21, 36b978f0)
- Subtasks & DAGs (Conductor) — decompose a task into a task.yaml DAG; the Conductor spawns and supervises a child session per node with dependency-ordered execution, durable resume after interruption, and union subtask progress on the tile. Subtask nodes render colored, localized state pills. (ea814eba, 5702ea6b, 36b978f0, b03a9357, 937f0ff3)
- Repair loop — when a task's output doesn't meet its acceptance criteria, a repair pass re-checks and dispatches fixes. (36b978f0)
Beta notice: the Kanban board and Tasks (including Conductor DAG runs) ship as beta. Expect rough edges — board behaviors and the task spec format may still change between releases. Please report issues and ideas on GitHub.
Features
- Background agents stay alive across turns — background subagents are no longer killed when the turn that spawned them ends (previously surfacing as "stopped by user"). A persistent query keeps the session alive; when an agent finishes while the session is idle, the session wakes itself and presents the result. On by default — set CRAFT_KEEP_BG_AGENTS_ALIVE=0 to opt out. Running work shows as chips above the input (including live Workflow fan-out counts) with a floating completion chip when results land. Fixes #953. (32072bca, c0a726a6, c4201dd0, 5e7490ae)
Improvements
- Pi SDK 0.80.3 — removes a hardcoded 20s SSE timeout that cut off long Codex responses on OpenAI-protocol connections. (c9e456bd)
- UI fully localized — every hardcoded English string flagged by the i18n lint is gone (aria-labels, drawer titles, menus, crash screen, code-block chrome); the permission-mode selector and the standalone br

v0.10.5 — Claude Sonnet 5
Features
- Claude Sonnet 5 now available — Anthropic's newest Sonnet-class model (claude-sonnet-5, released 2026-06-30) is now selectable in the model picker, with a 1M-token context window and adaptive thinking. It joins the lineup as the primary Sonnet, with Sonnet 4.6 kept available as the previous generation. Bedrock routing (US / EU / Global inference profiles) and connection defaults are wired up so it works across both Anthropic and Amazon Bedrock connections. The default model is unchanged (Opus 4.8). (5bf361b6)
Improvements
- Claude Agent SDK uplifted to 0.3.197 — the bundled SDK moves from 0.3.170 to 0.3.197 (Claude Code v2.1.197 parity), so the bundled CLI is itself aware of Sonnet 5 (the sonnet alias, context-window, and effort defaults). Includes upstream fixes since 0.3.170, including a Windows CLI subprocess console-flash fix. Full typecheck and shared test suites pass clean. (5bf361b6)
Bug Fixes
- None.
Breaking Changes
- None.

v0.10.4 — GLM-5 family on z.ai + Pi SDK uplift
Features
- GLM-5.2 and the GLM-5 family now available on z.ai — the Pi SDK uplift regenerates the model catalog from models.dev, so z.ai's GLM-5.2, GLM-5.1, and GLM-5-turbo surface automatically on the existing z.ai provider, with correct thinking / reasoning_effort handling (fixed upstream). Previously GLM-5.2 was only reachable through a hand-rolled OpenAI-compatible endpoint where reasoning effort misbehaved. (694d4b1c)
- Automatic config.json backups on startup — Craft Agent now snapshots config.json on every startup and keeps the newest three backups, so a corrupted or bad config write can be recovered. (a5e62bc8)
Improvements
- Pi SDK uplifted to 0.79.9 (scope migration to @earendil-works) — the three Pi packages (pi-ai, pi-coding-agent, pi-agent-core) move from the now-frozen @mariozechner/* scope at 0.73.1 to the rebranded @earendil-works/* scope at 0.79.9, across all manifests and the lockfile. Newer models for existing Pi providers (OpenRouter, z.ai, and others) appear automatically via the regenerated models.dev catalog. GitHub Copilot device-code login now consumes the SDK's structured onDeviceCode callback in place of a fragile free-text regex. No user-facing API breakage; the full typecheck and targeted Pi test suites pass clean. (694d4b1c)
Bug Fixes
- Session titles now follow your chosen language — AI-generated titles read your Appearance → Language setting directly from disk instead of the app's in-memory locale, which could still be on its English fallback when the first title generated. If you haven't picked a language, titles now auto-detect the language you're writing in rather than defaulting to English. Fixes #885. (3ed130c0)
- Auto-update now keeps a diagnostic log — the update lifecycle is recorded to an always-on log at ~/.craft-agent/logs/auto-update.log, even in release builds where the normal debug log is off. This makes "the update downloaded but never installed" reports diagnosable. Partially addresses #891 (diagnostics only; the ShipIt install hand-off fix is a follow-up). (0e2005ca)
- Config backups no longer overwrite an existing same-day backup — the earliest good snapshot from each day is preserved instead of being clobbered by a later same-day startup. (42c79906)
Breaking Changes
- None.

v0.10.3 — Claude Fable 5
Features
- Claude Fable 5 is here — Anthropic's most capable widely released model (GA 2026-06-09) is now available in Craft Agent on the Claude Agent SDK path, with a full 1M-token context window. It works across direct Anthropic connections and AWS Bedrock (us/eu/global inference-profile variants), with the model description localized in all 7 languages. Opus 4.8 remains the default — Fable is offered alongside it, not in front of it — and it is added to the Pi → Anthropic auth-bridge preferences below Opus. Because Fable 5 (and the Mythos 5 class) run with adaptive thinking always on and reject an explicit disabled thinking option, the thinking resolver now maps the "off" and minimize-thinking cases to low-effort adaptive thinking for these models; Opus, Sonnet, and Haiku behavior is byte-for-byte unchanged. (22c7ae99)
Improvements
- Claude Agent SDK uplifted to 0.3.170 — @anthropic-ai/claude-agent-sdk moves from 0.3.154 → 0.3.170 (root plus the core and shared peer dependencies, lockfile refreshed). No API breakage; the full typecheck passes clean. (22c7ae99)
Bug Fixes
- None.
Breaking Changes
- None.

v0.10.2 — Link labels, Anthropic account visibility, and prompt-cache fixes
Features
- Link label value type — Labels can now carry a link value alongside the existing string/number/date types. A link value renders as a clickable chip (the URL scheme is stripped for display) that opens in your external browser, with a keyboard-accessible "Open link" action in the value popover. The type is threaded through the zod schema, CLI, and agent-prompt layers, and the clickable value appears consistently in the session list, chat display, and chat-input badges. Documented in labels.md with validate/format/resolve tests. (8aae50ad)
- Resolved Anthropic account & org shown per OAuth connection — Settings → LLM Connections now displays the real Anthropic identity ("email · org") each Claude OAuth grant resolves to, captured from the token-exchange response and persisted on the connection. When two connections in the same workspace resolve to the same Anthropic account (shared quota — a previously invisible mis-route), an amber warning is surfaced. This also fixes a load-bearing bug where updateLlmConnection rebuilt connections from a hardcoded field allowlist and silently dropped any unlisted field on every save; the identity fields are now preserved with a round-trip test. Fixes #838. (a4b8868e)
- Last sent message restored to the input on Stop — Clicking Stop (an explicit cancel, not a redirect or new send) now copies your last sent message back into the chat input so you can tweak and resend it. It is append-safe: a draft typed while the agent was running is preserved, and the restored text is appended below it. Queued, never-sent messages continue to be handled by the existing restore path. (cfbe802b)
Improvements
- None.
Bug Fixes
- Pi prompt-cache no longer busts every turn — The Pi adapter folded volatile context (minute-precision date/time, session_state, source state) into the system prompt, which pi-ai caches as the prefix ahead of all history — so re-stamping it each turn dropped cacheRead to 0 every turn (notably with non-Claude models such as DeepSeek V4). PromptBuilder.buildContextParts is now split into buildVolatileContextParts + buildStableContextParts; the Pi path keeps only stable blocks (workspace capabilities, working directory) in the cached prefix and routes volatile blocks to the user-message tail, mirroring what the Claude path already did. The Claude path stays byte-identical. Fixes #862. (80b478e3)
- Accept-Plan chevron rotates when the dropdown opens — Both Accept-Plan triggers render via Radix asChild, so data-state="open" landed on the host <button>, never on the nested chevron <svg> — the self-scoped data-[state=open]:rotate-180 never fired. The rotation is now scoped to a named group on the button (group/accept + group-data-[state=open]/accept:rotate-180). Fixes #840. (f97f483b)
Breaking Changes
- None.

v0.10.1 — Claude Opus 4.8 default, localized session titles, and Apple Silicon–only macOS builds
Features
- None.
Improvements
- Claude Opus 4.8 is now the default Opus model — Upgraded the Claude Agent SDK to 0.3.154 and set Claude Opus 4.8 as the default Anthropic Opus model, while keeping Opus 4.7 selectable. The deprecated Opus 4.6 is removed from the model pickers. Existing direct-Anthropic Opus 4.7 defaults move to Opus 4.8, and Opus 4.5/4.6 selections migrate to the best available current Opus model. Bedrock mappings, Pi fallback handling, model-migration tests, docs, and UI examples were updated to match. (0a1357b2)
Bug Fixes
- Session titles honour the Settings → Appearance language — The main-process i18n instance has no language-detection plugin (there is no localStorage in Node) and was never re-hydrated from disk, so it always sat at the en fallback after restart. Title generation (generateTitle / refreshTitle) and the system prompt's "Preferred language" line read that main-process language, so they came back in English even when Appearance was set to another language. The chosen UI language is now persisted to an internal, validated uiLanguage field and the main-process i18n is hydrated from it on startup; the renderer also pushes its resolved language to main once on upgrade so the value is learned without re-picking. Partially addresses #815 and #738. (4a743ef2)
- Text-selection highlight stays aligned when scrolling a preview — Selecting text inside a scrollable markdown-preview block (the max-h-[400px] overflow-auto wrapper) and adding a follow-up left the yellow range highlight and its numbered chip anchored to the original coordinates, so they drifted away from the text as the block scrolled. A capture-phase scroll listener now recomputes the overlay geometry on any nested scroll — a coords-only fast path coalesced with requestAnimationFrame, with a no-op short-circuit when there are no annotations — keeping the highlight pinned to its text. The same fix is mirrored in the viewer's annotation path. (14bba50c)
Breaking Changes
- macOS Intel (x64) builds discontinued — Apple Silicon only — Craft Agents for macOS now ships an Apple Silicon (arm64) build only; Intel-based Macs are no longer supported. The x64 .dmg is no longer produced (v0.10.0 was the last release with an Intel build). Apple Silicon (M-series) Macs, Windows, and Linux are unaffected.
- Legacy language preference field removed — The free-text, user-editable language field in preferences.json is replaced by an internal, validated uiLanguage field driven by Settings → Appearance. Existing config files are tolerated (the schema is passthrough) and the stale language key is scrubbed on read, so no user action is required. The update_user_preferences tool no longer accepts a language argument.

v0.10.0 — Remote browser_tool bridging, per-workspace browser tab isolation, and #824 basic-auth fix
Features
- Remote browser_tool bridged into the user's local Electron browser — Agents running on a remote workspace (headless server, docker, WebUI) can now drive the user's local desktop BrowserPaneManager end-to-end. Adds a client:browser:invoke WS capability (advertised on handshake, server invokes client, plain Error with .code preserved through both directions), mirroring how shell.openExternal already works for OPEN_URL. Transport gains hasClientCapability / findClientsWithCapability for routing; Electron gets a new __browser:invoke IPC dispatcher with per-method owner-key authorization, no-manual-window-reuse for remote callers, session-scoped listInstances, and screenshot Buffer↔Uint8Array conversion across the wire. server-core gets a RemoteBrowserPaneManager (session-bound IBPM impl) and SessionManager.getBrowserPaneManagerForSession with capability-aware host-client fallback and per-session pin cleanup on disconnect. uploadFile is blocked over the bridge; evaluate is gated by a local allowRemoteEvaluate setting. The Pi runtime learns friendly error mappings for BROWSER_NO_CAPABLE_CLIENT, CAPABILITY_UNAVAILABLE, CLIENT_DISCONNECTED, CLIENT_REQUEST_TIMEOUT, BROWSER_INSTANCE_NOT_OWNED, BROWSER_REMOTE_UPLOAD_NOT_SUPPORTED, and BROWSER_REMOTE_EVALUATE_BLOCKED, and now mirrors Claude's getBrowserToolEnabled gate so Pi no longer advertises browser_tool when the toggle is off. 27 new tests cover wire packaging, per-method authz, capability introspection, host-client fallback, screenshot round-trip, error-code preservation, and the Pi error-mapping contract. (1d926c33)
- Browser tabs isolated per workspace — BrowserPaneManager is process-global and STATE_CHANGED used to broadcast { to: 'all' }, so a chat in workspace A saw browser tabs and status banners owned by sessions in workspace B. Every BrowserInstance (and the BrowserInstanceInfo DTO) now carries a nullable workspaceId; STATE_CHANGED routes to { to: 'workspace', workspaceId } when set (falling back to { to: 'all' } for unbound manual windows); the browserPane.LIST handler filters by ctx.workspaceId; and the renderer reads a new browserInstancesForWorkspaceAtomFamily keyed by activeWorkspaceId. Windows still run in parallel as real BrowserWindows — this is a UI visibility filter, not a sandbox. REMOVED / INTERACTED stay broadcast-to-all (id-only payloads, harmless no-op on workspaces that never saw the entry). workspaceId ships optional on the DTO, so old renderers tolerate missing values (treats undefined as null → passes the filter, equivalent to today's behavior). 17 new tests across atom filter, BPM stamping, and broadcast/LIST routing. (af817192)
Improvements
- markdown-preview block documented in the rich-output reference — The markdown-preview block (shipped in v0.9.6) is now covered in apps/online-docs/go-further/rich-output.mdx alongside the existing html-preview / pdf-preview / image-preview entries, so users discovering the block in-chat can find usage examples and the src / items field reference in the docs. (2d9693b1, 70c2955f)
- Browser-bridge wiring is observable from server logs — Three sessionLog.info lines in SessionManager now confirm whether setRpcServer ran at bootstrap and whether the browser-pane-forwarding block executed at agent init. Makes it possible to diagnose remote workspaces that still hit "Browser window controls are not available" from server logs alone, without attaching a debugger. (ad26e61d)
Bug Fixes
- Renderer accepts both local AND remote workspace ids when filtering tabs — When connected to a remote workspace, a renderer has two relevant workspace ids: activeWorkspaceId (the LOCAL Craft Agents window's identity, used for locally-opened manual tabs) and activeWorkspace.remoteServer.remoteWorkspaceId (the REMOTE server's id, used by the remote agent when it stamps tabs through the WS bridge). The first iteration of the workspace-

v0.9.6 — Auto-update window restoration, mid-session credential refresh, and #807/#798/#804 fixes
Features
- Workspace name in window title when multiple windows are open — With one window open the title stays as the app name ("Craft Agents"). Open a second window and each window's title becomes the workspace name it belongs to, so windows are trivial to tell apart in Cmd-Tab, Mission Control, and the Windows taskbar. The renderer's static <title> is suppressed so it can't clobber the main-process title. (5a49b6ca)
- Inline markdown-preview block for rendering .md files — A new markdown-preview code-block type mirrors html-preview / pdf-preview / image-preview: reference a .md file by absolute path and the chat renders it through the shared Markdown component, with a disablePreviewBlocks guard that prevents markdown-preview-inside-markdown-preview recursion without disabling other nested preview blocks. Registered in both minimal (assistant chat) and full modes; supports multi-item items arrays. Partially addresses #807. (45760a6f)
Improvements
- Online-docs coverage for multi-window titles and auto-update restoration — apps/online-docs/go-further/workspaces.mdx now documents the new window-title behavior and the auto-update window-state preservation, so users discovering these features in-app can find rationale and edge-case notes in the docs. (755a8b77)
- Online docs introduction polish — Tightened wording in the getting-started introduction (Mintlify dashboard edit). (3a5378db)
- Messaging gateway docs reflect the 0.9.5 fallback fix — Clarifies that progress and final_only modes now both fall back to the most recent assistant text when a run ends on a tool call without a non-intermediate text_complete, instead of leaving a thinking bubble (progress) or staying silent (final_only). Genuinely empty runs are still silent. (0cfd1ffb)
Bug Fixes
- Multi-window state survives auto-update — electron-updater (Squirrel.Mac) destroys all BrowserWindows between quitAndInstall and before-quit firing, so the existing window-state save ran with an empty snapshot and clobbered ~/.craft-agent/window-state.json with { windows: [] }. Users lost their multi-window setup every time they accepted an update. installUpdate now fires a setBeforeUpdateQuitHook callback that captures and saves window state while windows still exist, and the late before-quit path adds an empty-snapshot guard so the pre-update save can't be overwritten. (3db842e0)
- API source credentials refresh mid-session — Sources with bearer/header/query/basic auth captured the credential as a static string at tool-creation time. After refreshing an expired token via source_credential_prompt, the in-process tool kept sending the stale value (401 until full session restart) even though source_test confirmed the new token worked. Non-OAuth API sources now route through a credential getter that reads the vault on every call, mirroring the existing OAuth / renew-endpoint path. OAuth and renew-endpoint sources are untouched — they already have refresh via TokenRefreshManager. (5b6a0588)
- Stale source_apikey credential no longer leaks when flipping authType to 'none' — SourceCredentialManager.getCredentialId() maps 'none', 'header', and 'query' to the same source_apikey slot. Flipping a source from a credential-bearing authType to 'none' left the stored credential addressable under that slot, where it could later override defaultHeaders on a rebuild — the bare credential value got sent as a Cookie header instead of the new defaultHeaders.Cookie value. saveSourceConfig now best-effort-deletes the source_apikey slot when the new config is an API source with authType:'none'. Cleanup never throws and never blocks the config write. (d0c70f23)
- Blocked URL schemes now explain why + DOM hrefs are sanitized — When react-markdown's defaultUrlTransform stripped a file: / javascript: URL to empty, the anchor handler fell back to anchor text and new URL(text) rejected it as "Invalid URL" — a gene

v0.9.5 — Compact-mode UX polish and MCP / branching stability
Features
- Compact-mode drawer for the session row menu — Session row context menus on compact / mobile layouts now open as a vaul drawer instead of a clipped Radix popover, matching the drawer pattern already used for status, labels, share, and messaging actions. (a0468049)
- Drawer-based working-directory selector in compact mode — The working-directory picker switches to a drawer on narrow widths so the path list, recent dirs, and "Choose folder…" action remain reachable without horizontal scroll. The desktop dropdown is unchanged. (6e147c03)
- Expandable chat input on compact layouts — Users can now tap to expand a collapsed chat input on compact / mobile views so longer prompts are easier to compose without leaving the active session. (15fe589b)
- Drawer-based AcceptPlan picker in compact mode — The plan-acceptance UI uses a drawer in compact / mobile layouts so action rows aren't pushed off-screen on narrow widths. (f8e92a11)
- Compact model selector in the web UI's mobile / auto-compact view — The web UI's model picker collapses into a compact selector when the shell width crosses the mobile breakpoint, matching the Electron app's container-query behavior. (75c7825d)
Improvements
- Shared useWorkingDirectoryState hook — Path resolution, recent dirs, and folder-picker handling are extracted into a single hook so the desktop dropdown and compact drawer can't drift in behavior. (aa01895d)
- Windows RTK install docs — Clarifies the Windows install path in the RTK setup docs so the "RTK not found" detection in Settings → AI → Performance matches what users see on Windows. (129c28ff)
Bug Fixes
- Branching on the latest turn no longer drops the last assistant message — The Pi-backed branching path now captures the conversation anchor after the SDK appends the new entry, so branching off the most recent turn no longer truncates the message that triggered the branch. Fixes #782. (0988d0d6)
- Stdio MCP source_test gives real diagnostics instead of a fake timeout — source_test for stdio MCP servers now runs a single process with a proper stderr-activity watchdog and surfaces actual startup output, replacing the false "Server startup timeout" that fired for any slow-but-healthy server. Fixes #787. (a7fa0a3a, 69d74167)
- Parallel source_test calls no longer wedge a session — When a source-activation abort interrupts a turn, sibling tool_result blocks are now drained before the abort propagates so the SDK doesn't end up with orphaned tool_use IDs that permanently jam the next turn. Fixes #790. (58087fbf)
- AcceptPlan dropdown positioning and compact visibility — The desktop AcceptPlan dropdown is now built on Radix DropdownMenu, fixing anchor drift and viewport clipping that could push actions off-screen. The web UI also now actually renders the dropdown in compact / mobile views — previously hidden behind a wider-only branch. (01d062e3, d32f390d)
- Chat view no longer stuck on "Thinking…" when a turn ends on a tool call — Mirrors the messaging-gateway fix on the Electron main UI side. When the Pi agent emits intermediate text plus a tool call and then completes without a non-intermediate text_complete, groupMessagesByTurn now uses the session's isProcessing=false signal to mark the open turn complete and promote the intermediate text as the response. Pre-fix the chat sat on "Thinking…" forever and the user had to send a follow-up message to unstick it.
- Messaging gateway delivers the final assistant message in progress / final_only modes — Automations whose last action is a tool call (e.g. the agent uses a tool to send the Telegram message and never emits a clean non-intermediate text_complete) used to leave progress frozen on 💭 thinking… and final_only silent. The renderer now tracks the most recent assistant text regardless of isIntermediate and falls back to it on complete when no clean final arrived; genuinely empty runs still leave the status label in place. Thanks

v0.9.4 — RTK token optimization, compact-session fixes, Codex transport stability
Features
- Optional RTK Bash token compression — Adds opt-in support for RTK in Settings → AI → Performance. Craft Agent still shows and permission-checks the original Bash command, but when RTK is installed (>=0.23.0) and enabled, the execution path is rewritten through rtk so common development-command output is compressed before it reaches the model. The settings UI now detects missing/outdated installs, offers Get RTK + Re-check actions, and shows saved-token / efficiency stats once RTK has processed commands. (754d254c, a96b8706, 57452664)
Improvements
- Backend packaging cleanup after Pi consolidation — Removes stale Copilot/Codex binary packaging entries, import guards, runtime fields, and docblocks that survived the move to the actual two-backend architecture (ClaudeAgent + PiAgent). This trims dead build metadata and makes the codebase less misleading for future backend work. (a96b8706)
- OSS README metadata refresh — Adds the Trendshift badge to the OSS README so repository discovery metadata stays current. (7feae925)
Bug Fixes
- OpenAI/Codex long-running session instability — Upgrades the Pi SDK stack to 0.73.1, picking up the upstream Codex transport fixes: WebSocket setup can fall back to SSE before streaming starts, and cached WebSocket sessions are closed properly during session shutdown. This targets the 1011 keepalive timeout, 1006 disconnect, and certificate-verification failures reported for long-running ChatGPT Plus / Codex OAuth sessions. Fixes #747. (e30762b6)
- Compact session menu no longer clips nested actions — On narrow/compact layouts, the chat title menu now uses a vaul drawer with iOS-style drill-in panes for Status, Labels, Share, and Messaging instead of Radix nested dropdowns anchored inside the container-query panel. This keeps status and label actions reachable on small widths while leaving the desktop dropdown unchanged. The follow-up review fixes close leaked desktop dropdown state when switching layouts and prevent drawer actions from retargeting if navigation changes the active session while the drawer is open. (db9f506f, 084547e8)
- Rapid session-label toggles are race-safe — Extracts shared session-menu behavior into useSessionMenuActions and moves label toggling to optimistic state that compounds rapid taps, avoids Strict Mode double side effects, and resets correctly when switching sessions. This fixes compact drawer multi-label editing and removes behavior drift between desktop and compact session menus. (f0c12ed1, 88f8d733)
- Skills “Show in Finder” opens the real skill folder — Skills list, skill detail, and skill menu actions now reveal the authoritative skill.path directory instead of constructing a synthetic SKILL.md path, and failures surface via a platform-aware Finder/Explorer toast instead of silently no-oping. Fixes #756. (d44e5691)
Breaking Changes
- None. All changes are backward-compatible.

v0.9.2 — OAuth refresh ordering, Pi system prompt persistence, cross-machine spawn guard, i18n key restore, source_test/streaming/browser hotfixes
Bug Fixes
- OAuth tokens silently refresh before agent build — On cold sessions, SessionManager.sendMessage ran getOrCreateAgent → buildServersFromSources → refreshOAuthTokensIfNeeded in that order. The first build saw stale tokens, emitted AUTH_REQUIRED, and the wrapper called markSourceNeedsReauth — flipping isAuthenticated=false on disk. The user saw a brief "needs auth" UI flicker before the late refresh restored state. Three coordinated changes: (1) refresh now runs before getOrCreateAgent so its internal cold-session build sees fresh tokens, the old refreshOAuthTokensIfNeeded (refresh + conditional rebuild) is gone, and a single post-refresh build is the only build per send; (2) buildServersFromSources reclassifies AUTH_REQUIRED → TOKEN_EXPIRED when the credential is merely expired-but-refreshable and skips markSourceNeedsReauth in that case (prevents flicker if refresh is skipped, e.g. during cooldown); (3) TokenRefreshManager.ensureFreshToken failure branches now mirror markSourceNeedsReauth's disk write to source.config in memory, so isSourceUsable() returns false and the failed source is excluded from intendedSlugs by the post-refresh build. Closes #710. (347820ab)
- Pi backend silently dropped the Craft system prompt — Pi SDK 0.72.1's session.prompt() wipes agent.state.systemPrompt back to _baseSystemPrompt on every turn (agent-session.js ~L796), so pi-agent-server's direct assignment was silently dropped — taking preferences/notes, <session_state>, <sources>, working directory, and all other Craft-built system-prompt content with it. The Anthropic backend was unaffected because it uses the SDK's append: option, which the Claude SDK appends verbatim. New applySystemPromptOverride() helper stamps all three SDK private fields (state.systemPrompt, _baseSystemPrompt, _rebuildSystemPrompt) so the prompt survives both per-turn resets and tool-change rebuilds. Pattern matches OpenClaw's applySystemPromptOverrideToSession — same SDK, same constraint, validated workaround until upstream exposes a public API. Wired into both call sites that previously did state.systemPrompt = …: the per-turn prompt handler and the ephemeral queryLlm session. Regression test asserts all three fields are stamped (catches the original single-field regression and any future drop of one of the writes). Closes #648. (13f6e63e)
- SDK spawn guard against stale branch cwd from cross-machine imports — The SDK's Claude Code native binary not found at … error is a misleading wrapper around spawn ENOENT, which fires whenever the subprocess cwd is also missing — not just when the binary is missing. Cross-machine session imports preserve branchInfo.sdkCwd from the source machine, so a Send to Workspace recipient hits this error on first chat even though the bundle is fine. A pre-spawn guard now catches stale branchFromSdkCwd in chatImpl() and routes through the same branch-fallback recovery the post-failure path already uses (parent summary injection preserved). A new onBranchForkInvalidated callback persists all four fork fields atomically — fixing a pre-existing bug where onSdkSessionIdCleared only persisted sdkSessionId, leaving stale branch fields on disk to reload on next launch. ENOENT classification was lifted out of the if (isProcessError) gate (it never fired for the SDK wrapper, which is a ReferenceError, not a process-exit error); now it disambiguates binary vs cwd vs unknown causes, surfaces typed sdk_binary_missing / sdk_cwd_missing errors after retry exhaustion, and keeps the 2-second auto-retry for the auto-update bundle-swap window. Helpers extracted to agent/spawn-helpers.ts for unit testing (34 new tests; existing branching suite extended with v2 callback coverage). (12166f4a, e40a3801)
- source_test lastTestedAt now persists — UI no longer always shows "Never" — The source_test handler wrote an