mirror of
https://github.com/nexu-io/open-design.git
synced 2026-10-27 09:51:12 +01:00
agent-runtime
72
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 CaprikaandGitHub
|
06dbde51f9 |
[codex] Add Cursor Agent auth diagnostics (#1538)
* Add Cursor Agent auth diagnostics * Handle Cursor not logged in auth status * Address Cursor auth review feedback * Classify Cursor stdout auth failures |
||
|
CaprikaandGitHub
|
a3276ec542 |
[codex] Add visual draw annotation context (#1547)
* feat(web): add visual draw annotation context * Fix visual draw annotation staging * Fix concurrent visual annotation IDs |
||
|
lefarcen
|
5172e37217 |
Merge origin/main into release/v0.7.0 to prepare merge-back PR
Resolves 7 conflicts via hybrid strategy: - apps/web/src/components/EntryView.tsx: take main (Discord+X pills are forward feature) - apps/web/src/components/Icon.tsx: take main (switch-case refactor) - apps/web/src/components/NewProjectPanel.tsx: take release (preserve #1514 dropdown UX validated in 0.7.0 acceptance) - apps/web/src/index.css: take main (project-target-platforms / instructions chip styles) - apps/web/tests/components/FileViewer.inspect-empty-hint.test.tsx: accept main's deletion - nix/package-daemon.nix, nix/package-web.nix: take main pnpmDepsHash Non-conflicting hunks from #1519 (AppChromeHeader), #1428 (PostHog analytics call sites), and #1540 (release light background) are preserved via auto-merge. |
||
|
kamiandGitHub
|
4f76e836ae |
feat(audio): add ElevenLabs audio support (#1384)
* docs: add ElevenLabs audio support design * docs: add ElevenLabs audio implementation plan * feat(daemon): add ElevenLabs speech renderer * feat(daemon): add ElevenLabs sound effects renderer * fix(daemon): preserve ElevenLabs sfx durations * feat(web): expose ElevenLabs media providers * feat(daemon): document ElevenLabs audio contract * feat(audio): add ElevenLabs voice selection * chore: ignore superpowers scratch docs * fix(daemon): cache ElevenLabs voice options * fix(audio): expand ElevenLabs voice and SFX selection * fix(audio): align ElevenLabs SFX controls * fix(audio): tighten ElevenLabs SFX prompt budget * fix(audio): preflight ElevenLabs SFX prompt length * fix(audio): surface ElevenLabs lookup failures * fix(audio): sanitize ElevenLabs prompt errors |
||
|
RockyandGitHub
|
cd68c8a80a |
fix(daemon+web): emit conversation-created SSE event when routine run starts (#1523)
* fix(daemon+web): emit conversation-created SSE event when routine run starts When a Routine fires in "Reuse an existing project" mode, the daemon creates a new conversation in the project and writes a queued/running assistant message to the database, but the open `ProjectView` has no way to learn that anything happened: the project events SSE stream only carries `file-changed` and `live_artifact*` events, and `ProjectView` reloads conversations only when `project.id` changes. The result is the user's own routine "Run now" appears to do nothing until they exit and re-enter the project (#1361). Fix: - Add a `conversation-created` payload type to the existing project events stream in `apps/web/src/providers/project-events.ts`. The payload carries `projectId`, `conversationId`, `title`, and `createdAt`. Mirror the existing `file-changed` listener pattern with explicit malformed-payload handling. - In `apps/daemon/src/server.ts`, after `insertConversation` runs in the routine `setRunHandler` (both reuse-an-existing-project and new-project paths), broadcast a `conversation-created` event through the existing `activeProjectEventSinks` map. The function body was already generic so it was renamed from `emitProjectLiveArtifactEvent` to `emitProjectEvent` and the two pre-existing callers updated. - In `apps/web/src/components/ProjectView.tsx`, when `handleProjectEvent` receives a `conversation-created` event whose `projectId` matches the currently-viewed project, refetch the conversation list via `listConversations`. The active conversation is intentionally NOT changed — per maintainer guidance on #1361, auto-switching is a separate UX decision left for a follow-up. - `projectEventToAgentEvent` returns null for `conversation-created` so it doesn't get routed into the live-artifact path. Tests (`apps/web/tests/providers/project-events.test.ts`): - A single `conversation-created` event reaches the consumer with the parsed payload. - Two consecutive `conversation-created` events from concurrent routine runs both reach the consumer (covers the multiple-concurrent-runs case reported in #1502). - Malformed `conversation-created` payloads are swallowed without throwing, matching the existing `file-changed` / `live_artifact` defensive behavior. Manual verification: - Built locally with `pnpm exec tools-pack mac build --to app --portable` and installed. - Created a routine in `Reuse an existing project` mode targeting an existing project. - With the project view open, clicked `Run now`. The new "Routine" conversation appeared in the project's conversation list within about a second, without exiting and re-entering the project, and the active conversation was not changed. - Clicked `Run now` twice in quick succession; both new conversations appeared in the list, covering the concurrent-runs case in #1502. - `pnpm guard` and `pnpm --filter @open-design/web typecheck` clean; full web test suite is 1016/1016 passing. Fixes #1361 * fix(#1523 review): share SSE type via contracts; guard conversation refresh against project-switch and reordering races Addresses Codex P1 + lefarcen P2 inline review on #1523 (#1361): 1. Move `ProjectConversationCreatedSsePayload` to `@open-design/contracts` (`packages/contracts/src/sse/chat.ts`) so the daemon producer and the web consumer share one type. The web provider re-exports it under the local `ProjectConversationCreatedEvent` name to keep the existing import shape stable for callers; the daemon emit site picks up the same shape via a JSDoc typedef so producer and consumer can't drift as this stream grows. (Addresses lefarcen P2 on project-events.ts:17.) 2. Guard the `conversation-created` async refresh in `ProjectView` against two distinct races: - Project-switch race: capture `project.id` at dispatch time and re-check it via a live `projectIdRef` after `listConversations` resolves; bail if the user switched projects while the request was in flight. The existing project-load effects use the same cancellation pattern. (Addresses Codex P1 on ProjectView.tsx:767.) - Concurrent-refresh re-ordering race: bump a monotonic `conversationsRefreshTokenRef` on every dispatch and capture each request's token; only the request whose captured token still equals the live ref at await-return applies its result. Two rapid `conversation-created` events (the #1502 concurrent Run-now case) can no longer drop the newest conversation when the earlier request resolves last with a stale, shorter list. (Addresses lefarcen P2 on ProjectView.tsx:767.) Both guards are documented inline with comments that point back at the review threads. The existing project-events tests (single delivery, concurrent delivery, malformed payloads) are unchanged — the new guards are defensive logic on the consumer, not new event shapes. `pnpm guard`, `pnpm --filter @open-design/web typecheck`, `pnpm --filter @open-design/daemon typecheck`, and the full web test suite (1016/1016) remain green. |
||
|
lefarcenandGitHub
|
dc7791ef9d |
feat(analytics): add project_id + project_kind to studio/artifact events (#1509)
Product tracking doc 260513 added project_id + project_kind to studio_view (artifact), studio_click (share_option), and artifact_export_result. The Studio funnel can now group by project type without joining run_created on the back end. - contracts: 3 props gain required project_id + project_kind - ProjectView → FileWorkspace → FileViewer: thread projectKind down, converting metadata.kind via projectKindToTracking once at the top - FileViewer + HtmlViewer: populate the three call sites |
||
|
lefarcen
|
e2952acd05 |
Revert "fix(web): restore consistent app header layout (#1432)"
This reverts commit
|
||
|
sukumarp2022andGitHub
|
b167991d7c |
feat: add project-level and user-level custom instructions (#1304)
* feat: add project-level and user-level custom instructions Implements #510 — editable custom instructions that get injected into every model message, at both user level (Settings → Memory) and project level (pencil icon in project header). - Add customInstructions to Project, AppConfigPrefs contracts - Add custom_instructions column migration to projects table - Inject user + project instructions into system prompt (after memory, before design system; project-level wins on conflict) - Add Settings textarea for user-level instructions - Add inline editor bar in ProjectView for project-level instructions - Sync user-level instructions through daemon app-config round-trip * fix: address PR review — validation, draft reset, length limit - Reset instructionsDraft on Cancel and toggle close (stale draft bug) - Thread customInstructions through POST /api/projects create handler - Add type + length validation (5000 chars) in PATCH handler - Enforce length cap in app-config applyConfigValue - Add maxLength={5000} to both UI textareas - Resync draft via useEffect when editor is closed - Remove stray run.sh from commit * fix: address maintainer review — save race condition, precedence wording - Make handleSaveInstructions async with await + revert on failure - Add instructionsSaving state to disable Save/Cancel/textarea during save - Clarify precedence wording with concrete example in both prompt composers - UpdateProjectRequest already has customInstructions (verified) * fix: use server-returned project in save handler, drop optimistic update The previous optimistic-update + revert approach captured a stale project snapshot in the useCallback closure. On failure, reverting with the captured object could clobber unrelated project fields that changed during the async request. Switch to pessimistic update: wait for patchProject to succeed, then call onProjectChange(result) with the server-returned project object. The instructionsSaving flag disables the editor UI during the round-trip. * fix: align create/PATCH validation for customInstructions Create endpoint now rejects invalid types and >5000 char values with 400 instead of silently truncating, matching the PATCH handler behavior. |
||
|
+14Siri-RayGitHubPrantik Medhi이용진Nicholas-XiongHesamYuhao ChenchaoxiaochechaoxiaocheCursoreggward han@aaronjmars <61592645+aaronjmars@users.noreply.github.com>BryanDevForgeAI CI/CD EngineermrzhangkrislaomoNagendhra MadishettiNagendhraMasonYiang YiyanRockynetteeshangxinyu1Matt Van Horn |
3d3119333c |
fix(web): restore consistent app header layout (#1432)
* docs: add NotebookLM GitHub export script (#1062) * docs: add NotebookLM GitHub export script * fix: make NotebookLM export TOC anchors work * fix: escape TOC link text markdown chars * fix: include merged PRs when exporting --prs all * fix: allow --prs merged mode * fix: treat --limit as total export budget * fix: avoid starving buckets under global --limit * fix: support --issues none and handle repos w/ issues disabled * fix: avoid underfilling export when buckets empty * fix: keep disabled-issues fallback quiet * fix: silence disabled issues fallback * fix: satisfy script typecheck * prevent duplicate saves and add template deletion (#1294) * prevent duplicate template entries on repeated save * add delete button to saved template list Templates can now be removed from the template picker via a hover x button, calling the existing DELETE /api/templates/:id endpoint. * add missing onDeleteTemplate prop in test fixtures * add template deletion flow test for NewProjectPanel * reject template names longer than 100 characters * preserve original createdAt on template update * feat: add FAQ page skill (#1162) * fix: set writable OD_DATA_DIR default for nix run Fixes #1157 When running via 'nix run github:nexu-io/open-design', the daemon attempted to create runtime state under the Nix store package path: /nix/store/.../lib/open-design/.od/projects The Nix store is read-only at runtime, causing startup to fail with ENOENT when mkdir() tried to create the projects directory. This commit updates the nix run wrapper to export OD_DATA_DIR with a writable default ($HOME/.od) when the variable is unset. Users can still override it by setting OD_DATA_DIR before running. The Home Manager and NixOS modules already set OD_DATA_DIR, so they are unaffected by this change. * feat: add FAQ page skill Add a new skill for generating Frequently Asked Questions pages with: - Collapsible accordion sections for Q&A pairs - Real-time search functionality - Category filtering (Billing, Account, Technical, General) - Smooth animations and transitions - Keyboard navigation support - Mobile-friendly responsive design - Semantic HTML with proper ARIA attributes The skill includes: - SKILL.md with triggers, workflow, and output contract - example.html demonstrating a complete FAQ page with 12 questions Use cases: help centers, support pages, product documentation * fix: address PR review feedback for FAQ page skill - Fix craft slugs: use accessibility-baseline and state-coverage instead of non-existent slugs - Remove overly broad 'questions and answers' trigger - Add edge case handling for insufficient/excessive FAQs - Remove search highlighting requirement (XSS risk) - Update self-check to reflect filtering instead of highlighting Addresses review comments from @lefarcen and @chatgpt-codex-connector * feat: add localized copy for faq-page skill Add German, French, and Russian translations for the FAQ page skill example prompt to fix validation test failure. - DE: FAQ-Seite mit Akkordeon-Abschnitten, Suchfunktion und Kategoriefilterung - FR: Page FAQ avec sections accordéon, recherche et filtrage par catégorie - RU: Страница FAQ со складными секциями-аккордеонами, поиском и фильтрацией * fix: escape apostrophe in French translation Use double quotes to avoid syntax error with d'auth * fix(platform): add legacy ~/.fnm path to wellKnownUserToolchainBins (#1110) * fix(platform): add legacy ~/.fnm path to wellKnownUserToolchainBins fnm legacy installations use ~/.fnm/node-versions. Closes #1102 * fix: remove stray .fnm token from type declaration * docs: add Windows troubleshooting guide (#478) (#1170) * docs: add Windows troubleshooting guide (#478) Add docs/windows-troubleshooting.md with step-by-step fixes for the most common native-Windows setup errors: - Node 24 / nvm-windows gotchas (fake nvm file in System32) - pnpm not found after installation - Build scripts blocked by pnpm 10 (better-sqlite3, sharp) - Visual Studio / gyp build errors - Starting the dev server - Optional OpenCode CLI setup Also update CONTRIBUTING.md and QUICKSTART.md to link to the new guide instead of the vague "file an issue if it doesn't" note. * docs: fix Windows guide command accuracy (#1170) Address all 6 inline review comments from lefarcen: - Pin npm-global pnpm install to @10.33.2 (matches packageManager field) - Use where.exe instead of bare where (PowerShell alias conflict) - Fix OpenCode package: opencode-ai (not opencode), binary is opencode - Add EPERM fallback note for corepack enable on protected installs - Add Python check for gyp ERR! find Python - Expand diagnostic checklist with corepack, python, execution policy Also remove redundant corepack pnpm --version from checklist. * feat(daemon): inject compiled design-system tokens + fixture into prompts (#1385) * feat(daemon): inject compiled design-system tokens + fixture into prompts Follow-up to #1231. The prior PR landed the structured form of two brands (`default` + `kami`) and codified the schema; this PR teaches the daemon to actually consume those files when assembling the system prompt, so agents stop having to re-derive token names from DESIGN.md prose every turn. Gated behind `OD_DESIGN_TOKEN_CHANNEL=1` for the smoke-test phase — flag-off keeps the daemon byte-equivalent to today's behavior, flag-on appends two new prompt blocks (the brand's `tokens.css` :root contract and its `components.html` reference fixture) right after the existing DESIGN.md block. Brands without those sibling files (every brand except `default` and `kami` today) skip silently in either mode. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(daemon): only swallow ENOENT/ENOTDIR in readFileOptional, rethrow rest Reviewer feedback (nettee, #1385). The prior catch-all hid permission errors, EISDIR, and broken packaged-resource paths behind the same "undefined = absent" branch the legacy ~138-brand fallback uses, which would let `OD_DESIGN_TOKEN_CHANNEL=1` silently degrade to the DESIGN.md-only prompt while reporting success. That corrupts the exact signal the smoke-test rollout depends on. Now `readFileOptional` only returns undefined for ENOENT / ENOTDIR (real "file does not exist" cases) and rethrows everything else. Added a focused test that plants a directory at the tokens.css path to exercise the EISDIR branch, plus a partial-presence regression test to confirm the stricter contract preserves the legacy fallback. Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: chaoxiaoche <chaoxiaoche@192.168.10.16> Co-authored-by: Cursor <cursoragent@cursor.com> * feat(daemon): make connection-test timeouts configurable (#1222) * feat(daemon): make connection-test timeouts configurable Provider and agent connection tests had hardcoded 12s / 45s budgets, which are too tight for slow networks or distant providers (the user sees "timeout" in Settings with no way to extend the budget). - Add OD_CONNECTION_TEST_PROVIDER_TIMEOUT_MS (default 12_000) - Add OD_CONNECTION_TEST_AGENT_TIMEOUT_MS (default 45_000) - Invalid values (non-numeric, zero, negative, fractional) emit a console.warn and fall back to the default, so a typo in the env never silently disables the safety timeout. - Export resolveConnectionTestTimeoutMs for unit testing; cover the three resolution paths (fallback / honored override / invalid). 41 connection-test tests pass (+3 new), full daemon suite 1170/1170. * fix(daemon): reject connection-test timeout overrides above Node's setTimeout maximum Node's `setTimeout` silently clamps any delay above `2^31-1` ms (2_147_483_647) to ~1 ms with a TimeoutOverflowWarning. The previous `Number.isInteger(n) && n >= 1` check accepted oversized values unchanged and passed them straight to `setTimeout`, so an override that *intended* to raise the budget — e.g. `OD_CONNECTION_TEST_AGENT_TIMEOUT_MS=3000000000` — instead caused every connection test to fail almost immediately. The safety timeout was effectively disarmed. Add `MAX_CONNECTION_TEST_TIMEOUT_MS = 2_147_483_647` and switch the guard to `Number.isSafeInteger(n) && n >= 1 && n <= MAX...`. The boundary value is still accepted; one millisecond past it falls back with a warn. Regression test exercises `3_000_000_000`, `2_147_483_647`, and `2_147_483_648`. Addresses #1222 review feedback from @chatgpt-codex-connector, @mrcfps, and @lefarcen. * fix(security): strip trailing dot in normalizeBracketedIpv6 (FQDN SSRF bypass) (#1122) * fix(security): strip trailing dot in normalizeBracketedIpv6 (FQDN bypass) new URL('http://192.168.1.5./').hostname returns '192.168.1.5.' — the trailing dot is the RFC 1034 absolute-FQDN form and resolves identically to '192.168.1.5'. parseIpv4 fails on the dotted form, so 169.254.169.254. slips past the metadata-service block, 192.168.1.5. slips past the LAN block, and localhost. slips past the loopback identification. Strip trailing dots in normalizeBracketedIpv6 so all downstream checks (isLoopbackApiHost, isBlockedExternalApiHostname, isBlockedIpv4, IPv6 range tests) see the canonical form. Adds 6 vitest cases covering loopback FQDN forms (localhost., foo.localhost., 127.0.0.1.) and SSRF FQDN bypasses (169.254.169.254., 192.168.1.5., 10.0.0.5.). Refs nexu-io/open-design#1119 review feedback (P2 from @lefarcen). * test(connectionTest): tighten trailing-dot coverage per #1122 review Two issues from #1122 review: 1. (P2 from @mrcfps + codex bot) The original `foo.localhost.` case asserted error===undefined on validateBaseUrl, which only proves the URL passed validation — not that the host is identified as loopback. Replaced with direct isLoopbackApiHost(...) assertions on the actual loopback FQDN forms (localhost., 127.0.0.1., 127.0.0.5.) so the test exercises the loopback path the comment claims. 2. (P3 from @lefarcen) Original blocked-FQDN tests covered only 3 of 7 ranges that isBlockedIpv4 handles. Added a dedicated case per range (0.0.0.0/8, 10/8, 100.64/10, 169.254/16, 172.16/12, 192.168/16, multicast >=224) so future regressions in normalizeBracketedIpv6 surface against the full coverage. * docs: drop misleading foo.localhost./endsWith claim in normalizer comment @lefarcen review feedback: isLoopbackApiHost only accepts exact 'localhost', '::1', loopback IPv4, and mapped loopback IPv4 — there's no subdomain or endsWith handling, so referencing 'foo.localhost.' overstates what the trailing-dot strip enables. Rewrite the comment to match actual call sites (isLoopbackApiHost equality + isBlockedIpv4 numeric parse). * feat(daemon): export self-contained HTML via /export/*?inline=1 endpoint (#1312) * test(daemon): add Red unit tests for inlineRelativeAssets helper 14 cases pinning the behavior contract for the upcoming apps/daemon/src/inline-assets.ts helper: - link/script inlining with verbatim body preservation - non-src script attrs preserved (type=module, defer, crossorigin) - relative path resolution (root + nested + deep-nested owners) - self-closing and single-quoted attr forms - negative cases: missing rel, rel=preload, absolute/data/blob/leading-slash - escaping: </style and </script inside body - null-fileReader graceful degradation - duplicate identical tags fully replaced (diverges from apps/web/src/components/FileViewer.tsx:5313's first-match-only; locked decision per plan §3.3) - HTML-escaped data-od-inline-asset attr Tests intentionally Red — module ../src/inline-assets.js does not yet exist. Phase B-G of plan declarative-roaming-gosling.md will turn them green by porting FileViewer.tsx:5248-5354 server-side. Refs nexu-io/open-design#368. * feat(daemon): port inlineRelativeAssets server-side for export endpoint Adds apps/daemon/src/inline-assets.ts — a pure helper that takes (html, ownerFileName, fileReader closure) and returns the HTML with every relative <link rel=stylesheet> and <script src> contents inlined into <style data-od-inline-asset="…">/<script>…</script> blocks. The fileReader closure keeps the helper free of fs/Express coupling so the route handler owns the filesystem boundary. Port source: apps/web/src/components/FileViewer.tsx:5248-5354 — five functions (inlineRelativeAssets, resolveProjectRelativePath, baseDirFor, readHtmlAttr, escapeHtmlAttr). The fetch hop becomes the fileReader closure; replace-all replaces first-match-only per locked design decision §3.3 (inline comment in inline-assets.ts cites the divergence from FileViewer.tsx:5313 and notes the web inline path is on a deprecation track since PR #384 made URL-load the default). Phase B-G of plan declarative-roaming-gosling.md. All 14 unit cases from the Red commit ( |
||
|
lefarcenandGitHub
|
e1bc83a476 |
feat(analytics): PostHog product analytics (P0 events, consent-gated, packaged) (#1428)
* feat(analytics): scaffold PostHog product-analytics integration
- Add @open-design/contracts/analytics subpath with the 17 P0 event
payload types, header constants, and code↔CSV enum mapping helpers.
- Add apps/daemon/src/analytics.ts with env-gated posthog-node client,
request-scoped analytics context reader, and artifact-id anonymizer.
- Expose GET /api/analytics/config so the web bundle never embeds the
PostHog key at build time; daemon owns POSTHOG_KEY / POSTHOG_HOST.
- Add apps/web/src/analytics module (identity + lazy posthog-js client
+ React provider) and mount it under <I18nProvider> in app/layout.
No event wiring yet — that lands in the next commit alongside trigger
points (App.tsx, EntryView, NewProjectPanel, SettingsDialog, FileViewer,
runs.ts).
* feat(analytics): wire app_launch, home_view, home_click, project_create_result
- App.tsx: fire app_launch once after first effect tick. handleCreateProject
now emits project_create_result on both success and failure paths.
- EntryView.tsx: home_view (page) gated on agents loading so
has_available_cli isn't transiently false; home_view (asset_panel) fires
per top-tab change with the right result_count.
- NewProjectPanel.tsx: home_click create_button fires before delegating to
the parent; a fresh request_id is generated here and threaded through
onCreate so the matching project_create_result stitches via $insert_id.
- contracts/analytics: tighten createTabToTracking and topTabToTracking
for the worktree branch's renamed tabs (live-artifact, templates).
* feat(analytics): wire settings_view + 3 settings_click events
- settings_view fires on dialog mount and on every section switch,
carrying the active section (mapped via settingsSectionToTracking
for the 16-section worktree layout), execution_mode, and the
selected CLI provider id when present.
- settings_click execution_mode_tab: setMode now emits before/after
values whenever the user toggles between Local CLI and BYOK.
- settings_click cli_provider_card: agent card onClick reports
cli_provider_id via agentIdToTracking (kiro → other).
- settings_click byok_field: onFocus added to api_key, model select,
and base_url inputs; provider_id widened to include google so the
worktree's Gemini protocol slot type-checks.
* feat(analytics): wire studio_view + studio_click chat, studio_view artifact
- packages/contracts/src/analytics/artifact-id.ts: FNV-1a 64-bit helper
produces a 16-hex anonymized id for (projectId, fileName). Stable
cross-platform so the daemon and the web bundle resolve the same id
without a Web Crypto round-trip; daemon now re-exports it.
- ChatComposer: studio_view chat_panel fires once per project mount,
studio_click chat_composer fires on attachment + send buttons with
estimated user_query_tokens (length/4) and has_attachment.
- FileViewer: studio_view artifact fires once per (project, file) at
the dispatcher level, before any sub-viewer renders, with
artifact_kind derived from the renderer registry / file.kind table.
- Widen TrackingExportFormat to include markdown and cloudflare_pages
so the worktree branch's full share menu can emit verbatim.
* feat(analytics): wire studio_click share_option + artifact_export_result
HtmlViewer's share menu now emits both events per click via a
fireShareExport helper:
- studio_click share_option fires immediately on click with the chosen
export_format and a fresh request_id.
- artifact_export_result fires when the export resolves — success for
sync exporters (html, markdown, template) the moment the call
returns, success/failed for async exporters (pdf, zip, deploy)
via .then/.catch. The same request_id threads both events so
PostHog stitches click → result via $insert_id.
DEPLOY_PROVIDER_OPTIONS maps to the CSV's vercel / cloudflare_pages
slots; markdown is now a first-class export_format value.
Also ignore .env.local so local POSTHOG_KEY / .env-style secrets
don't get committed.
* feat(analytics): emit run_created and run_finished from the daemon
POST /api/runs now reads the analytics context off the
x-od-analytics-* headers the web client sets on every fetch, then:
- Captures run_created with project_id, conversation_id, run_id,
model_id, agent_provider_id (mapped via agentIdToTracking),
skill_id, design_system_id, plus the token_count_source marker.
- Schedules a run_finished capture on runs.wait(run) resolution,
mapping succeeded/canceled/failed to success/cancelled/failed and
reporting total_duration_ms.
Both events use a stable insert_id derived from the same uuid so
PostHog dedupes the daemon-side mirror against any future
web-side capture without double-counting.
Token sub-fields (user_query_tokens/system_prompt_tokens/...) stay
omitted in v1 — the claude-stream parser only exposes input/output
totals today. See tracking-doc-issues.md §3.2.
* feat(analytics): emit settings_cli_test_result + settings_byok_test_result
The original BLOCKING-list assumed these CSV P0 events were not
implementable in this branch because main lacked Test buttons. The
worktree HEAD actually wires `handleTestAgent` and `handleTestProvider`
in SettingsDialog, so both events are now in scope.
- handleTestAgent emits settings_cli_test_result on success and
failure paths with cli_provider_id mapped via agentIdToTracking,
result drawn from result.ok / catch branch, error_code from
result.kind or the thrown error name, and duration_ms timed via
performance.now().
- handleTestProvider emits settings_byok_test_result analogously,
using apiProtocol (anthropic|openai|azure|ollama|google) directly
as provider_id — wider than the CSV's 5-value enum, documented in
tracking-doc-issues.md §2.5.
Contracts: add SettingsCliTestResultProps / SettingsByokTestResultProps
plus matching track* helpers. AnalyticsEventName union now covers all
14 P0 events this branch supports.
* feat(analytics): gate PostHog on the existing telemetry.metrics consent
The integration now reuses the same first-launch privacy banner +
Settings → Privacy toggle that gates Langfuse, so a single user
decision controls both telemetry sinks.
- /api/analytics/config now consults the persisted AppConfigPrefs:
it returns enabled=true only when POSTHOG_KEY is set AND the user
has chosen "Share usage data" (telemetry.metrics === true). The
response also echoes installationId so the web client uses the
same anonymous id Langfuse keys off of — one identity per install,
shared across both sinks.
- Web AnalyticsProvider:
- Bootstrap fetch resolves installationId and threads it through
the x-od-analytics-anonymous-id header on every /api/* fetch,
so daemon-side captures (run_created / run_finished /
project_create_result) land on the same person record.
- Exposes a setConsent(granted) method that calls posthog-js's
opt_in_capturing / opt_out_capturing, wired from App.tsx via a
useEffect watching config.telemetry?.metrics. Toggling Privacy
→ metrics now stops/resumes events immediately, no reload.
- app_launch additionally gates on telemetry.metrics so a freshly-
declined user fires nothing, and a freshly-opted-in user fires on
the next reload.
* feat(packaging): bake POSTHOG_KEY into packaged daemon spawn env
Wires PostHog product analytics through the same Langfuse-style build-
secret pipeline so official Open Design builds ship with the key while
fork builds compile without it (the integration short-circuits cleanly
when POSTHOG_KEY is absent).
tools/pack
- resolveToolPackConfig reads POSTHOG_KEY / POSTHOG_HOST from
process.env at packaging time, validates them (no whitespace in the
key, http(s) URL for host, trailing-slash strip), and stamps them on
ToolPackConfig. Fork builds without the env vars simply omit the
fields; the daemon-side gate keeps things off in that case.
- Mac, Windows, and Linux packaged-config writers each append the two
fields to open-design-config.json next to the existing
telemetryRelayUrl entry.
apps/packaged
- RawPackagedConfig / PackagedConfig surface posthogKey / posthogHost
so the Electron entry and headless entry both forward them to the
daemon sidecar.
- buildPackagedDaemonSpawnEnv emits POSTHOG_KEY / POSTHOG_HOST into
the daemon child env when present. The daemon's existing analytics
module reads these via process.env — no daemon-side changes needed.
- The headless packaged path falls back to process.env for fields the
builder hasn't injected, mirroring how OPEN_DESIGN_TELEMETRY_RELAY_URL
is read there.
CI
- release-beta.yml and release-stable.yml expose POSTHOG_KEY (secret)
and POSTHOG_HOST (var) at workflow-env scope so every packaging job
inherits them. PR / fork builds without these set simply skip the
bake step.
Tests
- tools/pack: config.test.ts covers bake-through, fork-build omission,
whitespace rejection, invalid-URL rejection, and trailing-slash
normalization.
- apps/packaged: sidecars.test.ts covers buildPackagedDaemonSpawnEnv
forwarding the keys when present and omitting them when null.
* feat(analytics): enable PostHog autocapture + perf + exceptions
Flip on the PostHog SDK's automatic diagnostic features so we capture
click paths, page transitions, web vitals, dead clicks, and browser
exceptions without scattering instrumentation through the codebase.
Privacy defense lives in one place — apps/web/src/analytics/scrub.ts —
wired in via posthog-js's `before_send` hook so every outgoing event
passes through the same audit point:
- $autocapture / $rageclick / $dead_click / $copy_autocapture:
strips $el_text and value/placeholder/aria-label attrs from any
input, textarea, password input, or contenteditable element. PostHog
autocapture does not capture input.value by default, but $el_text
on a <textarea> reflects the typed content — that's the prompt
body for us, so it has to be scrubbed every time.
- $pageview / $pageleave: drops query string and fragment from
$current_url / $referrer so any future ?q=… can't leak.
- $exception: rewrites file:// and absolute filesystem paths in
stack frames to app://apps/<repo-relative> so we don't ship the
user's home directory.
- Suppresses $opt_in entirely — duplicate of our explicit
setConsent toggle in App.tsx.
Element-level defense in depth is limited to the single most sensitive
surface: the chat composer textarea gets `ph-no-capture` so PostHog
never even generates an event for clicks inside that subtree. Every
other input relies on scrub.ts — sprinkling the class through every
form would be noisy and easy to forget on new surfaces.
The existing Privacy → "Share usage data" toggle continues to gate
every new feature: posthog-js's opt_out_capturing() halts autocapture,
$pageview, $exception, web vitals, and dead clicks alongside the
explicit capture() calls — one global switch.
11 unit tests pin the scrub rules in apps/web/tests/analytics-scrub.test.ts.
* ci(nix): bump pnpmDepsHash for posthog-js + posthog-node additions
Adding posthog-js to apps/web and posthog-node to apps/daemon changed
pnpm-lock.yaml, which Nix's fixed-output pnpmDeps derivation pins by
sha256. The CI nix flake check failed with:
specified: sha256-KF3Mld72/iau+pJmA7HvnanRx8VLtDP0N624SKrtrrc=
got: sha256-PGFgX4lYyeH2TRAXfUq52A3EOa6bb1gO59hPsXhEk3s=
Copy the new hash into both nix/package-web.nix and
nix/package-daemon.nix per the procedure documented in nix/README.md
§"First-build hash pinning".
* feat(analytics): unify PostHog identity with Langfuse installationId
PostHog's distinct_id is the installationId stamped by /api/analytics/
config; Langfuse already reads the same id off app-config.json to
populate trace.userId. With both sinks keying off the same anonymous
identity, dashboards can correlate user actions (PostHog events) with
LLM runs (Langfuse traces) without re-identifying.
Two gaps closed:
1. applyConsent(false) — clear posthog-js's persisted ph_*_posthog
localStorage entry on opt-out via posthog.reset(). Without this, a
user who opts out, then clicks Delete my data, then re-opts in
would see PostHog stitch their new session to the deleted identity
because bootstrap.distinctID only takes effect on first init.
2. applyIdentity(newInstallationId) — Delete my data rotates the
installationId in app-config; App.tsx now watches config.installationId
and calls posthog.reset() then identify(newId) so the next event
batch is fully decoupled from the deleted one. Idempotent on
same-id re-renders so benign config refreshes don't churn PostHog
identities.
The fetch wrapper's x-od-analytics-anonymous-id header also flips to
the new id on rotation so daemon-side captures (run_created /
run_finished) land on the same person record from the very next API
call, not after a reload.
The end-to-end rotation flow is verified against a live PostHog
project; these unit tests pin the safety guards (no-client paths, null
inputs) since stubbing posthog-js's init-loaded callback chain is
brittle.
* fix(langfuse): require both metrics AND content consent for trace reports
Tightens the Langfuse gate so a user who shares anonymous metrics but
NOT conversation content stops emitting Langfuse traces entirely —
Langfuse is used for turn-quality evals which only make sense with
prompt/output bodies. PostHog (product analytics, content-free) stays
gated on `metrics` alone and is unaffected.
i18n: "Conversation content" → "Conversation and tool content" with
hints expanded to mention tool inputs/outputs so the consent surface
matches what the trace actually carries (en + zh-CN).
Bundled here per PR scope — change originated outside this PostHog
PR but lands cleanly on the same files; gating Langfuse strictly
on `content` makes the dual-sink consent model (PostHog = metrics,
Langfuse = metrics + content) symmetric across both i18n locales and
the daemon-side gate.
* feat(analytics): wire byok_provider_option + fix PR review P1s
Adds the BYOK protocol-chip click event (5-value provider_id mirroring
the apiProtocol Settings UI) and resolves four P1 review threads on
PR #1428.
byok_provider_option:
- New SettingsClickByokProviderOptionProps in contracts (provider_id =
anthropic|openai|azure|google|ollama; maps to CSV's 5 values per
tracking-doc-issues.md §2.5).
- trackSettingsClickByokProviderOption helper in apps/web/src/analytics.
- SettingsDialog hooks it on the protocol-chip onClick alongside the
existing setApiProtocol call; is_selected reflects whether the chip
was already active.
Review fixes:
1. client.ts (Siri-Ray): clear `initPromise` when the resolution is
null so a Privacy → metrics opt-in after a previous decline triggers
a fresh /api/analytics/config fetch. Without this, the disabled
response was cached forever — first-session opt-in needed a reload
to start sending PostHog events.
2. provider.tsx (Siri-Ray): replace `url.includes('/api/')` with a
strict same-origin + /api/ pathname check (shared
`isSameOriginApiCall` helper). Outbound third-party URLs containing
`/api/` (e.g. provider.example.com/api/x) no longer receive our
x-od-analytics-* headers.
3. provider.tsx (codex-connector, lefarcen): gate header injection on
`resolvedAnonId` being non-null. When Privacy → metrics is off,
/api/analytics/config returns enabled=false → resolvedAnonId stays
null → wrapper never installs → daemon can't read consent-bearing
headers → no daemon-side PostHog event. setConsent now also clears
resolvedAnonId on opt-out and re-fetches on opt-in.
4. daemon/analytics.ts (defense in depth): createAnalyticsService now
takes dataDir and capture() re-reads app-config to check
telemetry.metrics inside the fire-and-forget wrapper. Even if a
stale header somehow reaches the daemon after opt-out, the capture
is dropped before posthog-node.capture is called.
* fix(web): place "Share usage data" on the right in privacy consent banner
Swap button order in PrivacyConsentModal and the in-settings ConsentCard
so the affirmative "Share usage data" lands on the right and "Not now"
on the left. Matches the OK-on-the-right pattern users expect for
primary actions.
Both buttons keep equal visual prominence (same .privacy-consent-action
styling) so the swap doesn't change the EDPB equal-prominence stance
called out in the original Langfuse telemetry spec.
* feat(analytics): populate run_finished token totals from claude-stream usage
Daemon's claude-stream parser already emits agent usage events with
input_tokens / output_tokens totals; the run service buffers them in
run.events and Langfuse reads them out the same way. The run_finished
PostHog event was leaving these fields empty.
Scan run.events for the most recent agent usage frame on terminal
transition and emit input_tokens / output_tokens / total_tokens when
present. token_count_source flips to 'provider_usage' only when at
least one count landed; runs without provider-side usage data keep
'unknown'.
Provider does not break the input down into the 7 sub-fields the
tracking doc lists (memory / context / attachment / system_prompt /
…); those stay omitted until a parser change exposes them.
* feat(analytics): estimate user_query_tokens from prompt length
The user_query_tokens field for run_created / run_finished was hardcoded
to 0. We can't tokenize without bundling a model-specific tokenizer, but
the character/4 heuristic is the industry-standard estimate when one
isn't available and is enough for funnel analysis (prompt-length cohorts,
short-vs-long-query conversion rates).
Extracted from req.body via the same telemetryPromptFromRunRequest
pattern the daemon already uses for langfuse-bridge (currentPrompt then
message fallback). Only the integer count goes to PostHog — the prompt
text itself never leaves the daemon.
token_count_source flips appropriately:
- run_created with a prompt: 'estimated' (was 'unknown')
- run_created with no prompt: 'unknown'
- run_finished with provider usage: 'provider_usage' (overrides
baseProps' 'estimated' value)
- run_finished without provider usage: inherits 'estimated' or 'unknown'
from baseProps so input/output absent doesn't mask the estimate.
|
||
|
|
09a8fa8d64 |
feat(web): Critique Theater Phase 8 (8 Theater components, barrel, role-keyed CSS) (#1314)
* feat(web): pure reducer for Critique Theater states (Phase 7.1)
Pure CritiqueState reducer driven by the contracts-level PanelEvent
(the same shape both the live SSE stream and the recorded transcript
emit), so a single reducer powers both the in-flight panel and the
rerun replay. Lifecycle covers run_started → running → (shipped /
degraded / interrupted / failed), with panelist_open / dim /
must_fix / close / round_end events building per-round
CritiquePanelistView entries as they arrive.
Defensive behaviour that surfaced while writing the spec tests:
- Terminal phases (shipped / degraded / interrupted / failed) are
sticky against further lifecycle events for the same run, except
for parser_warning which can land late and is recorded in a side
channel without changing phase.
- A new run_started for a different runId at any time discards the
prior state and reboots, so the UI can launch consecutive runs
without an explicit reset action.
- Events whose runId does not match the active run return the same
state reference, so React's useReducer doesn't re-render
subscribers on stray traffic.
- Round bookkeeping keys by round number rather than "always last",
so an out-of-order panelist_dim for round 1 arriving after a
round 2 dim does not corrupt the round 2 bucket.
Test coverage: 18 cases covering each transition, the runId guard,
sticky-terminal behaviour, the out-of-order round invariant, and
the stable-identity guarantee. Sets up Phase 7.2 and 7.3 to wire
SSE + replay into the same reducer.
* feat(web): useCritiqueStream hook subscribes to SSE and feeds reducer (Phase 7.2)
createCritiqueEventsConnection is a pure connection manager that
mirrors apps/web/src/providers/project-events.ts: opens an
EventSource at /api/projects/:id/events, listens for every name in
CRITIQUE_SSE_EVENT_NAMES, decodes each frame back into a PanelEvent
(stripping the critique. prefix and merging the data payload), and
hands it to the caller's onEvent. Reconnect uses exponential
backoff (1s → 30s) and resets on `ready`; malformed payloads drop
with a dev-mode warning rather than tearing the stream.
useCritiqueStream wraps the manager in a useReducer that owns the
CritiqueState. enabled=false or a null projectId tears down the
connection cleanly; switching projectId closes the old connection
and opens a fresh one. The returned dispatch lets local UI
synthesise actions (e.g. an Esc keypress firing a synthetic
interrupted while a kill request is in flight); production traffic
comes from the SSE stream.
Test coverage:
- sse.test.ts (10 cases, node env): subscription set covers every
CRITIQUE_SSE_EVENT_NAMES channel; payload decoding lifts the wire
shape back to PanelEvent; malformed JSON is swallowed and does
not stop the stream; exponential backoff schedule and ready-reset
semantics are pinned with a setTimeout seam; close() cancels
pending reconnects and shuts the live source; no-op fallback
when EventSource is unavailable.
- useCritiqueStream.test.tsx (6 cases, jsdom env): idle pre-event,
reducer driven by synthetic actions, no connection when disabled
or projectId is null, clean close on unmount, projectId change
reopens cleanly.
* feat(web): useCritiqueReplay hook drives reducer from transcript file (Phase 7.3)
Fetches the per-run NDJSON transcript (one PanelEvent per line),
parses every line via the shared isPanelEvent predicate, and
dispatches into the same CritiqueState reducer the live SSE stream
uses. A single reducer means the UI rendering a replay can be
identical to the live panel, and a UI mounting both
useCritiqueStream and useCritiqueReplay in parallel does not have
to reconcile two state shapes.
speed knob is `paused | instant | live | { intervalMs: N }`.
- instant flushes every event synchronously, useful for opening a
finished run already at its terminal state.
- intervalMs paces dispatches at a fixed cadence so the reviewer
can watch the run unfold.
- paused parses the transcript but holds events back until the
caller advances speed (consumers can drive a scrubber later).
- live is reserved for the future "playback at original cadence"
feature, currently treated as instant; replay timestamps are not
yet persisted with each event so honest pacing requires a
follow-up Phase 7+ task.
gunzip seam handles `.ndjson.gz` transcripts via
DecompressionStream when present; the production fetch path picks
between text and arrayBuffer based on the URL extension. Both seams
are injectable so the unit tests don't need to spin up a real
network or a real gzip pipeline.
Test coverage (8 cases, jsdom env):
- Idle status before any URL is provided.
- speed=instant flushes the full transcript synchronously to
shipped state.
- speed={intervalMs:N} paces with the setTimeout seam, reaching
done after the last tick.
- speed=paused leaves status=playing with no dispatches.
- Empty transcript reports done with state still idle.
- Fetch rejection surfaces an error status with the message.
- Malformed NDJSON lines are skipped; valid events around them
still land.
- .gz transcripts route through the gunzip seam.
Closes the Phase 7 plan tasks 7.1 / 7.2 / 7.3 (reducer + stream +
replay), all on one branch ready for review. Phases 8+ (Theater
components) consume these from this PR.
* fix(web): close payload-override gap + paused-resume bug in Critique Theater hooks (Phase 7 review)
Two P1 fixes from lefarcen's review on PR #1307:
SSE payload override
`sseToPanelEvent` previously spread `data` after the channel-derived
`type`, so a payload-provided `type` could override the channel and
route a `critique.run_started` frame into the reducer as a `ship`
action. Reversed the spread so the channel-derived `type` is
authoritative, and revalidated the resulting object through the
contracts-level `isPanelEvent` predicate before returning. Frames
that fail validation (missing runId, empty runId, unknown type) are
dropped, so a malformed or compromised SSE frame can no longer
dispatch a wrong-shape action into the reducer.
Three new sse.test.ts cases pin the regression: hostile `type:'ship'`
in the payload still resolves to `run_started`, missing runId is
dropped, empty runId is dropped.
Replay pause/resume
`useCritiqueReplay` had one big effect keyed on `transcriptUrl`
only, so flipping `speed` from `paused` to `instant` never re-fired
and the held events sat undispatched. Split into a parse effect
(depends on URL, fetches and stores events in state) and a pace
effect (depends on parsed-events + speed, owns the cursor + timers).
The playback cursor lives in a ref that survives pause/resume
cycles, so flipping `paused` -> `instant` flushes from the current
position rather than restarting (which would double-dispatch
`run_started` and reset the reducer).
Two new useCritiqueReplay.test.tsx cases:
- paused-then-instant transitions from `playing` to `done` and
reaches the shipped terminal phase
- intervalMs paced playback dispatches one event, pauses to drain
the next scheduled timer, flips to instant, and confirms the
remaining transcript drains exactly once (cursor was preserved)
Doc consistency
The earlier source comment in useCritiqueReplay.ts claimed `live`
"paces by recorded timestamps" while the impl used zero-delay
timers and the PR body said it behaves like `instant`. Aligned to
reality: `live` currently behaves like `{ intervalMs: 0 }` (events
drain on successive microtasks via setTimeoutFn) because transcripts
do not yet carry per-event timestamps. Honest timestamp-driven
pacing is queued as a Phase 7+ follow-up.
Validated: pnpm guard, pnpm --filter @open-design/web typecheck,
Theater suite 47/47 (up from 42, +3 sse + 2 replay), full web suite
96 files / 888 tests.
* feat(i18n): seed Critique Theater key block (en + zh-CN; other locales fall back via spread)
* feat(web): Theater PanelistLane component (Phase 8.1)
* feat(web): Theater ScoreTicker component (Phase 8.2)
* feat(web): Theater RoundDivider component (Phase 8.3)
* feat(web): Theater InterruptButton component with Escape keybind (Phase 8.4)
* feat(web): Theater TheaterDegraded chip (Phase 8.5)
* feat(web): Theater TheaterCollapsed post-run summary (Phase 8.6)
* feat(web): Theater TheaterTranscript replay surface (Phase 8.7)
* feat(web): Theater TheaterStage top-level container (Phase 8.8)
* feat(web): Theater CSS using existing semantic tokens (no hex literals)
* feat(web): Theater public exports barrel
* fix(web): resolve P2 + P3 review feedback on Phase 8 (PR #1314)
Addresses all 4 P2 + 3 P3 items from codex, Siri-Ray, and lefarcen.
State-lifecycle fixes (3 x P2)
1. Reducer learns a synthetic `__reset__` action (`CritiqueResetAction`).
Host hooks dispatch it when their gating prop changes so a stale
run from a prior project / transcript cannot bleed into the next
context. Reset is idempotent on idle (returns the same reference).
2. `useCritiqueStream` dispatches `__reset__` at the top of its
connection effect, so a workspace switch from project A (which
streamed a critique) to project B clears the reducer before the
new EventSource opens. enabled=false also clears.
3. `useCritiqueReplay` dispatches `__reset__` at the top of its
parse effect, so transcriptUrl swaps (including swap-to-null after
a replay reached `shipped`) lift the reducer back to idle before
the new fetch starts.
SSE validation (1 x P2)
4. `sseToPanelEvent` now runs a per-variant `hasValidVariantShape`
check after the cheap `isPanelEvent` predicate. A
`critique.ship` frame missing `composite` / `round` / `status` /
`artifactRef` is rejected before reaching the reducer, so
TheaterCollapsed can no longer crash on `undefined.toFixed(1)`.
Every variant's required fields are validated: run_started
(protocolVersion, non-empty cast, maxRounds, threshold, scale),
panelist_* (round, role, plus variant-specific shape), round_end
(round, composite, mustFix, decision in {continue,ship}, reason),
ship (round, composite, status, artifactRef.{projectId,artifactId},
summary), degraded (reason, adapter), interrupted (bestRound,
composite), failed (cause), parser_warning (kind, position).
Reducer correctness (1 x P2)
5. `panelist_open` now materializes the round + an empty panelist
view (`{dims: [], mustFixes: []}`) so TheaterStage can highlight
the in-progress lane the instant the tag opens. Before this, a
stream that emitted only `panelist_open` after `run_started` left
`rounds = []` and the UI rendered no current round until a later
`panelist_dim` arrived.
Polish (3 x P3)
6. Brand role tint swaps from `var(--magenta, var(--accent))` to
`var(--purple, var(--accent))`. `--purple` is actually defined
across the design systems; `--magenta` is not, so Brand was
silently falling through to `--accent` and looking identical to
Designer.
7. New i18n key `critiqueTheater.interruptedSummary` for the
interrupted-collapse copy ("Interrupted at round N, best
composite X.X"). Previously the interrupted branch reused
`shippedSummary` and the UI read "Shipped at round..." for a run
that specifically did not ship. Native value in en + zh-CN; other
locales fall back via `...en` spread.
8. `TheaterDegraded` heading id comes from `useId()` instead of a
hardcoded `theater-degraded-heading`, so two chips rendered on
the same page (chat history with multiple completed runs) keep
their aria-labelledby references unambiguous.
Tests (15 new cases)
- reducer.test.ts (+5): __reset__ on running/terminal/idle, panelist_open materializes round, panelist_open does not stomp prior panelist data.
- sse.test.ts (+6): variant-level rejection for ship without required fields, degraded without adapter, run_started with empty cast, panelist_dim with non-numeric score, round_end with unknown decision, plus a positive fully-formed ship.
- useCritiqueStream.test.tsx (+2): state reset on projectId change, state reset on enabled flip false.
- useCritiqueReplay.test.tsx (+1): state reset on transcriptUrl swap to null after a replay reached shipped.
- TheaterCollapsed.test.tsx (text-pinning update): asserts the interrupted branch reads "Interrupted at round 1" + "best composite 7.9", and explicitly NOT "Shipped at round...".
- TheaterDegraded.test.tsx (+1): two chips on the same page get unique aria-labelledby ids that each resolve to an `<h3>`.
Validated
- pnpm guard clean
- pnpm --filter @open-design/web typecheck clean
- Theater suite: 13 files, 101 tests (was 86 on the first Phase 8 push, +15 new)
- tests/i18n/locales.test.ts 5 of 5 across 18 locales
* fix(web): tighten isPanelEvent in contracts so enum + numeric fields are checked end-to-end (Siri-Ray round-3 P1 on PR #1314)
The variant validator on the web SSE path previously accepted any
`typeof === 'string'` for closed-enum fields (ship.status,
panelist_*.role, degraded.reason, failed.cause, parser_warning.kind,
run_started.cast[]) and any `typeof === 'number'` for numeric fields,
which let NaN / Infinity through. Downstream components index i18n
tables by enum value, so an unknown status or role would land
`SHIP_BADGE_KEY[final.status]` on undefined and crash the translator.
The replay parser had a separate gap: `useCritiqueReplay.parseTranscript`
called the cheap `isPanelEvent` header check directly, so a recorded
line like `{"type":"ship","runId":"r"}` reached the reducer with
composite, status, round, artifactRef, summary all undefined and
TheaterCollapsed then called `final.composite.toFixed(1)` on undefined.
Resolution: move all wire-side validation into the contract guard.
- Export const arrays for the closed enums:
SHIP_STATUSES, DEGRADED_REASONS, FAILED_CAUSES, PARSER_WARNING_KINDS,
ROUND_DECISIONS (PANELIST_ROLES already existed).
- Rewrite `isPanelEvent` in packages/contracts/src/critique.ts to be the
single deep validator: header (known type + non-empty runId) plus
every variant-specific required field plus closed-enum membership
plus Number.isFinite on every numeric field. Documented as the wire
source of truth.
- Drop the local `hasValidVariantShape` from web/sse.ts; sseToPanelEvent
now relies entirely on the contract guard, and parseTranscript in
useCritiqueReplay (which already uses isPanelEvent) gets the deeper
validation for free.
Tests (TDD, red-first):
- packages/contracts/tests/critique.test.ts: 13 new cases pinning the
strict guard directly (well-formed across every variant, every
rejection path: unknown type, empty/non-string runId, unknown enum,
non-finite numeric, missing variant field).
- apps/web/tests/components/Theater/state/sse.test.ts: 9 new cases for
each closed-enum rejection on the wire path plus a positive sweep
across every legal enum value across every variant.
- apps/web/tests/components/Theater/hooks/useCritiqueReplay.test.tsx:
2 new cases for incomplete and unknown-enum transcript lines.
Verified:
- pnpm --filter @open-design/contracts test 4 files / 30 tests green.
- pnpm --filter @open-design/contracts build clean.
- pnpm --filter @open-design/web typecheck clean.
- pnpm --filter @open-design/web test 107 files / 976 tests green.
* fix(contracts): enforce numeric domains in isPanelEvent (lefarcen P2 on PR #1314 round 4)
The strict guard from PR #1314 round 3 enforced enum membership and
Number.isFinite, but accepted any finite number where the contract
intends a specific domain: scale: 0 (ScoreTicker divides by it),
negative thresholds, fractional rounds, negative mustFix, etc.
ScoreTicker.tsx writes `var(--scale, ${state.scale})` into inline
CSS and divides by it for tick width, so a guard-passing scale: 0
shipped Infinity into the rendered style. Negative composite /
score values reached downstream code that assumes >= 0.
Resolution: mirror the daemon-side Zod domain constraints in the
runtime guard.
Three new helpers in packages/contracts/src/critique.ts:
- isPositiveInt(v): integer with v > 0. Used for round, maxRounds,
scale, protocolVersion (all 1-indexed in the orchestrator).
- isNonNegativeInt(v): integer with v >= 0. Used for mustFix,
position, bestRound. bestRound: 0 is the valid sentinel for
'interrupted before any round closed'.
- isNonNegativeFinite(v): finite number with v >= 0. Used for
composite, score, dimScore, threshold. Threshold may be
fractional (e.g. 8.5 on a scale of 10).
Cross-field check inside run_started: threshold <= scale (the daemon
Zod schema enforces this with an epsilon refine, the wire guard
matches the same intent).
Tests (TDD, red-first) added in packages/contracts/tests/critique.test.ts:
- 22 new rejection cases across every numeric field that
previously slipped through: scale: 0, negative scale, fractional
scale, maxRounds: 0, fractional maxRounds, protocolVersion: 0,
fractional protocolVersion, negative threshold, threshold > scale,
round: 0, fractional round, negative dimScore / score, negative /
fractional mustFix, negative composite, ship round: 0, negative /
fractional bestRound, negative interrupted composite, negative /
fractional parser_warning position.
- 3 positive boundary cases that must still pass: threshold == scale,
fractional threshold within [0, scale], interrupted with
bestRound: 0 (no round completed before interrupt), parser_warning
with position: 0 (start of stream).
Verified:
- pnpm --filter @open-design/contracts build clean.
- pnpm --filter @open-design/contracts test: 4 files / 59 tests green
(was 37 before the new domain cases).
- pnpm --filter @open-design/web typecheck clean.
- pnpm --filter @open-design/web test: 110 files / 1004 tests green;
no regression on Theater suite, sse validator, replay parser, or
assistant-feedback widget tests.
---------
Co-authored-by: Nagendhra <nagendhra405@gmail.com>
|
||
|
MasonandGitHub
|
2f51f3c1ae |
feature: refine assistant artifact feedback (#1379)
* feature: refine assistant artifact feedback * fix: clear hidden custom feedback reason * test: update assistant feedback expectations |
||
|
@aaronjmarsandGitHub
|
377d65b7e4 |
fix(security): strip trailing dot in normalizeBracketedIpv6 (FQDN SSRF bypass) (#1122)
* fix(security): strip trailing dot in normalizeBracketedIpv6 (FQDN bypass)
new URL('http://192.168.1.5./').hostname returns '192.168.1.5.' — the
trailing dot is the RFC 1034 absolute-FQDN form and resolves identically
to '192.168.1.5'. parseIpv4 fails on the dotted form, so 169.254.169.254.
slips past the metadata-service block, 192.168.1.5. slips past the LAN
block, and localhost. slips past the loopback identification.
Strip trailing dots in normalizeBracketedIpv6 so all downstream checks
(isLoopbackApiHost, isBlockedExternalApiHostname, isBlockedIpv4, IPv6
range tests) see the canonical form.
Adds 6 vitest cases covering loopback FQDN forms (localhost.,
foo.localhost., 127.0.0.1.) and SSRF FQDN bypasses (169.254.169.254.,
192.168.1.5., 10.0.0.5.).
Refs nexu-io/open-design#1119 review feedback (P2 from @lefarcen).
* test(connectionTest): tighten trailing-dot coverage per #1122 review
Two issues from #1122 review:
1. (P2 from @mrcfps + codex bot) The original `foo.localhost.` case
asserted error===undefined on validateBaseUrl, which only proves the
URL passed validation — not that the host is identified as loopback.
Replaced with direct isLoopbackApiHost(...) assertions on the actual
loopback FQDN forms (localhost., 127.0.0.1., 127.0.0.5.) so the test
exercises the loopback path the comment claims.
2. (P3 from @lefarcen) Original blocked-FQDN tests covered only 3 of 7
ranges that isBlockedIpv4 handles. Added a dedicated case per range
(0.0.0.0/8, 10/8, 100.64/10, 169.254/16, 172.16/12, 192.168/16,
multicast >=224) so future regressions in normalizeBracketedIpv6
surface against the full coverage.
* docs: drop misleading foo.localhost./endsWith claim in normalizer comment
@lefarcen review feedback: isLoopbackApiHost only accepts exact 'localhost',
'::1', loopback IPv4, and mapped loopback IPv4 — there's no subdomain or
endsWith handling, so referencing 'foo.localhost.' overstates what the
trailing-dot strip enables. Rewrite the comment to match actual call sites
(isLoopbackApiHost equality + isBlockedIpv4 numeric parse).
|
||
|
|
140a4e1ff6 |
Improve responsive preview and design handoff outputs (#1224)
* feat: improve responsive design handoff * feat: refine cross-platform design outputs Changelog:\n- Add auto-fit responsive preview behavior for tablet/mobile frames.\n- Add landing page and OS widgets metadata options with project header chips.\n- Strengthen prompt contracts for modern breakpoints, app-specific modules, CJX-ready UX, and final product surfaces.\n- Require cross-platform outputs to use separate platform files instead of tabbed demo selectors.\n- Add DESIGN-MANIFEST.json plus richer handoff guidance to daemon/client exports.\n- Update archive/export tests for manifest and responsive viewport matrix. * feat: enforce screen-file design outputs Changelog:\n- Enforce screen-file-first generation for landing pages, app screens, platform surfaces, and OS widgets.\n- Update design handoff and manifest exports so coding tools map each screen file to separate routes/surfaces.\n- Strengthen minimal-brief visual guidance to avoid monochrome or unstyled design outputs. * fix: address responsive handoff review feedback * fix: address handoff review blockers * fix: preserve proxy auth and normalized export entry * fix: narrow frame wrapper filter to directory paths only * fix: make artifact save failure banner generic --------- Co-authored-by: Huy Hoàng <macos@MacBook-Pro-Hoang.local> |
||
|
|
8962088c75 |
feat(daemon): guard against agent-emitted stub artifact regressions (#1171)
* feat(daemon): guard against agent-emitted stub artifact regressions
When an agent emits an <artifact> block whose body is a placeholder
("see other-file.html in this project", a bare filename string, a tiny
fallback page) instead of the full document, the daemon writes the
placeholder to disk verbatim. Users see a 25-500 byte HTML file where
their previous version had tens of kilobytes of real markup.
Add a structural regression guard in writeProjectFile: before writing
an html/deck artifact whose manifest carries metadata.identifier, scan
the project dir for prior siblings matching <identifier>(-\d+)?\.html?
and compare sizes. If the new body is below minRetainedRatio (default
0.2) of the largest prior sibling >= minPriorBytes (default 4096),
flag a regression. Three modes via env:
- OD_ARTIFACT_STUB_GUARD=warn (default) writes the file and attaches
stubGuardWarning to the response so the frontend can surface it.
- OD_ARTIFACT_STUB_GUARD=reject throws ArtifactRegressionError before
fs.writeFile; the route returns 422 ARTIFACT_REGRESSION with the
prior sibling's name and size in error.details.
- OD_ARTIFACT_STUB_GUARD=off skips the guard entirely.
Cross-agent by design: anchored on size delta + identifier match,
no agent-specific stub-phrase regex, so works for any agent backend
behind the agent-adapter abstraction.
The body-then-manifest write order pre-dates this change; the reject
path throws before fs.writeFile so rejections never leave a partial
state behind.
24 unit + 8 HTTP tests cover happy paths, all three modes, deck kind,
.htm extension sibling detection, ratio=1 edge case, and verify
rejected writes leave neither the html nor its manifest sidecar on
disk.
* fix(stub-guard): close same-name, nested-dir, and non-slug bypasses
Code review on PR #1171 (lefarcen, Codex, mrcfps) found three holes
where the stub guard could be silently bypassed. All three are now
closed with HTTP test coverage.
Same-name overwrite (lefarcen P1): the writer's prior-sibling scan
deliberately skipped the file at safeName, but for an in-session
overwrite (persistArtifact reuses the same fileName when
savedArtifactRef.current matches) that file is the prior content,
not the new entry. Drop the exclude-by-name filter; the current
on-disk size at scan time is always the prior because the overwrite
happens after this check.
Subdirectory scoping (Codex/mrcfps P2): writeProjectFile creates
parent directories for nested paths like reports/overview.html, but
the guard only scanned the project root. Pass path.dirname(target)
as scanDir so nested artifacts are evaluated against their real
sibling set.
Non-slug identifier (Codex/lefarcen/mrcfps P2): the web's
persistArtifact slugifies the filename basename but stores the raw
identifier in the manifest, so an identifier like "Landing Page"
yields filename landing-page.html with metadata.identifier="Landing
Page". Build the sibling regex from both the raw identifier and a
slugified variant (mirroring the frontend's slugifier) so either
form matches the same priors.
Also surface warn-mode warnings in the web UI: ProjectView now
checks file.stubGuardWarning after writeProjectTextFile and renders
the warning via setError. Reject-mode 422 surfacing requires
restructuring writeProjectTextFile's return contract and is
deferred.
API change inside the daemon: evaluateArtifactStubGuard /
findPriorArtifactSiblings drop excludeSafeName and rename projectDir
to scanDir. Tests updated.
Tests: 4 new HTTP cases (same-name overwrite preserves prior body,
nested subdir rejects, slug-form match rejects, plus the existing
warn/off/deck/.htm cases) and 1 new unit case (slug-form sibling
match). 44 tests pass.
* fix(stub-guard): empty-slug fallback + reject-mode UI surface
Round 3 review on PR #1171 (lefarcen, mrcfps) found two remaining
holes after
|
||
|
DongsenandGitHub
|
fd67b680d7 |
fix(contracts): pin API-mode override above discovery layer (#313) (#1207)
* fix(contracts): pin API-mode override above discovery layer (#313) The old streamFormat='plain' rule was appended at the BOTTOM of the composed prompt, but DISCOVERY_AND_PHILOSOPHY is pinned at the TOP with its own 'these override anything later' header — so its hard rules ('TodoWrite on turn 3', 'brand-spec extraction via Bash + Read + WebFetch') still won precedence in API mode. With no real tools wired through to the Anthropic Messages path, the agent narrated pseudo-tool markup (<todo-list>...</todo-list>, [读取 X]) instead of emitting structured tool_use events the UI could render. Move the API-mode override to the absolute top of the prompt so it beats the discovery layer, name every unavailable tool, and explicitly forbid the pseudo-tool / fake-protocol markup observed in #313. <artifact> output and <question-form> discovery are still allowed — both are markup the UI parses, not tool calls. * fix(daemon): mirror API-mode override above discovery layer (#313) Address Codex + mrcfps review on #1207: the daemon has its own copy of composeSystemPrompt that is hit by any adapter declaring streamFormat: 'plain' (e.g. DeepSeek) via server.ts:6190. That copy still appended the obsolete bottom '## API mode rule', which loses the precedence war against DISCOVERY_AND_PHILOSOPHY's 'these override anything later' header — so plain-stream daemon agents could still narrate <todo-list> / [读取 X] pseudo-tool markup. Mirror the same top-anchored API_MODE_OVERRIDE here (byte-identical to the contracts copy) so both code paths produce the same behaviour. Adds 8 daemon-side tests including the indexOf-based positional assertion that pins the override above the discovery layer header. |
||
|
|
b5eb8c1647 |
feat: generic skills + split skills/design-templates + finalize-design API (#955)
* feat: general-purpose skills with @-mention composition and user import
Lift skills from "one mode-bound skill per project" to a generic capability
the user can compose per turn:
- Daemon: scan multiple skill roots (user-skills under runtime data, then
the bundled `skills/`); user-imported skills can shadow built-ins by id.
- New `POST /api/skills/import` and `DELETE /api/skills/:id` endpoints,
with CONFLICT/BAD_REQUEST/NOT_FOUND error codes and built-in delete
protection.
- ChatRequest gains `skillIds: string[]`; the chat run concatenates each
picked skill's body (and merges craftRequires) into the system prompt
for that turn only — the project's persistent `skillId` is untouched.
- Web composer: `@` popover now lists skills alongside project files;
picks render as removable chips above the textarea and ride along with
the request as `skillIds`.
- Settings → Library: import form (name/description/triggers/body),
per-card delete for user skills, "user" origin badge.
* chore(web): drop welcome pet teaser + add ds→prompt-template mapping util
- SettingsDialog: remove the inline pet adoption teaser from the welcome
panel so the first-run modal stays focused on configuration.
- New `inferPromptTemplateCategoriesForDs(ds)` helper that maps a design
system's authored metadata to prompt-template gallery categories.
Imported by the design-system gallery wiring on a sibling branch; no
callers in this branch yet.
* feat: split skills/design-templates and add finalize-design API
Phase 0 of the skills/design-templates refactor (specs/current/
skills-and-design-templates.md):
- Move ~104 rendering catalogue entries from skills/ to design-templates/
and keep skills/ for the small set of functional skills that *do work*
on user input (utilities, briefs, packagers).
- Add design-templates/AGENTS.md and skills/AGENTS.md describing the
contract, and a brand-agnostic craft/ surface for opt-in craft rules.
- Daemon: add DESIGN_TEMPLATES_DIR / USER_DESIGN_TEMPLATES_DIR roots and
an /api/design-templates surface mirroring /api/skills. Asset/example
routes still span both registries so existing srcdoc URLs keep
resolving across the rename.
- Web: split LibrarySection into SkillsSection + DesignSystemsSection,
rename the EntryView "Examples" tab to "Templates", and update locales
+ the New-project picker accordingly.
Adds the finalize-design endpoint:
- New apps/daemon/src/finalize-design.ts and packages/contracts/src/api/
finalize.ts — one-shot synthesis of a project's transcript + active
design system + current artifact into <projectDir>/DESIGN.md via the
Anthropic Messages API. Per-project .finalize.lock mirrors the
transcript-export hygiene from PR #493; provider credentials are not
persisted by the daemon.
Other supporting changes:
- README + AGENTS.md updates to document the new directory split and
craft/ surface, plus i18n strings across 13 locales.
- Test refactors and new coverage (finalize-design, runs, sidecar
server, plus refreshed daemon integration tests).
- .gitignore: scope the *.exe ignore to /OpenDesign.exe so legitimate
vendor binaries are no longer hidden.
* fix(merge): move clinical-case-report to design-templates/
Origin/main added the clinical-case-report skill under skills/ before
the skills/design-templates split landed. Its od.mode is prototype, so
per specs/current/skills-and-design-templates.md it is a design template
and belongs alongside the other rendering catalogue entries — not under
the slimmed-down functional skills/ root. Moving it keeps the EntryView
Templates tab consistent with origin/main's intent.
* feat(skills): curated design/creative catalogue + collapsible Settings rows
Seed ~100 curated design/creative skill stubs under skills/ sourced from
awesome-claude-skills (ComposioHQ) and awesome-agent-skills (VoltAgent).
Each stub carries an od.category tag so the new filter pill row in
Settings -> Skills can group them. The seed script
(scripts/seed-curated-design-skills.ts, pnpm seed:curated-design-skills)
is idempotent: it only creates folders that don't already exist, so
hand-edited stubs are never overwritten.
- Daemon: parse and surface od.category on SkillInfo with a strict slug
normaliser; mirror the field on SkillSummary in @open-design/contracts.
Category is purely a UI hint — system-prompt composition is unchanged.
- Web: rewrite SkillsSection from a left-list / right-detail grid into a
vertical stack of collapsible rows mirroring the External MCP panel
(header always visible with name + mode/source/category pills + per-row
enable toggle; SKILL.md preview, file tree and inline edit form expand
on demand). Add a Category filter row above the list. Reorder Settings
nav so Skills + External MCP sit above the Composio/MCP cluster. Update
composer placeholder/hint across 17 locales to advertise '@ files or
skills · / for commands'.
- Docs: extend skills/AGENTS.md with the curated catalogue rules
(idempotency, category vocabulary, no upstream vendoring).
Co-authored-by: Cursor <cursoragent@cursor.com>
* test(skills): teach localized-content + system-prompt tests about the skills/design-templates split
mrcfps blocking review on PR #955: the skills/design-templates split
(
|
||
|
|
e254d1280b |
feat(memory): auto-memory store with chat-protocol-aware extraction (#999)
* feat(memory): auto-memory store with chat-protocol-aware extraction Markdown memory store at <dataDir>/memory/ with two extractors — heuristic regex for explicit "remember:" / "我是 X" markers, and a small-model LLM pass after each turn — folded into the system prompt so cross-chat preferences, role, and ongoing-work context survive restarts. Settings UI: - Memory tab lists entries, exposes a hand-edited MEMORY.md index, and shows an extraction history with per-attempt phase/skip/failure rows. - Memory model picker is inline next to the chat model picker (CLI and BYOK) so the choice "which fast model mines facts each turn?" sits next to the chat-model decision instead of a separate panel. The picker reuses the same SUGGESTED_MODELS table and "Custom..." pattern the chat picker uses. LLM extractor supports all four protocols (anthropic / openai / azure / google); pickProvider takes the chat agent id from the chat handler and constrains its auto-pick to the chat's protocol family — Claude Code chats no longer surprise users by silently extracting on whatever OpenAI key happens to be in media-config. When no matching key is configured the attempt records as 'skipped: no-provider' instead of quietly switching vendors. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(memory): keep hint outside <label> and disambiguate Model selectors The inline Memory model picker wrapped its hint paragraph inside the <label>, which made the hint's "API key" / "model" wording bleed into the <select>'s accessible name and broke Playwright's getByLabel('API key') / getByLabel('Model') strict-mode matching in the existing settings-api-protocol e2e suite. - Move the hint <p> out of the <label> in MemoryModelInline so the select's accessible name is just "Memory model". - Switch the chat-Model selectors in settings-api-protocol.test.ts from getByLabel('Model') to getByRole('combobox', { name: 'Model', exact: true }) so they no longer collide with the new "Memory model" select that sits next to the chat Model picker. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(memory): address review changes — BYOK wiring, MEMORY.md index, /v1, label wrapper Addresses the four blocking review threads on PR #999. 1. MemoryModelInline accessibility (mrcfps) The inline picker still wrapped its select + custom input + flash + hint inside a single <label>, which made the select's accessible name absorb every text descendant — including the "API key" / "model" hint copy. The previous fix moved only the hint outside; the reviewer asked for a non-label wrapper. Switch to <div className="field"> and associate just the short title with the controls via `aria-labelledby` / `aria-label`. The select's accessible name is now exactly "Memory model" so `getByLabel` strict-mode locators on the surrounding chat form stop cross-matching the memory copy. 2. Respect the hand-edited MEMORY.md index (mrcfps + codex) `composeMemoryBody()` was reading every *.md file in the memory dir, ignoring the index. Removing a `- [Name](id.md)` line had no effect on future prompts. Parse the index's `INDEX_LINK_RE` bullets and filter `listMemoryEntries()` to the linked id set, so the editor's "delete this line to disable injection" promise actually holds. 3. Versioned OpenAI-compatible base URLs (codex) `callOpenAI` and `callAnthropic` hard-coded `/v1` onto `provider.baseUrl`, breaking custom endpoints whose saved URL already includes `/v1` (`/v1/v1/chat/completions`). Apply the same conditional `appendVersionedApiPath` helper the chat proxy and connection-test routes already use. 4. Wire memory into BYOK / API-mode chats (mrcfps + codex) The previous PR's daemon-only memory hook never fired for BYOK, leaving the Memory tab + model picker as a no-op for that mode. Add the missing surface and wire it through ProjectView: - contracts: extend `composeSystemPrompt` with `memoryBody`, mirroring the daemon's local composer; add `MemorySystemPromptResponse` and the `attemptedLLM` flag on `ExtractMemoryResponse`. - daemon: expose `GET /api/memory/system-prompt` (returns the composed body) and turn `POST /api/memory/extract` into a two-phase endpoint — heuristic-only when only userMessage is supplied (pre-turn), LLM-only when assistantMessage is also supplied (post-turn), so the extraction-history doesn't double up. - web: ProjectView's BYOK branch now fetches the memory body before composing the system prompt, runs the heuristic extractor before the run (so "remember:" markers in this turn reach this turn's prompt), accumulates assistant text during streaming, and queues the LLM extractor on `onDone` — fire-and- forget so it never blocks the chat round-trip. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(memory): re-sync BYOK memory override when chat config drifts The inline memory-model picker captured `apiProtocol` / `chatApiKey` / `chatBaseUrl` / `chatApiVersion` into the saved override only at the moment the user clicked a model. If they later swapped the BYOK protocol tab, rotated the API key, or edited the base URL in the same settings flow, the daemon's background extractor kept calling the *old* vendor / credential — directly contradicting the picker's "borrows the surrounding chat picker's protocol, key, base URL, and api-version automatically" promise. Add a debounced effect that compares the persisted (masked) shape against the live chat props and re-PATCHes /api/memory/config when they drift. The masked config exposes `apiKeyTail` (last 4 chars), so key rotation is detectable without ever round-tripping the secret back to the browser. The 300 ms debounce coalesces the keystroke- granularity prop updates the parent settings dialog streams during its autosave loop, so a user editing the base URL doesn't trigger one PATCH per character. Background re-syncs are silent — the "Saved!" flash only fires for explicit user clicks, so the picker doesn't feel like it's fighting them as they edit unrelated chat fields. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(memory): thread BYOK chat config through /api/memory/extract default path Leaving the BYOK memory picker on "Same as chat" still broke the default LLM extraction path: `MemoryModelInline` clears the override for that option, both `/api/memory/extract` calls in `ProjectView` only sent the messages, and the daemon never persists BYOK creds, so `extractWithLLM(..., { chatAgentId: null })` always reached `pickProvider()` with no chat context and fell through to env / media-config — the wrong vendor for a BYOK chat that works for inference. Thread the live BYOK chat config through the extract endpoint as a per-call snapshot: - contracts: extend `ExtractMemoryRequest` with an optional `chatProvider` (provider/apiKey/baseUrl/apiVersion/model) and add `'chat-byok'` to the credentialSource enum. - daemon: parse + validate `chatProvider` on `/api/memory/extract` (provider must be one of the five known shapes) and forward to `extractWithLLM` as a new option. `pickProvider()` gets a new path 2 that uses the snapshot directly with the per-protocol fast-model default — so a memory pass on `gpt-4o` / `claude-sonnet-4-5` silently turns into a cheap `gpt-4o-mini` / `claude-haiku-4-5` call instead of paying chat-tier rates for sediment work. Override and CLI-agent-constrained paths still win when they apply. - web: `ProjectView` snapshots `apiProtocol` / `apiKey` / `baseUrl` / `apiVersion` from the live `AppConfig` on each BYOK extract call (both pre-turn heuristic-only and post-turn LLM phases). The picker's existing drift-resync effect already covers explicit overrides; this snapshot covers the implicit "Same as chat" default that the override flow can't reach. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(memory): treat empty apiKey on PATCH as a real clear MemoryModelInline silently re-PATCHes /api/memory/config whenever the surrounding BYOK chat creds drift. The previous reuse branch lumped `apiKey === ''` together with `apiKey === undefined`, so clearing the chat API key from the picker quietly preserved the old daemon-side secret and kept calling the provider on a stale credential. Distinguish four states for the apiKey field: - absent -> preserve stored secret (form re-save without re-typing) - '' -> clear stored secret (user removed it from the picker) - 'sk-...' -> replace - new provider -> ignore stored secret entirely Add tests/memory-config-route.test.ts covering all four cases. Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: Cursor <cursoragent@cursor.com> |
||
|
|
e11e86d468 |
feat(hyperframes): land HTML-in-Canvas across web + skills (#866)
* feat(hyperframes): land HTML-in-Canvas across web + skills Ships HTML-in-Canvas as a first-class HyperFrames video path: - 7 new video prompt templates (liquid glass, iPhone+MacBook, portal, shatter, magnetic, liquid background, text-cursor reveal). - skills/hyperframes/references/html-in-canvas.md, surfaced via SKILL.md description+triggers and the system-prompt pre-flight references list. - ChatPane starter prompts now branch by project kind and video model, so the hyperframes-html surface shows HTML-in-canvas-shaped prompts instead of the generic prototype trio. - NewProjectPanel propagates a picked template's model+aspect onto the project, and defaults videoModel to hyperframes-html when the hyperframes skill resolves for the video tab. Polish bundled in the same branch: - DesignFilesPanel empty state becomes a centered pill with a "New sketch" CTA; designFiles.empty copy simplified across 19 locales. - Topbar project title + meta render on one baseline row separated by a middot. - scripts/seed-test-projects.ts hardens daemon URL discovery against pnpm engine warnings on stdout. * fix(new-project): preserve explicit video model choice across tab revisits Latch a videoModelTouched guard once the user picks a model via the dropdown or via a template that declares one, so the hyperframes-html auto-default no longer silently overwrites the override when the Video tab is re-entered. Generated-By: looper 0.6.1 (runner=fixer, agent=claude-code) * fix(i18n): register hyperframes html-in-canvas templates, category, and tags Adds the seven new prompt-template ids, the "VFX / HTML-in-Canvas" category, and the new tag set to the de/ru/fr i18n bundles so the e2e localized-content coverage test passes. Generated-By: looper 0.6.1 (runner=fixer, agent=claude-code) * fix(daemon): inject html-in-canvas preflight for hyperframes runs The contracts-side derivePreflight() learned about references/html-in-canvas.md when this PR landed, but the daemon copy at apps/daemon/src/prompts/system.ts kept the older five-ref allowlist. server.ts:4138 wires composeSystemPrompt from the daemon copy into live chat runs, so the main HyperFrames flow this PR is meant to improve still wasn't auto-injecting the preflight directive in production. Mirror the html-in-canvas case into the daemon composer and lock it behind a daemon-side test so the two copies cannot drift again on this reference. The broader live-artifact preflight gap (artifact- schema / connector-policy / refresh-contract) is pre-existing drift and is intentionally out of scope here. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(web): restyle designs empty state as centered card on grid backdrop Swap the horizontal pill for a stacked card and add a faint grid backdrop so the empty designs surface reads as an intentional canvas rather than a gap. Title now wraps instead of truncating; container is taller. * fix(new-project): pin skillId to hyperframes when videoModel is hyperframes-html When the Video tab resolves its skill it used to fall back to `list[0]?.id` if no skill declared `default_for: video`. That list is built from an unsorted `readdir()` in apps/daemon/src/skills.ts, so a freshly mounted project could land on `video-shortform` even when the user had explicitly chosen the HyperFrames-HTML model (or one of the new `hyperframes-html-in-canvas-*` templates). The agent then ran without the hyperframes SKILL body or its `references/html-in-canvas.md` preflight — the exact regression PR #866 was meant to land. `skillIdForTab` now pins to `hyperframes` whenever the current video model is `hyperframes-html`, regardless of discovery order. Added a unit test that mounts both `video-shortform` and `hyperframes` (with hyperframes last, simulating the bad readdir order) and asserts the create payload routes through `hyperframes`. --------- Co-authored-by: Cursor <cursoragent@cursor.com> |
||
|
shangxinyu1andGitHub
|
b19aa6c907 |
Improve Codex CLI path fallback UX (#1205)
* Improve Codex CLI path fallback UX (#1193) * Handle ENOENT Codex shim fallback |
||
|
|
32fa0c23bb |
feat(daemon): Critique Theater Phase 6.2 (artifact extraction + endpoint) (#1085)
The orchestrator was leaving artifactPath = null on every shipped run because the SHIP <ARTIFACT> body never made it past the parser. Reviewers caught this on PR #1006: a rerun-style endpoint built on top of that null could not return a usable prior-art reference, and tests that synthesized artifactPath via insertCritiqueRun were hiding the gap rather than covering the feature. This PR closes that gap. The parser now hands the orchestrator a ShipArtifactPayload (round, mime, body) through a side-channel callback, and the orchestrator writes the bytes to <artifactsDir>/<projectId>/<runId>/ artifact.<ext> via a new artifact-writer module. The row's artifactPath is the absolute on-disk path. The web layer never sees that path: it fetches the bytes through GET /api/projects/:projectId/critique/:runId/artifact, which the new artifact-handler module serves with a mime-derived Content-Type, X-Content-Type-Options: nosniff, a CSP header for HTML and SVG, and the same cross-project leak guard pattern the interrupt handler uses. The body and mime intentionally never travel on the SSE wire. The SHIP PanelEvent (which doubles as the SSE payload shape) keeps its lightweight artifactRef, and the orchestrator strips body/mime before bus.emit, so a multi-megabyte artifact does not broadcast to every subscriber. The new orchestrator test asserts this explicitly. Defense in depth in the writer + handler: - mime allowlist with text/html, text/css, text/markdown, text/plain, application/json, image/svg+xml; everything else falls through to application/octet-stream + .bin so unknown payloads can't be misinterpreted as a known type; - UTF-8 byte-length cap, configurable via cfg.parserMaxBlockBytes, so multi-byte payloads can't sneak past a JS .length check; - atomic write through a sibling tmp file + rename so a daemon crash mid-write can't leave a half-written artifact under the canonical name; - path-traversal guard on the GET endpoint that resolves the row's artifactPath against the artifacts root and refuses anything that escapes it, refuses non-regular files (symlinks, dirs), and refuses files larger than the response cap. Folded in two non-blocking notes lefarcen left on PR #1016 (the contracts move) since persistence.ts was already in scope here: - P2: introduced CritiquePersistedStatus = CritiqueRunStatus | 'running' in the contracts package. CritiqueRunRow.status and CritiqueRunInsert. status now use it, and the inline `as CritiqueRunStatus | 'running'` widen in interrupt-handler.ts is gone. Public DTOs continue to use the terminal-only CritiqueRunStatus so a future endpoint can't leak a 'running' row through the wire. - P3: added AssertExhaustiveValues + a compile-time assertion that CRITIQUE_RUN_STATUSES covers every CritiqueRunStatus variant. Adding a value to ShipStatus or CritiqueRunStatus without updating the array now fails the build with a tuple naming the missing variants instead of silently dropping out of UI filters. Coverage: 174 critique tests across 14 files pass locally, including the new critique-artifact-writer (13 cases) and critique-artifact-endpoint (11 cases) suites, the inverted critique-lifecycle artifact-persistence test, and the orchestrator happy-path that asserts the SSE ship payload does NOT carry body or mime. Validated: pnpm guard, pnpm --filter @open-design/contracts build, pnpm --filter @open-design/daemon build (full tsc), pnpm --filter @open-design/web typecheck, pnpm --filter @open-design/daemon exec vitest run tests/critique (all green). This is step (b) of the four-step plan that PR #1006's closing comment laid out. Step (a) was the contracts move in PR #1016. Steps (c) (persist original_message_id / agent_id / model_id) and (d) (real rerun endpoint on top of (a)+(b)+(c)) follow. Co-authored-by: Nagendhra <nagendhra405@gmail.com> |
||
|
|
84f768d4a2 |
feat: add WeChat design system, login-flow skill, and fix API mode tool_calls bug (#1083)
* feat: add WeChat design system, login-flow skill, and fix API mode tool_calls bug - Add WeChat design system (design-systems/wechat/) with full brand spec including color palette, typography, and component rules for chat UI - Add login-flow skill (skills/login-flow/) for mobile authentication flows with P0 checklist, example HTML, and i18n registration across 3 locales - Fix DeepSeek V4 bug: API/BYOK mode (streamFormat=plain) models now receive a directive to emit only <artifact> HTML blocks and suppress tool_calls, since plain adapters proxy to external providers that cannot execute tools * fix: restore full server.ts and WeChat DESIGN.md from ad46d8cd commit Restore files that were corrupted in PR #1083 head branch. The WeChat DESIGN.md was reduced to a single line (filename only) and server.ts was reduced to ~1 line. Both are restored to their original ad46d8cd state with full content. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com> * fix: restore full server.ts and WeChat DESIGN.md from ad46d8cd Restore files corrupted in PR #1083: - apps/daemon/src/server.ts: restored 7106-line file - design-systems/wechat/DESIGN.md: restored 301-line WeChat design spec - skills/login-flow/SKILL.md: restored from local working state - skills/login-flow/example.html: restored 351-line example HTML * fix: only suppress tool_calls when streamFormat='plain' explicitly, remove nonexistent assets/template.html 1. streamFormat check now requires explicit 'plain' value instead of defaulting to 'plain' when undefined. This prevents normal tool-using chat runs from incorrectly inheriting the API/BYOK tool_calls suppression rule. 2. login-flow SKILL.md: removed reference to assets/template.html since that file does not exist in the skill bundle and derivePreflight() would inject a hard instruction to read it before any other tool, causing pre-flight to fail. * fix: thread streamFormat to composeSystemPrompt in server.ts call Previously the composeSystemPrompt call at line ~4940 omitted streamFormat, causing the composer to default to 'plain' and suppress tool_calls even for tool-using chat runs. Now streamFormat is passed through from the adapter definition so the API mode rule only fires when streamFormat='plain' is explicitly set. * fix: WeChat category metadata, font-family, and login-flow example interactivity WeChat DESIGN.md: - Add Category: Social & Messaging metadata so it appears correctly in picker - Fix font-family declaration: remove invalid -webkit-font-family prefix, use standard font-family so downstream CSS generation works correctly skills/login-flow/example.html: - Add password toggle click handler so show/hide actually works - Change Apple icon fill from hardcoded #fff to currentColor so it is visible on light backgrounds * fix: mirror streamFormat suppression in contracts composer and add WeChat i18n 1. packages/contracts/src/prompts/system.ts: Add streamFormat parameter to ComposeInput and ComposeInput interface, mirroring the same suppression rule from daemon prompts/system.ts. When streamFormat='plain' is passed, a directive is appended telling models not to emit tool_calls and to only output <artifact> HTML blocks. 2. apps/web/src/i18n/content.{ts,fr,ru}.ts: Add WeChat design system entries: - Add 'wechat' to DE/FR/RU_DESIGN_SYSTEM_IDS_WITH_EN_FALLBACK arrays - Add 'wechat' summary to DE/FR/RU_DESIGN_SYSTEM_SUMMARIES - Add 'Social & Messaging' category to DE/FR/RU_DESIGN_SYSTEM_CATEGORIES (matching the Category: Social & Messaging metadata in WeChat DESIGN.md) * fix: thread streamFormat='plain' into web composeSystemPrompt for api mode * test: focus localized content coverage on missing resources --------- Co-authored-by: Open Design Contributor <z@open-design.dev> Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com> Co-authored-by: mrcfps <mrc@powerformer.com> |
||
|
|
587c783dc0 |
feat(web): add Finalize design package + Continue in CLI buttons (#451) (#974)
* feat(daemon): expose resolvedDir on GET /api/projects/:id (#451 prereq)
Native projects (no metadata.baseDir) live at <projects root>/<id>, where
projects root is daemon-side state. The web client cannot reconstruct an
absolute path on its own, and shell.openPath on a relative path is
undefined behavior. Without resolvedDir, the upcoming Continue in CLI
button (#451) would render permanently disabled for native projects.
Mirrors PR #832's pattern of exposing designMdPath in its response.
Computed via the existing resolveProjectDir(...) helper. No behavior
change to existing callers; they ignore the new field.
Adds ProjectDetailResponse contract type and a focused projects-routes
test covering imported-folder, native, and unknown-id paths.
* feat(web): add parseProvenance helper for DESIGN.md staleness checks
Pure helper that extracts Project ID, design system, current artifact,
transcript message count, and generated UTC timestamp from the
`## Provenance` section emitted by the daemon's finalize synthesis
prompt (apps/daemon/src/finalize-design.ts). Used by useDesignMdState
to derive the Continue in CLI button's stale/fresh state without an
additional daemon endpoint.
Handles missing section, "none" sentinels for design system /
artifact, and malformed timestamps without throwing. Tests cover all
four branches.
* feat(web): add buildClipboardPrompt template for Continue in CLI
Inline single-source-of-truth template per #451 spec §3.4. Names the
project, the working directory, and the DESIGN.md-first operating
contract for the receiving `claude` CLI session. Trailing TODO is
the blank task slot the issue body specifies — left empty so the user
fills it in before submitting.
Also lands the shared copyToClipboard helper (jsdom-safe canonical path
+ execCommand fallback) so the new button and any future caller share
one fallback path, mirroring the inline pattern in FileViewer.tsx.
Tests cover happy-path field rendering, "none"/"unknown" sentinels
when DESIGN.md fields are absent, and both clipboard branches.
* feat(web): add useProjectDetail + useDesignMdState hooks
useProjectDetail wraps GET /api/projects/:id, surfacing the resolvedDir
field and falling back to metadata.baseDir for older daemons that don't
include it. Continue in CLI needs an absolute working directory so the
desktop bridge can openPath it; the web client never reconstructs the
path itself.
useDesignMdState fetches the project's file list, downloads DESIGN.md
when present, parses the Provenance section, and computes a stale
verdict by comparing the recorded generatedAt against the max mtime of
non-DESIGN.md files and the max conversation updatedAt. Drives the
button's three-state UI (disabled / fresh / stale) without a
daemon-side endpoint.
Tests cover happy path, fallback, and both stale branches plus the
pure computeStale helper for the null-timestamp edge case.
* feat(web): add useFinalizeProject hook with cancel + error-code mapping
Wraps POST /api/projects/:id/finalize/anthropic for the Finalize design
package button. Three concerns:
1. Lifecycle: idle → pending → success | error. Double-clicking the
button aborts the prior in-flight request before starting a new
one so the daemon never sees stacked finalize calls per project.
2. Cancellation: AbortController plumbed through fetch + a 130 s
timer (daemon timeout 120 s + 10 s buffer). Cancel returns to idle
cleanly — it's a user gesture, not an error surface.
3. Daemon error mapping: when the response is non-OK, body.error.code
drives the canonical user-facing toast string (table covers all
7 codes the daemon emits today plus a network-error catch-all).
body.error.details, when a string, surfaces alongside the category
message so account-usage-cap responses (Anthropic 400 →
UPSTREAM_UNAVAILABLE) can show the upstream's own reason instead
of just the daemon's category label — committed to lefarcen on
#450 verification reply.
Tests cover request body shape, all 8 error codes via it.each, the
network-error path, the details-surfacing branch, the cancel ⇒ idle
flow, and the unknown-code → catch-all message branch.
* feat(web): add useTerminalLaunch with electron/web detection
Capability-detected wrapper around window.electronAPI.openPath. On
desktop the bridge forwards to shell.openPath, which opens the OS
file manager at the project working directory (per Electron's
contract for directory paths — it is NOT a terminal launcher;
spawning a terminal application is deferred per #451 Non-goals). On
browser builds the hook reports web-fallback so the caller renders
a manual-instruction toast naming the working directory.
Treats any non-empty string return from shell.openPath as ok: false
so platform-specific failures surface the manual fallback toast.
Behavior is exercised end-to-end by the upcoming
ContinueInCliButton tests.
* feat(desktop): expose shell.openPath via electronAPI bridge
Adds an openPath bridge method that the Continue in CLI button (#451)
uses to surface the project working directory in the OS file manager.
shell.openPath is part of Electron's contract and resolves to '' on
success / a non-empty error string on failure; the IPC handler
forwards the result so the renderer can decide between the success
toast and the manual fallback toast without a separate error channel.
Empty / non-string inputs short-circuit to a self-describing error
string so the renderer never needs to worry about undefined-input
crashes from the main process.
Web side: extracts Window.electronAPI into a single global declaration
at apps/web/src/types/electron.d.ts so future bridge methods land in
one place. Two pre-existing inline declare-global blocks
(NewProjectPanel.tsx, providers/registry.ts) are deleted in favor of
that single source of truth — the inline ones each carried a partial
shape of the bridge and were diverging from the desktop preload.
* feat(web): add FinalizeDesignButton, ContinueInCliButton, ProjectActionsToolbar
Project-level toolbar that hosts the two new actions from #451.
Mounted between AppChromeHeader and the chat/workspace split (wiring
lands in the next commit). Per-file actions (Export PDF/PPTX/ZIP,
Deploy) stay in the FileViewer share menu.
FinalizeDesignButton has three idle labels driven by DESIGN.md
existence + staleness, plus a pending state with a spinner and a
cancel link that maps to useFinalizeProject's AbortController. Error
toasts are owned by ProjectView so the button doesn't carry its own
toast surface.
ContinueInCliButton renders disabled with a Finalize-pointing
tooltip when DESIGN.md is missing (so the workflow is discoverable
rather than hidden), enabled when fresh, and enabled with a stale
chip otherwise. Chip text is the spec's canonical "Spec is stale —
regenerate?" — N-turns-ago is deferred per spec §4.6.
Toast.tsx is a tiny transient component that mirrors
PromptTemplatePreviewModal's state-based toast pattern; supports a
secondary details line so daemon error envelopes that carry an
upstream explanation (e.g. Anthropic account-usage cap) can surface
the real reason alongside the daemon's category label.
CSS appends one block to apps/web/src/index.css mirroring the
existing app-project-title token usage; no CSS modules in this
repo (verified by grep).
* test(web): cover ContinueInCliButton states + interaction wiring
Three rendered states (DESIGN.md missing → disabled with the
Finalize-pointing tooltip; DESIGN.md fresh → enabled, no chip;
DESIGN.md stale → enabled with the canonical "Spec is stale —
regenerate?" chip), plus three onClick branches (no-op when
disabled, fires once when fresh, fires once when stale).
Click-handler integration with clipboard / shell.openPath / toast
lives in ProjectView (the button is presentational and takes the
handler in via props), so those are covered by Phase K's wiring +
the manual smoke test rather than the per-component test.
* feat(web): wire Continue in CLI + Finalize buttons into ProjectView
Mounts the new project-actions toolbar between AppChromeHeader and
the chat/workspace split, hidden when workspaceFocused so the
focus-mode artifact view stays uncluttered.
Wires the four hooks (useProjectDetail, useDesignMdState,
useFinalizeProject, useTerminalLaunch) to a single shared toast
surface. handleFinalize reads the request body from the existing
config: AppConfig prop and uses effectiveMaxTokens(config) to match
the chat-flow's maxTokens defaulting; on success it refreshes
useDesignMdState so the toolbar re-renders with the new chip state.
handleContinueInCli builds the literal clipboard prompt, copies it,
opens the working directory via shell.openPath on desktop /
falls through to a manual-instruction toast on browser, and surfaces
shell.openPath failures with a fallback toast that names the path.
Errors lift into the same toast surface (a useEffect tied to
finalize.error) so the daemon's category message + body.error.details
reach the user as the spec's two-line render — covered by hook test
16a in the prior commit.
⌘+Shift+K (mac) / Ctrl+Shift+K (others) is the keyboard
accelerator for Continue in CLI; capture-phase, platform-gated,
no-op when DESIGN.md is missing. Mirrors the existing FileWorkspace
shortcut idiom and does not collide with ⌘+P (Quick Switcher).
* fix(web): distinguish timeout abort from user cancel in useFinalizeProject
Addresses codex P2 finding on PR #974: the catch block treated every
AbortError as a user-initiated cancel and reset to idle silently. If
the internal 130 s timeout fired, users saw no failure signal but the
daemon's synthesis call may still have been in flight.
Adds a timedOutRef set inside the setTimeout callback before
controller.abort(), and branches in the catch: timeout → status
'error' with new TIMEOUT code ("Finalize timed out after 130 s. The
daemon may still be running."), user cancel → existing idle reset.
Reset the ref at the start of every trigger() so a previous timeout
doesn't poison the next call.
Adds one test using vi.useFakeTimers() that advances past 130_001 ms
and asserts the TIMEOUT error surface.
* fix(web): surface clipboard failures by rendering the prompt in the toast
Addresses codex P2 finding on PR #974: handleContinueInCli ignored
copyToClipboard's return value, so when both clipboard paths failed
(restricted browser context / insecure origin) the toast still said
"paste the prompt" though nothing had been copied — leaving users
with no manual-copy recourse in exactly the environments where the
fallback should help.
handleContinueInCli now branches on copyToClipboard's boolean return.
On failure the toast renders the prepared prompt in a scrollable
<pre> block and pins itself open (no auto-dismiss) so the user has
time to select-and-copy manually. Includes a Dismiss button + the
working directory in the secondary details line so the user has the
information needed to proceed.
The folder-open call is skipped on copy failure because there's
nothing to paste yet; the user copies first, then re-clicks Continue
in CLI when they're ready.
Toast component grows an optional Updating VS Code Server to version 41dd792b5e652393e7787322889ed5fdc58bd75b
Removing previous installation...
Installing VS Code Server for Linux x64 (41dd792b5e652393e7787322889ed5fdc58bd75b)
Downloading: 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 2% 2% 2% 2% 2% 2% 2% 2% 2% 2% 2% 2% 2% 2% 2% 2% 2% 2% 3% 3% 3% 3% 3% 3% 3% 3% 3% 3% 3% 3% 3% 3% 3% 3% 3% 3% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 5% 5% 5% 5% 5% 5% 5% 5% 5% 5% 5% 5% 5% 5% 5% 5% 5% 5% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 8% 8% 8% 8% 8% 8% 8% 8% 8% 8% 8% 8% 8% 8% 8% 8% 8% 8% 9% 9% 9% 9% 9% 9% 9% 9% 9% 9% 9% 9% 9% 9% 9% 9% 9% 10% 10% 10% 10% 10% 10% 10% 10% 10% 10% 10% 10% 10% 10% 10% 10% 10% 10% 11% 11% 11% 11% 11% 11% 11% 11% 11% 11% 11% 11% 11% 11% 11% 11% 11% 11% 12% 12% 12% 12% 12% 12% 12% 12% 12% 12% 12% 12% 12% 12% 12% 12% 12% 12% 13% 13% 13% 13% 13% 13% 13% 13% 13% 13% 13% 13% 13% 13% 13% 13% 13% 13% 14% 14% 14% 14% 14% 14% 14% 14% 14% 14% 14% 14% 14% 14% 14% 14% 14% 14% 15% 15% 15% 15% 15% 15% 15% 15% 15% 15% 15% 15% 15% 15% 15% 15% 15% 15% 16% 16% 16% 16% 16% 16% 16% 16% 16% 16% 16% 16% 16% 16% 16% 16% 16% 16% 17% 17% 17% 17% 17% 17% 17% 17% 17% 17% 17% 17% 17% 17% 17% 17% 17% 17% 18% 18% 18% 18% 18% 18% 18% 18% 18% 18% 18% 18% 18% 18% 18% 18% 18% 19% 19% 19% 19% 19% 19% 19% 19% 19% 19% 19% 19% 19% 19% 19% 19% 19% 19% 20% 20% 20% 20% 20% 20% 20% 20% 20% 20% 20% 20% 20% 20% 20% 20% 20% 20% 21% 21% 21% 21% 21% 21% 21% 21% 21% 21% 21% 21% 21% 21% 21% 21% 21% 21% 22% 22% 22% 22% 22% 22% 22% 22% 22% 22% 22% 22% 22% 22% 22% 22% 22% 22% 23% 23% 23% 23% 23% 23% 23% 23% 23% 23% 23% 23% 23% 23% 23% 23% 23% 23% 24% 24% 24% 24% 24% 24% 24% 24% 24% 24% 24% 24% 24% 24% 24% 24% 24% 24% 25% 25% 25% 25% 25% 25% 25% 25% 25% 25% 25% 25% 25% 25% 25% 25% 25% 25% 26% 26% 26% 26% 26% 26% 26% 26% 26% 26% 26% 26% 26% 26% 26% 26% 26% 26% 27% 27% 27% 27% 27% 27% 27% 27% 27% 27% 27% 27% 27% 27% 27% 27% 27% 28% 28% 28% 28% 28% 28% 28% 28% 28% 28% 28% 28% 28% 28% 28% 28% 28% 28% 29% 29% 29% 29% 29% 29% 29% 29% 29% 29% 29% 29% 29% 29% 29% 29% 29% 29% 30% 30% 30% 30% 30% 30% 30% 30% 30% 30% 30% 30% 30% 30% 30% 30% 30% 30% 31% 31% 31% 31% 31% 31% 31% 31% 31% 31% 31% 31% 31% 31% 31% 31% 31% 31% 32% 32% 32% 32% 32% 32% 32% 32% 32% 32% 32% 32% 32% 32% 32% 32% 32% 32% 33% 33% 33% 33% 33% 33% 33% 33% 33% 33% 33% 33% 33% 33% 33% 33% 33% 33% 34% 34% 34% 34% 34% 34% 34% 34% 34% 34% 34% 34% 34% 34% 34% 34% 34% 34% 35% 35% 35% 35% 35% 35% 35% 35% 35% 35% 35% 35% 35% 35% 35% 35% 35% 35% 36% 36% 36% 36% 36% 36% 36% 36% 36% 36% 36% 36% 36% 36% 36% 36% 36% 37% 37% 37% 37% 37% 37% 37% 37% 37% 37% 37% 37% 37% 37% 37% 37% 37% 37% 38% 38% 38% 38% 38% 38% 38% 38% 38% 38% 38% 38% 38% 38% 38% 38% 38% 38% 39% 39% 39% 39% 39% 39% 39% 39% 39% 39% 39% 39% 39% 39% 39% 39% 39% 39% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 41% 41% 41% 41% 41% 41% 41% 41% 41% 41% 41% 41% 41% 41% 41% 41% 41% 41% 42% 42% 42% 42% 42% 42% 42% 42% 42% 42% 42% 42% 42% 42% 42% 42% 42% 42% 43% 43% 43% 43% 43% 43% 43% 43% 43% 43% 43% 43% 43% 43% 43% 43% 43% 43% 44% 44% 44% 44% 44% 44% 44% 44% 44% 44% 44% 44% 44% 44% 44% 44% 44% 44% 45% 45% 45% 45% 45% 45% 45% 45% 45% 45% 45% 45% 45% 45% 45% 45% 45% 46% 46% 46% 46% 46% 46% 46% 46% 46% 46% 46% 46% 46% 46% 46% 46% 46% 46% 47% 47% 47% 47% 47% 47% 47% 47% 47% 47% 47% 47% 47% 47% 47% 47% 47% 47% 48% 48% 48% 48% 48% 48% 48% 48% 48% 48% 48% 48% 48% 48% 48% 48% 48% 48% 49% 49% 49% 49% 49% 49% 49% 49% 49% 49% 49% 49% 49% 49% 49% 49% 49% 49% 50% 50% 50% 50% 50% 50% 50% 50% 50% 50% 50% 50% 50% 50% 50% 50% 50% 50% 51% 51% 51% 51% 51% 51% 51% 51% 51% 51% 51% 51% 51% 51% 51% 51% 51% 51% 52% 52% 52% 52% 52% 52% 52% 52% 52% 52% 52% 52% 52% 52% 52% 52% 52% 52% 53% 53% 53% 53% 53% 53% 53% 53% 53% 53% 53% 53% 53% 53% 53% 53% 53% 53% 54% 54% 54% 54% 54% 54% 54% 54% 54% 54% 54% 54% 54% 54% 54% 54% 54% 55% 55% 55% 55% 55% 55% 55% 55% 55% 55% 55% 55% 55% 55% 55% 55% 55% 55% 56% 56% 56% 56% 56% 56% 56% 56% 56% 56% 56% 56% 56% 56% 56% 56% 56% 56% 57% 57% 57% 57% 57% 57% 57% 57% 57% 57% 57% 57% 57% 57% 57% 57% 57% 57% 58% 58% 58% 58% 58% 58% 58% 58% 58% 58% 58% 58% 58% 58% 58% 58% 58% 58% 59% 59% 59% 59% 59% 59% 59% 59% 59% 59% 59% 59% 59% 59% 59% 59% 59% 59% 60% 60% 60% 60% 60% 60% 60% 60% 60% 60% 60% 60% 60% 60% 60% 60% 60% 60% 61% 61% 61% 61% 61% 61% 61% 61% 61% 61% 61% 61% 61% 61% 61% 61% 61% 61% 62% 62% 62% 62% 62% 62% 62% 62% 62% 62% 62% 62% 62% 62% 62% 62% 62% 62% 63% 63% 63% 63% 63% 63% 63% 63% 63% 63% 63% 63% 63% 63% 63% 63% 63% 64% 64% 64% 64% 64% 64% 64% 64% 64% 64% 64% 64% 64% 64% 64% 64% 64% 64% 65% 65% 65% 65% 65% 65% 65% 65% 65% 65% 65% 65% 65% 65% 65% 65% 65% 65% 66% 66% 66% 66% 66% 66% 66% 66% 66% 66% 66% 66% 66% 66% 66% 66% 66% 66% 67% 67% 67% 67% 67% 67% 67% 67% 67% 67% 67% 67% 67% 67% 67% 67% 67% 67% 68% 68% 68% 68% 68% 68% 68% 68% 68% 68% 68% 68% 68% 68% 68% 68% 68% 68% 69% 69% 69% 69% 69% 69% 69% 69% 69% 69% 69% 69% 69% 69% 69% 69% 69% 69% 70% 70% 70% 70% 70% 70% 70% 70% 70% 70% 70% 70% 70% 70% 70% 70% 70% 70% 71% 71% 71% 71% 71% 71% 71% 71% 71% 71% 71% 71% 71% 71% 71% 71% 71% 71% 72% 72% 72% 72% 72% 72% 72% 72% 72% 72% 72% 72% 72% 72% 72% 72% 72% 73% 73% 73% 73% 73% 73% 73% 73% 73% 73% 73% 73% 73% 73% 73% 73% 73% 73% 74% 74% 74% 74% 74% 74% 74% 74% 74% 74% 74% 74% 74% 74% 74% 74% 74% 74% 75% 75% 75% 75% 75% 75% 75% 75% 75% 75% 75% 75% 75% 75% 75% 75% 75% 75% 76% 76% 76% 76% 76% 76% 76% 76% 76% 76% 76% 76% 76% 76% 76% 76% 76% 76% 77% 77% 77% 77% 77% 77% 77% 77% 77% 77% 77% 77% 77% 77% 77% 77% 77% 77% 78% 78% 78% 78% 78% 78% 78% 78% 78% 78% 78% 78% 78% 78% 78% 78% 78% 78% 79% 79% 79% 79% 79% 79% 79% 79% 79% 79% 79% 79% 79% 79% 79% 79% 79% 79% 80% 80% 80% 80% 80% 80% 80% 80% 80% 80% 80% 80% 80% 80% 80% 80% 80% 80% 81% 81% 81% 81% 81% 81% 81% 81% 81% 81% 81% 81% 81% 81% 81% 81% 81% 81% 82% 82% 82% 82% 82% 82% 82% 82% 82% 82% 82% 82% 82% 82% 82% 82% 82% 83% 83% 83% 83% 83% 83% 83% 83% 83% 83% 83% 83% 83% 83% 83% 83% 83% 83% 84% 84% 84% 84% 84% 84% 84% 84% 84% 84% 84% 84% 84% 84% 84% 84% 84% 84% 85% 85% 85% 85% 85% 85% 85% 85% 85% 85% 85% 85% 85% 85% 85% 85% 85% 85% 86% 86% 86% 86% 86% 86% 86% 86% 86% 86% 86% 86% 86% 86% 86% 86% 86% 86% 87% 87% 87% 87% 87% 87% 87% 87% 87% 87% 87% 87% 87% 87% 87% 87% 87% 87% 88% 88% 88% 88% 88% 88% 88% 88% 88% 88% 88% 88% 88% 88% 88% 88% 88% 88% 89% 89% 89% 89% 89% 89% 89% 89% 89% 89% 89% 89% 89% 89% 89% 89% 89% 89% 90% 90% 90% 90% 90% 90% 90% 90% 90% 90% 90% 90% 90% 90% 90% 90% 90% 90% 91% 91% 91% 91% 91% 91% 91% 91% 91% 91% 91% 91% 91% 91% 91% 91% 91% 92% 92% 92% 92% 92% 92% 92% 92% 92% 92% 92% 92% 92% 92% 92% 92% 92% 92% 93% 93% 93% 93% 93% 93% 93% 93% 93% 93% 93% 93% 93% 93% 93% 93% 93% 93% 94% 94% 94% 94% 94% 94% 94% 94% 94% 94% 94% 94% 94% 94% 94% 94% 94% 94% 95% 95% 95% 95% 95% 95% 95% 95% 95% 95% 95% 95% 95% 95% 95% 95% 95% 95% 96% 96% 96% 96% 96% 96% 96% 96% 96% 96% 96% 96% 96% 96% 96% 96% 96% 96% 97% 97% 97% 97% 97% 97% 97% 97% 97% 97% 97% 97% 97% 97% 97% 97% 97% 97% 98% 98% 98% 98% 98% 98% 98% 98% 98% 98% 98% 98% 98% 98% 98% 98% 98% 98% 99% 99% 99% 99% 99% 99% 99% 99% 99% 99% 99% 99% 99% 99% 99% 99% 99% 99%100%100%
Unpacking: 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29% 30% 31% 32% 33% 34% 35% 36% 37% 38% 39% 40% 41% 42% 43% 44% 45% 46% 47% 48% 49% 50% 51% 52% 53% 54% 55% 56% 57% 58% 59% 60% 61% 62% 63% 64% 65% 66% 67% 68% 69% 70% 71% 72% 73% 74% 75% 76% 77% 78% 79% 80% 81% 82% 83% 84% 85% 86% 87% 88% 89% 90% 91% 92% 93% 94% 95% 96% 97% 98% 99%100%
Unpacked 4009 files and folders to /home/bryan/.vscode-server/bin/41dd792b5e652393e7787322889ed5fdc58bd75b.
Looking for compatibility check script at /home/bryan/.vscode-server/bin/41dd792b5e652393e7787322889ed5fdc58bd75b/bin/helpers/check-requirements.sh
Running compatibility check script
Compatibility check successful (0) prop and the auto-dismiss
TTL is suppressed whenever code is present. CSS adds .od-toast-code
(monospace, max-height 240 with overflow-auto) and .od-toast-dismiss
styling.
Six new Toast tests cover details rendering, code rendering,
no-auto-dismiss when code is present, auto-dismiss when code is
absent, and the Dismiss button affordance.
* fix(web): make ContinueInCliButton disabled-state guidance visible
Addresses mrcfps's PR #974 review: native <button disabled> does
not fire hover/focus events in browsers we ship against, so a
`title` tooltip on the disabled button never surfaces. The only
guidance for the missing-DESIGN.md state was effectively invisible —
defeating the spec's "discoverable, not hidden" intent.
Renders the help text as a visible sibling <span> next to the
disabled button instead. Adds aria-describedby pointing the button
at the hint's id so assistive tech announces the explanation when
the disabled button gets focus. The native `disabled` attribute
stays so the button still can't be clicked or submitted.
CSS adds .project-actions-disabled-hint (muted italic, 11.5px,
matches the existing meta/secondary text style on this surface).
Test asserts the role="note" hint is in the DOM with the canonical
text and that the button's aria-describedby links to its id.
* fix(web): keep ProjectActionsToolbar at natural height inside the .app grid
The .app container was `grid-template-rows: auto 1fr` — only two
rows. Adding ProjectActionsToolbar as a third child between
AppChromeHeader and the chat/workspace split made the toolbar the
2nd grid item, so it took the `1fr` row (filling roughly half the
viewport) while the split got pushed into an implicit auto row at
its content's natural height. Surfaced as a screenshot from Bryan
showing the toolbar's background bleeding across most of the screen.
Extend grid-template-rows to `auto auto 1fr` and pin the split to
`grid-row: 3` explicitly. Now:
- Toolbar visible: row 1 = header (auto), row 2 = toolbar (auto),
row 3 = split (1fr, fills remaining viewport).
- Toolbar hidden via hidden=workspaceFocused → ProjectActionsToolbar
returns null, row 2 collapses to 0px (auto with no content), split
still fills row 3.
No JS changes; existing 609 tests still green.
* fix(web): guard useFinalizeProject state writes against superseded triggers
Addresses mrcfps's PR #974 P1 review on useFinalizeProject.ts:132
(also called out as P1.3 in lefarcen's deep-dive review).
Calling trigger() twice in quick succession aborted the first
controller and swapped abortRef to the new one, but the first
request's later AbortError catch still unconditionally called
setStatus('idle') / setError(null). That cleared the spinner and
re-enabled both toolbar buttons while the replacement finalize was
still pending — defeating the de-duplication this hook was meant to
enforce.
Adds an isCurrent() closure (`abortRef.current === controller`)
and gates every state-write site after the await: success path,
non-OK envelope path, AbortError-timeout, AbortError-cancel, and
network-error all bail early when the trigger has been superseded.
Per mrcfps: "make every state write request-scoped."
Regression test triggers twice in quick succession with a
never-resolving fetch, awaits the first promise (it rejects with
AbortError), and asserts status stays 'pending' rather than
collapsing to 'idle' under the replacement's lifetime.
* fix(desktop): allowlist-validate shell.openPath against registered project roots
Addresses mrcfps's PR #974 P1 review on runtime.ts:305 (also called
out as P1.2 in lefarcen's deep-dive review): the new
`shell:open-path` IPC handler accepted any renderer-supplied
string and forwarded it straight into Electron's `shell.openPath`,
widening the renderer→main trust boundary so XSS or a compromised
renderer dependency could open arbitrary local paths to the user.
Adds an explicit gate around the bridge:
1. validateExistingDirectory(p) — floor check that rejects empty
strings, relative paths, files, apps, and non-existent paths;
realpath-resolves so symlink games can't be used to register
one path and reach another.
2. createProjectRootGate() — Set-backed allowlist of
daemon-validated project working directories. The renderer
calls registerProjectRoot(absDir) once per project mount via
a new IPC method (preload bridge); the main process only
opens paths that pass both the floor check and the allowlist.
ProjectView wires the registration via a useEffect tied to
projectDetail.resolvedDir, so the active project's daemon-supplied
working directory is always the one being approved (not a renderer-
synthesized string).
Threat-model caveat documented in the runtime.ts comment block: an
attacker that fully controls the renderer can also call register
with arbitrary paths. Closing that gap fully requires a daemon-side
round-trip to derive the canonical resolvedDir from the daemon's
project registry, which is deferred to keep this PR focused.
Today's allowlist still defends against accidental misuse, bugs,
and common XSS payloads that don't know to call register first.
Adds apps/packaged/tests/desktop-project-root-gate.test.ts with 13
cases: floor-validation rejection cases (empty / relative / missing
/ file), happy-path resolution, symlink realpath canonicalization,
and the allowlist's register/isApproved/reset semantics. Mirrors
the existing apps/packaged/tests/desktop-url-allowlist.test.ts
pattern from PR #911 — the packaged workspace hosts the test
because apps/desktop has no vitest setup yet.
* fix(daemon): wire request-lifecycle abort signal through finalize route
Addresses mrcfps's PR #974 P1 review on
apps/daemon/src/server.ts:3831-3837 (also called out as P1.1 in
lefarcen's deep-dive review): `POST /api/projects/:id/finalize/anthropic`
called `finalizeDesignPackage(...)` without threading any
request-lifecycle abort, so cancelling the browser fetch only
aborted the UI-side request — the daemon's 60–120 s Anthropic call
kept running and still wrote DESIGN.md after the UI returned to idle.
Adds an AbortController inside the route handler, fired from
`res.on('close')`, and threads its signal into the existing
`signal?: AbortSignal` parameter on `FinalizeOptions`
(finalize-design.ts:70). `callAnthropicWithRetry` already passes
the signal through to the underlying fetch, so a client disconnect
now propagates all the way to the Anthropic SDK call.
Listener-event choice: `res.on('close')` is the canonical event
for "client disconnected before response was sent" in Express. The
common alternative `req.on('close')` fires whenever the *request*
stream finishes — for POST routes that means as soon as the
body-parser middleware drains the body, well before the route does
any work. Using req.on('close') would have flipped the abort
controller in every successful run; the test caught this empirically.
Caveat documented in the route's comment block: an abort fired
*after* the upstream response has been received but *before* the
atomic write completes still allows the write to land. The SDK
contract bounds the network round-trip, not the post-network disk
handoff.
Adds tests/finalize-route-abort.test.ts: spins up the test server,
mocks global fetch to capture the daemon-side AbortSignal at the
Anthropic call, sends the request via raw http (so we can destroy
the underlying socket), waits until the server reaches the
Anthropic call, then destroys the socket and asserts that the
daemon-side signal received an abort event within 5 s.
Three pre-existing project-watchers chokidar tests show flaky
timeouts under full-suite concurrency but pass in isolation;
unrelated to this fix.
* fix(daemon): refactor finalize-route-abort test to satisfy strict TS narrowing
The CI typecheck (`pnpm --filter @open-design/daemon typecheck`,
which runs both tsconfig.json and tsconfig.tests.json) caught what
my pre-push validation missed: TS narrowed `capturedSignal` to
literal `null` because vitest's mockImplementation closure can't
prove its callback runs, leaving the bare `let capturedSignal:
AbortSignal | null = null` permanently typed at its initial value.
At line 184 (`expect(capturedSignal?.aborted).toBe(true)`) the
right-hand side of the optional-chain became unreachable, and TS
flagged it as `Property 'aborted' does not exist on type 'never'`.
Switches to the standard ref-object pattern
(`const capture: { signal: AbortSignal | null } = { signal: null }`).
TS narrows let bindings inside closures conservatively but treats
object-property writes as opaque, so `capture.signal` reads
correctly across the closure boundary. Logic is unchanged.
(Pre-push oversight: ran `pnpm --filter @open-design/web typecheck`
but not the full repo `pnpm typecheck` after the daemon test
landed; the daemon's own typecheck would have caught this. Adding
`pnpm typecheck` back into the standard pre-push checklist.)
* fix(desktop): make shell.openPath gate daemon-controlled and reject .app bundles
Addresses lefarcen + mrcfps PR #974 P1 reviews on the previous path
allowlist (commit
|
||
|
|
617fb043fe |
feat(settings): add fetch models button for BYOK providers (#1034)
* feat(settings): add fetch models button for BYOK providers * fix(settings): exclude Ollama from fetch models, add manual-entry hint * fix(provider-models): classify non-JSON upstream errors by HTTP status * fix(i18n): drop redundant English overrides from non-English locales * fix(provider-models): allow ollama through allowlist, return unsupported_protocol --------- Co-authored-by: haolin122 <hl6593@nyu.edu> |
||
|
Justin GaoandGitHub
|
49b55761d3 |
feat(daemon+web): add install/uninstall for skills & design systems (#1003)
* feat(daemon+web): add install/uninstall for skills & design systems (#497 Phase 2) Phase 2 of the Library Settings feature. Adds the ability to install skills and design systems from GitHub repos or local paths, and uninstall user-installed items. Built-in items remain read-only. Daemon: - Multi-directory scanning: built-in + ~/.open-design/{skills,design-systems} - Install from GitHub (git clone --depth 1) or local path (symlink) - Uninstall removes cloned dirs or symlinks under user dir only - 4 new API routes: POST install + DELETE for skills and design systems - Reuse isBlocked() from linked-dirs.ts for path safety Web: - Install form with GitHub/Local tab switch in Library settings - Source badge ("Built-in" / "Installed") on cards - Uninstall button (trash icon) on installed items only - Native-language i18n for all 17 locales Contracts: - Add source field to SkillSummary and DesignSystemSummary - Add InstallInput, InstallSkillResponse, InstallDesignSystemResponse, UninstallResponse types * fix(daemon): follow symlinks in scanners and respect OD_DATA_DIR for user libraries Two fixes from PR review feedback: 1. Scanners (skills.ts, design-systems.ts) used Dirent.isDirectory() which skips symlinks on macOS/Linux, so locally-installed items never appeared in API responses. Now also accepts isSymbolicLink() entries. 2. User library directories were hardcoded under os.homedir(), ignoring OD_DATA_DIR. This broke test isolation and packaged runs. Now derived from RUNTIME_DATA_DIR so they respect the data dir override. |
||
|
|
039cc3d674 |
fix(settings): add install onboarding links for unavailable local CLIs (#985)
* fix(settings): add install onboarding links for unavailable local CLIs * fix(settings): rename Claude config dir label to config directory --------- Co-authored-by: mrcfps <mrc@powerformer.com> |
||
|
CIoudherdandGitHub
|
724d071c01 |
feat: add design file rename support (#894)
* feat(contracts): add project file rename contract * feat(daemon): add safe project file rename API * feat(web): support renaming design files * fix(daemon): handle case-only file renames * fix(web): prevent rename collisions with pending sketches * fix(daemon): preserve source names during rename * test(daemon): cover rename symlink escapes * fix(daemon): avoid clobbering rename targets * test(web): align rename tests after rebase * test(web): align rename tests with latest main |
||
|
Tom HuangandGitHub
|
643d0cf637 |
feat: add scheduled routines for unattended agent runs (#1033)
* feat: add scheduled routines for unattended agent runs
Generalizes Orbit's single hard-coded daily-digest scheduler into
user-defined routines: each one fires on a schedule (hourly / daily /
weekdays / weekly with IANA timezone) and starts a fresh agent
conversation, either inside an existing project or in a new project
minted on the spot.
Backend:
- New RoutineService with timezone-aware nextRunAt computed via
Intl.DateTimeFormat (no new dependency); two-pass tzWallToUtc so
DST transitions stay correct. Each fire chains rescheduleOne in
finally() to keep the cadence alive.
- routines + routine_runs SQLite tables; schedule_json is the
authoritative form, with legacy schedule_kind/value kept populated.
- /api/routines CRUD + /api/routines/:id/run + /api/routines/:id/runs.
- Run handler resolves agent (routine override -> app config -> first
available), creates project (or reuses configured one) and a fresh
conversation per fire, then dispatches into startChatRun.
UI (Settings -> Routines):
- Pill-chip schedule kind picker, time + timezone fields, weekday
picker for Weekly. Live preview line ("Runs daily at 9:00 AM GMT+8").
- Routine list with inline status pill, next/last meta, expandable run
history; each history row links into the project the run wrote to
via the existing router primitive.
* fix(daemon): swallow trailing finally rejection for inflight cleanup
Without a terminal `.catch`, the promise returned by `promise.finally(...)`
mirrors the original rejection and produces an unhandled rejection — fatal
in modern Node — when the run handler rejects before producing a start
handle. Callers still see the rejection on the returned `promise`.
Generated-By: looper 0.6.1 (runner=fixer, agent=claude-code)
* fix(daemon): handle DST spring-forward gap in tzWallToUtc
The two-pass conversion picked the pre-gap candidate when the requested
wall time fell inside a spring-forward gap (e.g. 02:30 in
America/New_York on 2026-03-08), so the resulting instant rendered back
as 01:30 local and a 02:30 routine fired an hour early on the
transition day. Routines are local wall-clock schedules, so firing
before the requested time breaks the contract.
Now we round-trip both candidates through partsInTimezone, return the
one whose wall-clock matches the request, and on a gap day where
neither matches return the later candidate so the routine fires at the
first valid post-gap instant on the same day.
Generated-By: looper 0.6.1 (runner=fixer, agent=claude-code)
* fix(daemon): preserve both wall-time candidates on DST fall-back day
On a fall-back day, the requested wall time inside the repeated hour
(e.g. 01:30 America/New_York) maps to two distinct UTC instants. The
previous tzWallToUtc collapsed them to the first (pre-transition) one,
so a daemon that woke between the two instants would skip the second
01:30 entirely and fire a day late once per fall-back. Replace it with
tzWallToUtcCandidates (returns all valid instants, ascending) plus a
gap-only fallback for spring-forward, and have nextWallTimeMatching
walk both ambiguous candidates before advancing to the next day. Adds
fixtures for the repeated-hour case so the intended behavior stays
locked in.
Generated-By: looper 0.6.1 (runner=fixer, agent=claude-code)
* fix(web): make routine timezone picker IANA-complete and DST-truthful
The timezone dropdown was a hardcoded subset, but the backend validator
accepts any IANA zone — so users could not pick zones like
`America/Phoenix` or `Africa/Johannesburg` unless they happened to be
local. And `gmtLabel()` always derived the offset from `new Date()`,
which drifted seasonally for DST-observing zones (a New York routine
created in winter rendered `GMT-5` while it would actually fire on
`GMT-4` after DST started).
Source the picker from `Intl.supportedValuesOf('timeZone')` (with a
curated fallback for older runtimes) and anchor the GMT label to the
routine's next fire time. When the next fire time is unknown (e.g.
the live preview while the form is open) and in the dropdown itself,
fall back to the IANA city, which is stable year-round.
Generated-By: looper 0.6.1 (runner=fixer, agent=claude-code)
* fix(web): always include UTC in routine timezone picker
`Intl.supportedValuesOf('timeZone')` returns only canonical region
names on current runtimes (Node 24, recent browsers) and omits `UTC`,
so the previous picker dropped the most common non-local zone unless
the runtime itself was already UTC. The backend validator and the
contract examples still accept `UTC`, so a user on a non-UTC machine
could not create a documented UTC routine from Settings.
Prepend `UTC` inside `listSupportedTimezones()` when the runtime list
omits it, so the picker stays aligned with the supported schedule
surface.
Generated-By: looper 0.6.1 (runner=fixer, agent=claude-code)
|
||
|
CaprikaandGitHub
|
8feb2e586c |
fix(connectors): preserve OAuth state and advertised tool counts (#1036)
* fix(connectors): preserve oauth state and advertised counts * test(connectors): type fixture for advertised count * docs(connectors): align tool count badge contract * docs(connectors): clarify curated tool names role |
||
|
|
66f84972cf |
feat(byok): Added Ollama CLoud as BYOK provider (#923)
* feat: add Ollama Cloud to KNOWN_PROVIDERS as OpenAI-compatible BYOK provider * feat: add ollama.com to isOpenAICompatible base URL detection * feat: add Ollama Cloud models to SUGGESTED_MODELS_BY_PROTOCOL fallback list * fix: use full Ollama Cloud model list from /api/tags, drop -cloud suffix * feat: add Ollama Cloud as native protocol with NDJSON streaming and connection test support * fix: remove ollama.com from OpenAI compatibility check * feat: add token overrides for Ollama Cloud models to prevent truncation * fix: extend inferApiProtocol and legacy migration to recognize ollama.com base URLs * fix: normalize Ollama Cloud base URL by stripping /api suffix during migration and in daemon --------- Co-authored-by: herediaron <aronheredi346@gmail.com> |
||
|
初晨andGitHub
|
9ef136ced5 |
fix: sync Orbit last run with selected prompt template (#937)
* fix(orbit): scope last run to selected template * fix(orbit): preserve legacy last run on upgrade * fix(orbit): pin legacy last-run fallback on refresh * fix(orbit): pin template id at run start * test(web): sync orbit fixtures with skill summary |
||
|
|
35900834be |
feat(contracts): move shared CritiqueRoundSummary / CritiqueRunStatus into the contracts package (#1016)
The daemon's persistence layer was the source of truth for three pieces of public critique state: the CritiqueRoundSummary record, the terminal CritiqueRunStatus union, and the CRITIQUE_RUN_STATUSES enumeration that defines the canonical display order. None of those are daemon-private: the web layer (Phase 7 reducer + Theater UI) needs the same shapes to render run history, status filters, and the rerun affordance, and the root AGENTS.md requires shared web/daemon DTOs to live in packages/contracts so the two sides can't drift. Move all three to packages/contracts/src/critique.ts and have apps/daemon/src/critique/persistence.ts re-export them so existing daemon imports (orchestrator, server, run-registry, tests) keep working unchanged. The CHECK-constraint-aware ALL_VALID_STATUSES set that adds the in-flight 'running' value stays daemon-side because it is strictly a DB invariant, not a public surface. This unblocks the web layer's eventual rerun / history wiring without forcing it to either duplicate the shapes or import daemon source (which would violate the apps/web → apps/daemon boundary). Co-authored-by: Nagendhra <nagendhra405@gmail.com> |
||
|
lefarcenandGitHub
|
afb331a288 |
feat: add opt-in Langfuse telemetry (#800)
* docs(specs): add langfuse telemetry change spec
Captures the design for forwarding completed agent runs to Langfuse,
including data-model mapping, field-budget caps, privacy gates,
build-secret injection, GDPR right-to-deletion approach, and the
resolved decisions on default consent, identifier shape, region, and
ownership.
* feat(daemon): add langfuse-trace module and telemetry prefs
Adds the dependency-free building blocks for forwarding completed
agent runs to Langfuse. Two layers:
- AppConfigPrefs gains installationId and a TelemetryPrefs object with
metrics / content / artifactManifest gates. The daemon validator
treats telemetry like agentModels — replace-on-write, drop-when-empty,
reject non-boolean inner values.
- New langfuse-trace.ts builds a {trace-create, generation-create}
pair from a ReportContext, capping prompt at 8 KB, output at 16 KB,
artifacts at 50 entries, and dropping any batch larger than 1 MB
before send. reportRunCompleted is no-op when LANGFUSE_PUBLIC_KEY /
LANGFUSE_SECRET_KEY are unset (so dev runs and forks never emit) and
short-circuits on prefs.metrics === false.
Server-side wiring into the run-close path lands in a follow-up.
* fix(langfuse): default to US Langfuse region
End-to-end smoke against the project's actual dev key on 2026-05-07
returned 401 from cloud.langfuse.com (EU) and 207 from
us.cloud.langfuse.com (US), confirming the org lives in US. Update the
default base URL, the matching test, and the spec's Q3 decision row to
match. Self-hosted or EU-region operators can still override via the
LANGFUSE_BASE_URL env var.
* feat(daemon): wire langfuse trace forwarding into run-close
Adds the daemon-side glue to forward completed agent runs:
- runs.ts gains an optional onTerminate hook fired once per run after it
reaches a terminal state. Errors thrown from the hook are caught and
logged, never propagated, so telemetry can never break the run path.
- New langfuse-bridge.ts assembles a ReportContext from the in-memory
run record, the conversation's persisted assistant message, and the
user's app-config preferences. It tolerates a missing message (e.g.
when web has not yet PUT the final delta) and a missing app-config.
- server.ts stashes the original user prompt on the run object inside
startChatRun so the bridge can include it without crossing the
createChatRunService boundary, and registers the hook callback when
building the run service.
Behavior remains a no-op unless LANGFUSE_PUBLIC_KEY / LANGFUSE_SECRET_KEY
are set in the daemon env AND telemetry.metrics is true in app-config.
A live smoke against us.cloud.langfuse.com on 2026-05-07 confirmed the
matching trace + generation schema is accepted (HTTP 207, both events
201 created).
* fix(langfuse): address PR #800 review feedback
P1 — Move trace forwarding off the daemon-internal run-close hook and
onto the message-persistence path. The original onTerminate hook ran
inside finish() the moment the SSE 'end' event was emitted, which is
*before* the web client's onDone handler refreshes project files and
PUTs producedFiles + final assistant content back to SQLite. Reading
SQLite at that moment routinely missed both. The fix: drop the runs.ts
hook entirely and trigger from PUT /api/projects/:id/conversations/:cid/
messages/:mid when the saved row carries a terminal runStatus. A
reportedRuns Set guards against the multiple PUT calls web makes per
turn (each retry / state update). Set entries auto-evict after the same
30 min TTL the runs map uses. Web persists a terminal-status message in
all three completion paths — onDone (succeeded), onError (failed), and
cancel (canceled) — so this catches every run shape.
P2 — postLangfuseBatch now parses the 207 Multi-Status response body.
Langfuse legacy ingestion always returns 207, and response.ok is true
for 207, so per-event validation errors used to slip through silently.
We now warn when body.errors is non-empty. Two new unit tests.
P2 — truncate() and the HARD_BATCH cap now compare UTF-8 byte length,
not String.length (which counts UTF-16 code units). A 4096-character
CJK prompt occupies 12 KB, well over the 8 KB input cap. truncate also
walks backwards to a UTF-8 leading byte so the cut never lands inside a
multi-byte codepoint. New unit test covers '设'.repeat(4096).
P2 — Spec R7 now lists the actual Langfuse trace deletion endpoint
(DELETE /api/public/traces/{traceId} for single, DELETE /api/public/traces
with body for batch). Verified by curl on us.cloud.langfuse.com:
DELETE /api/public/traces/X → 200; the path the original spec named
(POST /api/public/trace/X) returns 404. Reference link points at
langfuse.com/docs/administration/data-deletion.
P3 — Q4 (legacy ingestion vs OTel) moved from Open Questions to
Resolved Decisions. The implementation already commits to legacy and
the trade-off was discussed during design; the open-question status was
stale.
* feat(web): privacy consent surface + Settings → Privacy tab
Adds the user-facing half of the telemetry feature so the daemon-side
hook from PR #800 has something to talk to.
- AppConfig gains optional `installationId` (anonymous v4 uuid generated
on first opt-in; null after explicit decline; undefined when the user
has never seen the consent surface) and `telemetry: TelemetryConfig`
({metrics, content, artifactManifest}). syncConfigToDaemon round-trips
both fields so the bridge module sees the same prefs.
- SettingsDialog grows a Privacy section with two states. When the user
has never made a consent decision (typical first-run path), the
section renders the GDPR-aligned consent card: a kicker, the disclosure
body listing both metrics and conversation content as separate bullets,
and two equally-prominent buttons ("Share usage data" / "Don't share").
The Don't-share path keeps the app fully usable (core app must work
with all tracking declined). After a decision the same panel switches
to three independent toggles + the anonymous ID + a "Delete my data"
button that rotates the ID and turns everything off.
- App.tsx points the welcome modal at the new Privacy section so the
consent decision is the first thing a fresh installation sees.
- 17 i18n keys land in en + zh-CN + zh-TW with hand-translated copy,
and as English placeholders in the remaining 14 locales — enough for
the parity check to pass while leaving room for proper localisation
in a follow-up. Dict type updated.
- Minimal index.css for the consent card + toggle rows so the panel is
legible without depending on follow-up design polish.
Telemetry remains a no-op end-to-end until the user clicks Share usage
data: the daemon gate (prefs.metrics === true) keeps every code path
short-circuited otherwise.
* refactor(web): rebuild Privacy panel using project-native settings primitives
The first cut used custom .settings-privacy-* classes + raw HTML
checkboxes that didn't match any other Settings tab. Replace with the
shell other sections already use:
- settings-subsection containers with section-head + h4 + .hint
- seg-control / seg-btn pill toggles ("active" / "offline") for each of
the three telemetry preferences, mirroring NotificationsSection
- a 2-cell seg-control for the consent card so Share usage data and
Don't share carry identical visual weight (the GDPR equal-prominence
requirement that the previous accent / outline split missed)
- ghost button + readonly text input for the installation id row,
mirroring the API-key field pattern elsewhere
Drop the bespoke CSS block in favor of inheriting the existing
settings-section / seg-control / ghost styling. The only privacy-
specific style left is a tight definition list inside the consent card
for the metrics + content disclosure rows.
* refactor(web): use .toggle-row iOS switch for Privacy preferences
Active/offline pills (the seg-control single-cell pattern that
NotificationsSection uses) read awkwardly for a flat preference list.
Switch the three telemetry toggles to .toggle-row — the same control
NewProjectPanel uses for "speaker notes" / "animations": label + hint
on the left, iOS-style sliding switch on the right, full-row click
target. The consent card's two-button seg-control stays as-is — there
the equal-weight pill pair is exactly what GDPR equal-prominence wants.
* feat(web): standalone first-run privacy consent banner
Replaces the Settings-dialog-as-onboarding hack with a dedicated
bottom-right banner card that mounts whenever the user has never made
a privacy decision (cfg.installationId === undefined). The banner is
prominent (anchored to the corner with a soft shadow) but
non-blocking, mirrors cookie-consent UX, and shares the project's
panel styling — same .modal-elevated background, --radius-lg corners,
--shadow-lg lift.
Wiring:
- App.tsx imports PrivacyConsentModal and renders it at the root,
gated on installationId === undefined && !settingsOpen so it doesn't
double up with the Privacy tab's own consent card when Settings is
already showing.
- Share / Don't share both go through handleConfigPersist, so the
resulting installationId + telemetry prefs land in localStorage and
the daemon at the same time, reusing the existing autosave plumbing.
- The previous attempt that pinned the welcome SettingsDialog to the
Privacy section is reverted; onboarding now stays focused on agent
configuration, and the consent decision lives in its own surface.
* fix(web): keep privacy banner visible while Settings welcome modal is open
The banner gated itself on `!settingsOpen` to avoid double-rendering
with the Privacy tab's consent card. But the first-run path opens the
Settings welcome modal automatically when `onboardingCompleted=false`,
which fired immediately after bootstrap — so the banner flashed for a
moment and then vanished behind the modal backdrop.
Drop the `!settingsOpen` clause so the banner stays mounted whenever
the user has not yet made a privacy decision, and bump its z-index
above the modal backdrop (200 vs 100) so first-run users can actually
reach the consent buttons. The minor visual overlap with the Privacy
tab's own card is fine: clicking either copy resolves both surfaces.
* copy(privacy): soften consent button labels
Banner action buttons now read "Help improve Open Design" / "Not now"
(en, with hand translations in zh-CN / zh-TW and English placeholders
in the other 13 locales) instead of "Share usage data" / "Don't share".
The new wording aligns the affirmative action with the kicker copy
("Help us improve Open Design") and reads less alarming, while the
disclosure list above still names both data categories explicitly so
the consent stays informed under GDPR. The decline button stays as a
soft "Not now" rather than an aggressive "Don't share" so the reject
path doesn't read as hostile to the user.
No structural change — the two-cell seg-control still gives the buttons
identical visual weight, and the underlying side-effects are unchanged
(installationId is generated on Help / nulled on Not now, and the
telemetry prefs flip the same way).
* feat(telemetry): expand trace fields for evals & dataset construction
Each Langfuse trace now ships the full per-turn + per-install fact
sheet that the eval/dataset workflow needs, instead of only the bare
turn id + token count from before. Everything below is gated by
`prefs.metrics === true`; nothing here is content (those gates remain
separate).
Per-turn:
- model — first-class generation.model field, drives Langfuse cost
lookup and model-grouping in the UI; also mirrored in trace.metadata
and trace.tags so list-view filters work.
- reasoning — generation.modelParameters.{ reasoning } so the Model
Parameters card lights up; mirrored in metadata.
- skillId / designSystemId — metadata + tags, so dataset slices can
group by which skill/DS produced which output.
Per-process / build (constant within one daemon run, cached at start):
- appVersion / appChannel / packaged from app-version.ts
- nodeVersion (process.version), os (platform()), osRelease,
arch (os.arch())
- clientType — desktop vs web, derived from a new X-OD-Client header
the web layer sets in providers/daemon.ts (with a User-Agent sniff
fallback for third-party callers).
Plumbing:
- startChatRun stashes model / reasoning / skillId / designSystemId
on the run object alongside the existing userPrompt stash.
- POST /api/runs reads X-OD-Client and stores run.clientType.
- langfuse-bridge collects RuntimeInfo once per process and merges
per-run client carrier; ReportContext gains optional `turn` +
`runtime` blocks; existing fields stay backward compatible.
Spec gains a "Telemetry Fields Catalog" section enumerating every
field, its source, and the gate it lives under, so the eval team has a
single place to look up what's available without reading the trace
schema by example.
Tests:
- new langfuse-trace tests cover turn tags, runtime tags, generation
model/modelParameters promotion, modelParameters omission when
reasoning is unset, and metadata mirroring.
- langfuse-bridge gains an end-to-end "turn-level config" test that
threads model/reasoning/skill/DS/clientType + appVersion through
the bridge and asserts the Langfuse payload shape.
- existing tests adjusted to tolerate host-dependent os tag.
* copy(privacy): trim Share button to verb phrase only
"Help improve Open Design" overflowed the equal-width 2-cell
seg-control on the consent banner — the product name is already in
the kicker + headline above the buttons, so the button itself only
needs the verb phrase. Drop the product name from all locales:
- en: Help improve Open Design → Help improve
- zh-CN: 帮助改进 Open Design → 帮助改进
- zh-TW: 協助改進 Open Design → 協助改進
The decline button ("Not now" / "暂不" / "暫不") was already short, so
the two buttons now have comparable length and the equal-prominence
seg-control fits cleanly. Standalone Settings → Privacy panel uses
the same labels for consistency.
* fix(web): defer Settings welcome modal until privacy decision is made
Previously bootstrap raced two surfaces against each other on first
launch: the privacy consent banner (gated on installationId ===
undefined) and the Settings welcome modal (gated on
onboardingCompleted === false). The banner's higher z-index kept it
above the backdrop visually, but having two foreground surfaces at
once is still confusing UX.
Sequence them instead: bootstrap only opens the welcome modal when
the user has *already* resolved consent (installationId !== undefined).
Until then the banner owns the foreground alone. Once the user clicks
Help improve / Not now, the corresponding handler hands off to the
welcome modal if onboarding is still pending. End state matches what
it was before — just without the simultaneous-render flash.
* debug(privacy): log banner gate state to track sudden disappearance
Two console.log points to find which setCfg call (or stale bundle) is
flipping cfg.installationId from undefined to a value while the banner
is visible. To remove once the regression is reproduced.
* fix(privacy): keep installationId + telemetry out of localStorage
Daemon is now the single source of truth for the privacy decision.
Why this matters: the consent banner gates on
\`config.installationId === undefined\`, but loadConfig() merges
localStorage on top of the daemon's reply, so a stale uuid in
\`open-design:config\` (left over from a previous opt-in) was
re-hydrating the React state and immediately syncing back to the
daemon — defeating "Delete my data" and re-suppressing the banner
within milliseconds of every page load.
The deeper reason to fix it here, not just patch the gate: a privacy
identifier persisted in browser storage that the user can't see or
clear without DevTools is a compliance liability. Anything users can
revoke needs one canonical place to store it. Daemon \`app-config.json\`
already serves that role for everything else gated through
syncConfigToDaemon, so installationId + telemetry now ride that path
exclusively:
- saveConfig() strips both keys before writing localStorage.
- loadConfig() strips both keys when reading older stale payloads,
so existing installs migrate transparently on next launch.
- syncConfigToDaemon() / mergeDaemonConfig still round-trip them, so
the React state stays in sync with the daemon as before.
Net effect: clearing app-config.json (or hitting "Delete my data") now
fully resets the install identity, with no residual cohort key in
browser storage.
* feat(privacy): scrub secrets + PII from prompt/output before send
When prefs.content is on, daemon now runs the prompt and assistant
text through a regex scrubber (apps/daemon/src/redact.ts) before
posting to Langfuse. The scrubber is the simplest thing that gives
the user-facing copy a truthful claim — pure regex, zero new
dependencies, fully auditable in this Apache-2.0 repo (vs. pulling a
single-maintainer 5-month-old npm package into a core process).
Categories covered (each replaced with [REDACTED:<kind>]):
- Anthropic / OpenAI sk- keys (incl. proj/live/test/ant variants)
- Langfuse pk-lf- / sk-lf- (specific rule wins over generic sk-)
- GitHub gh[opsur]_ tokens
- AWS access key ids (AKIA + 16 uppercase)
- Google API keys (AIza + 35)
- Slack xox[abprs]- tokens
- Stripe live/test keys
- JWT header.payload.signature triples
- Bearer-header values (scheme word stays readable)
- Emails, IPv4, US-style phone numbers
- Credit cards — 13–19 digit runs that pass a Luhn check, so order ids
and unix-nanos timestamps that fail Luhn pass through unchanged
Not covered, stated openly in spec + i18n: names, postal addresses,
business-secret semantics, raw 40-hex tokens (too high a false-positive
cost for artifact slugs). Those would require an ML layer.
Wired in:
- apps/daemon/src/redact.ts — exports redactSecrets() +
redactSecretsWithCounts() helper for future audit-summary metadata.
- apps/daemon/src/langfuse-bridge.ts — runs both prompt and output
through redactSecrets() before they reach the trace builder.
- 18 unit tests cover every pattern plus negative cases (Luhn-failing
digit runs, out-of-range IPv4 octets, idempotence on re-redacted
text, ordinary prose passthrough).
- i18n privacyContentHint on en + zh-CN + zh-TW (plus 14 locale
placeholders) enumerates the categories so the consent disclosure
matches the implementation — the GDPR informed-consent requirement.
- spec gains a Pre-send Redaction subsection with the regex shape
table + intentional non-coverage list.
Drive-by: dropped the [privacy] debug logs that traced the now-fixed
bootstrap regression.
* fix(telemetry): make Langfuse reporting resilient
* feat(telemetry): nest Langfuse turn observations
* feat(telemetry): emit Langfuse tool spans
* fix(telemetry): report after finalized message writes
* fix(telemetry): honor persisted terminal status
* fix(web): let consent banner yield page clicks
* fix(telemetry): report current turn prompt only
|
||
|
|
05461c64fd |
fix(connectors): show stable curated tool count in connector card badge (#748) (#767)
* fix(connectors): show stable curated tool count in connector card badge (#748)
The connector card's "N tools" badge in `apps/web/src/components/
EntryView.tsx` rendered `connector.tools.length` for both pre- and
post-Composio-hydration states, so the displayed count lurched
without explanation:
- Before configuring a Composio API key, GitHub showed "2 tools"
(the static fallback catalog at `composio.ts:21-53`).
- After hydration, the daemon merged in the full Composio provider
inventory (`composio.ts:725-778`, ~868 GitHub tools), and the
same badge jumped to "868 tools".
Same field name, two different concepts — `connector.tools` is
"everything the provider ships" while the agent-callable subset
is `definition.allowedToolNames` (read-only auto-approval tools
that pass `isAgentPreviewListableTool()`).
Fix: surface `allowedToolNames` on the wire `ConnectorDetail` and
have the badge use that count instead. The detail drawer below
the card still iterates over `connector.tools` to enumerate the
full inventory — the count and the list are intentionally
different surfaces.
The badge now stays close to "tools the agent can actually
invoke" (≈2-30 for GitHub, depending on auto-allowed read tools)
instead of the raw provider inventory (~868). No 800x jump on
hydration.
Wire format change:
- packages/contracts/src/api/connectors.ts: add
`allowedToolNames: string[]` to ConnectorDetail
- apps/daemon/src/connectors/catalog.ts: same field on the
daemon-internal type, populated in `connectorDefinitionToDetail()`
as a defensive copy of `definition.allowedToolNames`
- packages/contracts/src/examples.ts: extend the example fixture
- apps/web/src/components/EntryView.tsx: badge call sites switch
to `connector.allowedToolNames?.length ?? connector.tools.length`
(the `??` keeps the badge alive against any older daemon build
that hasn't shipped the field yet)
Tests:
- 4 new daemon tests in connectors-service.test.ts pin the
contract: getConnector() emits the field, the array is a
defensive copy, and the #748 regression guard simulates the
Composio post-hydration shape (tools.length=801,
allowedToolNames.length=1) to prove the badge invariant
- web EntryView.test.ts fixtures updated to satisfy the new
required field
Verified locally:
- daemon vitest: 925/925
- web vitest: 332/332
- daemon/web/contracts typecheck clean
- i18n-check passes
- Live `/api/connectors/discovery` returns the new field; pre-
hydration GitHub/Notion/Google Drive badges all read "2 tools"
(no regression vs before this change)
Fixes #748
* fix(connectors): split drawer badge vs inventory counts; fix daemon test typecheck (#767 review)
Two follow-ups for @mrcfps's review on PR #767.
1) **P1: Drawer empty-state regression.** The previous commit reused
the curated `toolCount` for both the header badge AND the inventory
section's loading gate / empty-state branch / section count. The
inventory section renders `connector.tools` directly, so a hydrated
connector with raw provider tools but an empty allowlist (e.g. a
write-only Composio surface) would render "no tools available" and
hide the actual inventory list — exactly the contradiction my own
PR description warned against.
Fix splits the two surfaces:
- `badgeToolCount` (curated, via the new exported helper
`getConnectorBadgeToolCount`) feeds the card and drawer header
badge — the summary count, where the #748 stability matters.
- `inventoryToolCount = connector.tools.length` (inline) drives the
drawer's loading gate, section count, and empty-state branch —
the surfaces describing the actual rendered list.
The card has no inventory section so it stays on the badge helper
unchanged.
2) **CI: daemon test typecheck failed.** The connectors-service test's
`provisionedTools[0].safety` index access tripped daemon
`tsconfig.tests.json`'s `noUncheckedIndexedAccess` strict setting,
even though my local `tsc -p tsconfig.json --noEmit` was clean —
that config is a separate compilation. Bind through a defined-checked
local before reading `.safety`, per @mrcfps's exact suggestion.
Tests:
- 4 new web tests in `EntryView.test.ts` pin the
`getConnectorBadgeToolCount` contract, including the explicit
regression: a connector with `allowedToolNames=[]` and
`tools.length=800` returns badge=0 but the inventory length
stays at 800 — the drawer's empty-state branch must use the
inventory count, never the badge count.
- Existing daemon test fixed without losing assertion coverage.
Verified locally:
- daemon vitest: 921/921
- web vitest: 336/336 (was 332, +4)
- daemon `tsc -p tsconfig.json` and `tsc -p tsconfig.tests.json` (the CI killer): both clean
- web `tsc -b --noEmit` clean
- i18n-check passes
Process learning baked into this PR: from now on I'll always run the
`tsconfig.tests.json` separately before pushing, since the workspace
typecheck script chains both and the second one is what CI fails on.
* fix(connectors): pin badge to curated catalog count, not the dynamic execution allowlist (#767 review v2)
@lefarcen and @mrcfps both flagged that the previous iteration of
this PR (commit
|
||
|
netteeandGitHub
|
32d820e4ee |
fix(daemon): typecheck leaf modules (#943)
* update drift * fix(daemon): typecheck leaf modules * fix(daemon): decode Qoder stdout buffers Generated-By: looper 0.5.6 (runner=fixer, agent=opencode) |
||
|
CaprikaandGitHub
|
4f647f56ba |
[codex] Optimize Composio connector previews (#907)
* Optimize Composio connector previews * Fix partial connector tool preview hydration * Cancel pending connector authorization on daemon * Preserve Composio cached tool counts * Avoid pending state after OAuth launch failure * Preserve static tool count fallback * Fix connector preview retry state * Remove Composio auth config metrics * Hydrate unknown connector tool previews * Fix remaining connector review threads * Stop failed connector preview spinner * Hydrate only targeted agent connectors |
||
|
|
e13adf2e63 |
feat(daemon): finalize design package endpoint (closes #450) (#832)
* feat(daemon): scaffold /api/projects/:id/finalize/anthropic (refs #450) Phase C of the PR 2 plan for issue #450: scaffold the route + module shape so subsequent phases (D-I) land function bodies and tests against a stable surface that already passes typecheck. What lands here: - apps/daemon/src/finalize-design.ts: module-level constants (DEFAULT_BASE_URL, DEFAULT_MAX_TOKENS=16000, INPUT_BODY_CAP_BYTES=384KiB, LOCK_FILENAME=.finalize.lock, OUTPUT_FILENAME=DESIGN.md, DEFAULT_TIMEOUT_MS=120s); inline interfaces for the request/response shape (kept out of packages/contracts per scope rules); two error classes - FinalizePackageLockedError (mirrors PR #493's TranscriptExportLockedError) and FinalizeUpstreamError (carries upstream HTTP status for the route's error mapping); function stub that throws "not yet implemented". - apps/daemon/tests/finalize-design.test.ts: vitest harness with describe.skip placeholder so the file imports cleanly. Real cases land in phases D-I. Default-import of node:fs (per memory: vi.spyOn cannot redefine on the frozen ESM Module Namespace; CJS exports object is mutable). - apps/daemon/src/server.ts: route handler at POST /api/projects/:id/finalize/anthropic, slotted next to the existing :id/deploy* family. Validates apiKey/model non-empty, optional baseUrl via the existing validateExternalApiBaseUrl closure (forbidden -> 403, invalid -> 400), optional maxTokens positive number; calls getProject (404 on miss); calls finalizeDesignPackage (which throws, caught and mapped to 500 for now); maps known error classes (FinalizePackageLockedError -> 409, FinalizeUpstreamError -> 502) pre-emptively. Path shape rationale (Bryan-confirmed): project-scoped path matches every sibling /api/projects/:id/* route in server.ts (deploy, deployments, deploy/preflight); provider-namespaced segment leaves a clean expansion line for /api/projects/:id/finalize/openai etc. as follow-ups. Field-name rationale: apiKey, baseUrl, model, maxTokens match ProxyStreamRequest verbatim (packages/contracts/src/api/proxy.ts:8-19) so a future caller can reuse the same body shape. baseUrl is optional here (intentional divergence from the proxy at server.ts which requires it) so standard Anthropic users do not need to set it; Bedrock / self-hosted-proxy users still can. Verification: pnpm --filter @open-design/daemon typecheck exits 0; finalize-design.test.ts loads cleanly with 1 skipped placeholder; no other tests touched. Refs nexu-io/open-design#450 (PR 2 scaffold; pipeline body in subsequent commits) * feat(daemon): transcript truncation helper for /finalize prompt Phase D of the PR 2 plan for issue #450: lands the helper that bounds the transcript section of the synthesis prompt. Why this exists: real-world signal at authoring time was a local project transcript already at 3.95 MB. Anthropic's claude-opus-4-7 context cap is roughly 200K tokens (~700 KB at typical density). Inserting an unbounded transcript would 4xx upstream on the first real call. This helper keeps the on-disk .transcript.jsonl lossless (PR #493's contract) while making the prompt-inclusion bounded. Strategy: - Cap output at INPUT_BODY_CAP_BYTES (384 KiB) so the prompt has room for the system prompt + design system body + current artifact + room for the synthesis output. - Always preserve the header line - it carries projectId, schemaVersion, conversation/message counts, attachment counts; synthesis quality depends on knowing the original sizes. - Split equal byte budgets between head and tail so both project genesis and most-recent intent survive. Two thinking segments separated only by mid-session truncation lose the same kind of boundary that PR #493 preserves between thinking blocks - that's accepted; smarter semantic chunking is a follow-up. - Insert a single `{"kind":"truncated","reason":"size","omittedBytes":N}` sentinel JSON line between the head and tail so a synthesis consumer can detect the gap. omittedBytes is the difference between the original UTF-8 byte length and the output's UTF-8 byte length. - If the head + tail budgets together cover the whole body (e.g. all message lines are tiny), no marker is emitted - the output is the input verbatim. Tests: - "returns the input verbatim when the JSONL fits under the 384 KiB cap" pins that small transcripts pass through unchanged with no marker. - "head+tail truncates with a single marker line when the JSONL exceeds the 384 KiB cap" pins that output is bounded, header survives, exactly one marker emitted with non-zero omittedBytes, both ends of the body preserved, and at least one middle message omitted. Suite delta: +2 tests in finalize-design.test.ts. Refs nexu-io/open-design#450 * fix(daemon): resolve noUncheckedIndexedAccess in truncateTranscriptForPrompt D1 (0eaa123) shipped with `body[headIndex]` and `body[i]` typed as `string | undefined` under TypeScript's `noUncheckedIndexedAccess` strict mode. Local typecheck would have caught it but the prior verification piped through `tail` which masked the non-zero exit code of `tsc`. Coalesce each access via `?? ''` (the array is from `String.split('\n')` so undefined elements are not actually reachable; the coalesce is a type-narrowing convenience, not a behavior change). Verification: `pnpm --filter @open-design/daemon typecheck` exits 0; `pnpm --filter @open-design/daemon test finalize-design` shows 2/2 + 1 skipped, identical to the pre-fix run. Refs nexu-io/open-design#450 * feat(daemon): current-artifact resolver for /finalize Phase E of the PR 2 plan for issue #450: resolves which artifact (if any) accompanies the transcript + design system in the synthesis prompt. Priority order (Bryan-locked in plan §6): 1. The file referenced by tabs.is_active = 1 IF an <name>.artifact.json sidecar exists on disk. Sidecar presence is the discriminator: an inferred manifest from `inferLegacyManifest` (e.g. for a bare .html with no sidecar) does NOT count, and an active tab pointing at a non-artifact file (.md, .txt) falls through. 2. Newest project file with a real .artifact.json sidecar, sorted by manifest.updatedAt descending. Files without an updatedAt sort last so legacy pre-streaming manifests do not get accidentally promoted. 3. Returns null - "no artifact in scope". The Phase H caller will emit `artifact: null` in the response and the prompt's "Current artifact" section will read "none". Sidecar presence is checked via `existsSync` on the on-disk path, NOT via the `artifactManifest` field returned by readProjectFile/listFiles (those run inferLegacyManifest as a fallback for known kinds, which would otherwise cause a bare .html with no sidecar to look like an artifact). Tests: - "returns the active-tab artifact when its sidecar is present, even if a newer artifact exists elsewhere": pinned.html (older updatedAt) is in the active tab; newer.html (newer updatedAt) is not. Resolver returns pinned.html - intent (active tab) beats recency. - "falls through to newest .artifact.json when active tab points at a non-artifact file": README.md is the active tab (no sidecar); design.html has a real sidecar. Resolver falls through and returns design.html. - "returns null when no active tab and no .artifact.json sidecars exist": only a README.md is in the project; no tabs row. Resolver returns null. Suite delta: +3 tests in finalize-design.test.ts (5 active total). Refs nexu-io/open-design#450 * feat(daemon): synthesis prompt construction for /finalize Phase F of the PR 2 plan for issue #450: builds the system + user prompts that get sent to Anthropic's Messages API in the synthesis call. Pure function; no IO, no side effects. System prompt (literal, stored as a module-level constant): instructs Claude to emit a DESIGN.md document with a fixed 7-heading structure (# DESIGN.md / ## Summary / ## Brand & Voice / ## Information Architecture / ## Components & Patterns / ## Visual System / ## Open Questions / ## Provenance). The Provenance section is required to list project ID, design system, current artifact, transcript message count, and the UTC generation timestamp. User prompt (built at runtime): structured payload with the truncated transcript JSONL, the design system body, and the current artifact body, each under a ## heading. Missing inputs (no design system selected, no artifact in scope) produce explicit "none" headings + parenthetical placeholder body so Claude does not hallucinate content for absent sections. Truncation is the caller's concern - this function does not re-truncate. The caller (Phase H pipeline) feeds in a JSONL that has already been bounded by truncateTranscriptForPrompt. Tests: - "includes the transcript JSONL verbatim and the generation context": pins all section headings, the transcript body verbatim, the design system body verbatim, the artifact body verbatim, and every generation-context line. - "falls back to \"none\" + parenthetical when no design system is selected": designSystemId=null and designSystemBody=null -> heading reads "## Active design system: none" with the parenthetical body. - "falls back to \"none\" + parenthetical when no artifact is in scope": artifact=null -> heading reads "## Current artifact: none" with the parenthetical body. Suite delta: +3 tests in finalize-design.test.ts (8 active total). Refs nexu-io/open-design#450 * feat(daemon): Anthropic call + retry strategy for /finalize Phase G of the PR 2 plan for issue #450: lands the upstream Claude Messages API call with a single transient-error retry, plus the response extractor that turns Anthropic's content array into the DESIGN.md body. What lands here: - appendVersionedApiPath: inlined from the connectionTest helper at apps/daemon/src/connectionTest.ts:188-195 (it is not exported there). Appends /v1/messages when the base URL has no /vN segment, otherwise appends /messages directly. Same semantics; ~5 lines. - callAnthropicWithRetry: POSTs to <base>/v1/messages with the canonical Anthropic headers (content-type, x-api-key, anthropic-version: 2023-06-01) and body shape ({ model, max_tokens, system, messages, stream:false }). One retry on transient (HTTP 429 or 5xx); on terminal failure throws FinalizeUpstreamError carrying the upstream HTTP status and raw body text. The route handler in Phase I maps status to AUTH_FAILED / RATE_LIMITED / UPSTREAM_FAILED and runs the body through redactSecrets before exposing it as `details`. - extractDesignMd: concatenates content[].text for every block where type === 'text', preserving order. Throws FinalizeUpstreamError(502) on three malformed-response shapes: non-object payload, missing content array, zero text blocks. The route handler maps the throw to 502 UPSTREAM_FAILED so synthesis cannot land a half-empty DESIGN.md on disk. - Test-only `_sleepMs` injection on the call params so the retry-delay sleep is instant under vitest. Default sleep uses setTimeout. Retry posture (1 retry on transient) is opinionated; the maintainer's "standard exponential backoff" answer was directional and a single retry matches the existing daemon's posture (transcript export and connectionTest do zero retries) while staying inside the daemon's blocking-fast posture for /finalize. Tests: - callAnthropicWithRetry: throws on 401 with no retry; retries once on 429 and resolves on second 200; throws after both 5xx attempts; propagates AbortError when signal is pre-aborted. - extractDesignMd: concatenates ordered text blocks; throws on missing content array; throws on content with zero text blocks. A spurious typecheck error from `exactOptionalPropertyTypes` (signal typed as AbortSignal | undefined where RequestInit expects AbortSignal | null) was resolved by conditionally spreading signal into the RequestInit literal. Suite delta: +7 tests in finalize-design.test.ts (15 active total). Refs nexu-io/open-design#450 * feat(daemon): wire /finalize pipeline end-to-end Phase H of the PR 2 plan for issue #450: stitches together every phase D-G primitive into the full finalizeDesignPackage pipeline that the route handler in Phase I will expose over HTTP. Pipeline (in execution order, all inside a try/finally that always releases the lockfile): 1. getProject(db, projectId): defensive 404 (the route validates first; this throw catches direct CLI/script callers). 2. mkdirSync(<projectDir>, { recursive: true }): some projects have DB rows but no on-disk dir yet (PR #493's same fix). 3. fs.openSync(.finalize.lock, 'wx'): EEXIST -> FinalizePackageLockedError (mirror PR #493's TranscriptExportLockedError). 4. exportProjectTranscript(db, projectsRoot, projectId, { now }): produces .transcript.jsonl on disk; we read the body and run it through truncateTranscriptForPrompt to bound the prompt-inclusion size. 5. readDesignSystem(designSystemsRoot, designSystemId): returns null when the project has no design_system_id selected, when the design system directory does not exist, or when the DESIGN.md file is missing. 6. resolveCurrentArtifact(db, projectsRoot, projectId): active tab -> newest .artifact.json by manifest.updatedAt -> null. 7. buildSynthesisPrompt({...}): system + user prompt (per Phase F). 8. callAnthropicWithRetry({...}): one retry on 429/5xx; throws FinalizeUpstreamError on terminal failure. 9. extractDesignMd(payload): concatenates content[].text blocks; throws FinalizeUpstreamError(502) on malformed shape. 10. Atomic write: writeFileSync({flag:'wx'}) -> reopen for fsync -> rename. Errors unlink tmp before rethrowing. 11. Lock release in finally (always closeSync + unlinkSync). Bounded blocking: the function uses its own AbortController + 120s timeout when the caller does not supply a signal. Caller-supplied signal takes precedence. Type tightening: switched the local Db interface to `type Db = Database.Database` (better-sqlite3) so the function signature is compatible with `exportProjectTranscript`'s typed parameter. Source file already had a `better-sqlite3` import in claude-design-import area of the daemon, so no new dependency. Tests: - "writes DESIGN.md atomically on the happy path": end-to-end with seeded project + conversation + 2 messages + design system on disk; asserts file at exact path + body bytes match the fetch mock. - "response carries every documented field with correct types": designMdPath/bytesWritten/model/inputTokens/outputTokens/artifact/ transcriptMessageCount/designSystemId all present and typed. - "emits design system 'none' in the prompt when no design_system_id is set": fetch mock asserts on the body it receives. - "throws FinalizePackageLockedError when .finalize.lock is already held": pre-create lockfile; assert throw + DESIGN.md not written + pre-existing lock NOT unlinked (we did not own it). - "replaces an existing DESIGN.md atomically on a second finalize": inject a sentinel between two finalize calls; assert sentinel is gone after second run. - "cleans up tmp file AND lock file on every error path": mock fs.writeFileSync to throw on the tmp path; assert no DESIGN.md.tmp.* remain, no DESIGN.md, no .finalize.lock. - "uses the default https://api.anthropic.com baseUrl when baseUrl is omitted": fetch URL begins with the default; baseUrl=undefined path. vi.restoreAllMocks() now runs in afterEach so the writeFileSync spy from the cleanup test does not leak into subsequent tests. Suite delta: +7 tests in finalize-design.test.ts (22 active total). Refs nexu-io/open-design#450 * feat(daemon): /finalize HTTP route handler + error mapping Phase I of the PR 2 plan for issue #450: replaces the Phase C stub's catch-all 500 with status-aware error mapping that surfaces the right HTTP status + error code for each documented failure mode, and adds HTTP-layer tests that boot startServer to exercise the route's validation branches. Route handler changes: - :id format guard: an inline regex matching isSafeId at apps/daemon/src/projects.ts:556-558 rejects unsafe ids with 400 BAD_REQUEST before any DB or filesystem work. Without this, an id like 'bad!id' would either fail getProject as 404 (wrong code) or reach the function and throw 'invalid project id' (mapped to 500). - FinalizeUpstreamError mapping is now status-aware: - upstream 401 -> 401 AUTH_FAILED - upstream 429 -> 429 RATE_LIMITED - upstream 5xx (or our own 502 sentinel for malformed responses) -> 502 UPSTREAM_FAILED In all cases the upstream raw text is run through redactSecrets so the apiKey cannot leak through `details` even if the upstream echoes the inbound headers. - AbortError mapping: when the 120s AbortController fires (or the caller pre-aborted the signal), surface as 503 TIMEOUT. - Default case: console.error the error per daemon convention; client sees 500 INTERNAL with the message routed through redactSecrets. - Imported redactSecrets alongside the existing connectionTest imports (apps/daemon/src/server.ts:51). HTTP-layer tests (boot startServer({port:0,returnServer:true}) once in beforeAll, mirror the proxy-routes.test.ts pattern): - "400 BAD_REQUEST when baseUrl is not a valid URL (test #13)": baseUrl='not-a-url'. - "403 FORBIDDEN when baseUrl points at a private internal IP (test #14)": baseUrl='http://10.0.0.1'. Note: validateBaseUrl explicitly allows loopback (for local OpenAI-compatible servers) and only blocks non-loopback private IPs (10/8, 172.16/12, 192.168/16, fc00::/7, fe80::/10). - "400 BAD_REQUEST when apiKey is missing (test #15)": apiKey omitted. - "400 BAD_REQUEST when :id contains characters outside the safe-id regex (test #16)": id='bad!id' contains '!' which is not in [A-Za-z0-9._-]. Suite delta: +4 tests (26 active in finalize-design.test.ts). Full daemon suite: 1078/1078 pass; baseline+26 (the +5 above plan target reflects retry+extract split into more granular unit tests than originally enumerated; all real, none skipped). Refs nexu-io/open-design#450 * fix(daemon): tighten isSafeId to reject pure-dot project ids Addresses the P1 path-traversal finding from @lefarcen on PR #832 (https://github.com/nexu-io/open-design/pull/832#discussion_r3202512644). The pre-fix `isSafeId` at apps/daemon/src/projects.ts:556-558 used regex `/^[A-Za-z0-9._-]{1,128}$/` which permitted pure-dot ids (`.`, `..`, `...`) because `.` is in the character class. `projectDir` and `resolveProjectDir` both delegated to `isSafeId`, so an id of `..` would resolve to the PARENT of `.od/projects/` via `path.join`. Threat model (per @lefarcen): - An attacker creates a project row whose stored id is `..` (or another pure-dot variant) — for instance via a workflow that writes the row directly without going through the API. Subsequent finalize/write ops keyed by that id then escape the project tree. - A direct CLI / scripted caller passing `..` as the project id reaches the function without HTTP normalization saving us. (Express normalizes %2e%2e to .. and collapses path segments, which yields 404 for the URL `/api/projects/%2e%2e/...` in practice — but that's Express's protection, not ours.) Fix: - isSafeId now explicitly rejects pure-dot ids (`/^\.+$/.test(id)`) before the char-class regex check. Empty string and inputs longer than 128 chars are also rejected explicitly so the function fails closed on edge cases. - isSafeId is now exported from apps/daemon/src/projects.ts so the /finalize route handler in apps/daemon/src/server.ts can use the same validator instead of re-implementing the regex inline. This prevents drift between the route guard and the projectDir guard, which was how this hole originally appeared. Tests (in finalize-design.test.ts because that's where the threat was flagged; isSafeId is daemon-wide so a dedicated test file would also work): - isSafeId rejects `.`, `..`, `...`, `....` - isSafeId rejects ids with `/`, `\`, `!`, leading whitespace - isSafeId rejects empty string and >128 chars - isSafeId rejects non-string inputs (null/undefined/number) - isSafeId accepts plain ids, ids with mid-string dots, UUIDs, single chars Suite delta: +7 tests (33 active in finalize-design.test.ts). Full daemon suite: 1085/1085. Refs nexu-io/open-design#832 * fix(daemon): address PR #832 P1 findings — imported folders + network 502 Addresses two of the three P1 findings from @lefarcen on PR #832: 1. Imported-folder projects route DESIGN.md to metadata.baseDir (https://github.com/nexu-io/open-design/pull/832#discussion_r3202512656, also flagged independently by @chatgpt-codex-connector at #discussion_r3202430470) The pipeline previously called `projectDir(projectsRoot, projectId)` unconditionally, which resolves to `.od/projects/<id>`. For projects created via /api/import/folder the project row's `metadata.baseDir` carries the user's actual folder; without threading metadata through, finalize would silently land DESIGN.md in the hidden daemon data dir and the current-artifact resolver would miss the user's real files. Fix: switch from `projectDir` to `resolveProjectDir(projectsRoot, projectId, metadata)` in both `finalizeDesignPackage` and `resolveCurrentArtifact`. Thread `project.metadata` (from `getProject`'s normalized row) through both call paths. The resolver gets a new optional `metadata` parameter; native projects pass null and get identical behavior. 2. Network failures and JSON parse errors now map to 502 UPSTREAM_FAILED (https://github.com/nexu-io/open-design/pull/832#discussion_r3202512661) Pre-fix, only HTTP-non-OK responses were wrapped as FinalizeUpstreamError. DNS failures (ECONNREFUSED, ENOTFOUND), fetch TypeErrors, and `response.json()` SyntaxErrors fell through to the route's catch-all and surfaced as 500 INTERNAL — incorrect: those are upstream-level failures, not daemon bugs. Fix: - Wrap callAnthropicWithRetry in a try/catch that passes FinalizeUpstreamError and AbortError through verbatim, but rewraps any other thrown error as FinalizeUpstreamError(502, '', message). - Wrap response.json() in a try/catch that rewraps SyntaxError as FinalizeUpstreamError(502, '', "upstream Anthropic returned non-JSON body: ..."). - The route handler's existing FinalizeUpstreamError mapping then correctly maps these to 502 with the message in `details` (run through redactSecrets first). Tests: - "writes DESIGN.md under metadata.baseDir for imported-folder projects": inserts a project row with metadata.baseDir pointing at a user-folder temp dir; asserts result.designMdPath lands there AND the hidden .od/projects/<id> dir does NOT contain a DESIGN.md. - "rewraps fetch network rejection as FinalizeUpstreamError(502)": fetchImpl throws TypeError with cause.code='ENOTFOUND'; assert thrown error has name=FinalizeUpstreamError and status=502. - "rewraps 200 with non-JSON body as FinalizeUpstreamError(502)": fetchImpl returns 200 with text/html body; response.json() throws SyntaxError internally; assert FinalizeUpstreamError(502). Suite delta: +3 tests (36 active in finalize-design.test.ts). Full daemon suite: green at last check; will re-verify before push. Refs nexu-io/open-design#832 * refactor(daemon): move /finalize DTOs to contracts + map error codes + validate active-tab Addresses the P2 and P3 findings from @lefarcen on PR #832: P2 — Error codes + DTOs not in packages/contracts https://github.com/nexu-io/open-design/pull/832#discussion_r3202512673 Reverses my plan's locked decision #10 ("no contracts changes in this PR; inline the request/response types"). That rule came from the predecessor PROMPT brief's anti-pattern table; @lefarcen's review is fresher signal and supersedes it. Drift risk between the daemon's inline types and any future PR 3 web client is real. - New contracts module: packages/contracts/src/api/finalize.ts with FinalizeAnthropicRequest / FinalizeArtifactRef / FinalizeAnthropicResponse. Re-exported from the package root and made addressable via `@open-design/contracts/api/finalize` subpath. - Daemon source imports the canonical types from contracts and re-exports the public type names so internal references keep working without touching every call site. - Daemon-local error codes remapped to existing ApiErrorCode union members (apps/daemon/src/server.ts), per @lefarcen's suggested mapping: FINALIZE_IN_PROGRESS -> CONFLICT AUTH_FAILED -> UNAUTHORIZED UPSTREAM_FAILED -> UPSTREAM_UNAVAILABLE TIMEOUT -> UPSTREAM_UNAVAILABLE (status 503) INTERNAL -> INTERNAL_ERROR HTTP status codes are unchanged; only the `code` field in the error JSON body changed. P3 — Active-tab name not validated before sidecar probe https://github.com/nexu-io/open-design/pull/832#discussion_r3202512684 resolveCurrentArtifact now runs the active tab's name through validateProjectPath BEFORE composing it into a path.join expression. An invalid tab (traversal segments, absolute path, null byte, reserved segment) causes resolveCurrentArtifact to fall through to the newest-artifact branch rather than abort or probe outside the project directory. Tests: - "falls through (does not throw) when active tab name contains traversal segments": injects a malformed `tabs.name = '../../../etc/passwd'` row directly via SQL (bypassing production tab-creation validation), seeds a real artifact, asserts the resolver returns the real artifact rather than the malformed name. Suite delta: +1 test (37 active in finalize-design.test.ts). Full daemon suite: 1089/1089 green. Refs nexu-io/open-design#832 * fix(contracts): publish /api/finalize as standalone runtime entrypoint Addresses @mrcfps's CI-red review on PR #832 (https://github.com/nexu-io/open-design/pull/832, inline comment on packages/contracts/package.json). The previous J3 commit added `./api/finalize` as a type-only subpath: the entry had only a `types` field, no `default`. That broke the contracts package-runtime gate (packages/contracts/tests/package- runtime.test.ts:38-47) which asserts every exports entry exposes both a `.mjs` runtime and a `.d.ts` types target. mrcfps proposed two fixes; this commit takes path B — make finalize a first-class published module rather than a type-only re-export from the package root. Path B vs path A (a peer-AI second opinion via /collaborate confirmed): under NodeNext + ESM with exports-map semantics, TypeScript validates re-exported symbols against the published module-identity surface. Because the previous J3 had `./api/finalize` neither declared as an exports-map entry nor materialized as a standalone .mjs, TS omitted the re-exported names during package boundary analysis. Even at runtime `import('@open-design/contracts').FINALIZE_SCHEMA_VERSION` worked from the bundled index.mjs but the type-checker rejected it. Path B aligns the runtime and declaration surfaces. Changes: - packages/contracts/esbuild.config.mjs: add `./src/api/finalize.ts` to entryPoints so dist/api/finalize.mjs is generated as a standalone module rather than only inlined into the bundled root. - packages/contracts/package.json: re-add `./api/finalize` to the exports map with both `default: ./dist/api/finalize.mjs` AND `types: ./dist/api/finalize.d.ts`. Mirrors `./api/connectionTest`'s shape (the canonical pattern for first-class submodule entries). - packages/contracts/src/api/finalize.ts: keep the runtime export `FINALIZE_SCHEMA_VERSION = 1` (giving the standalone module a real value to emit beyond the type-only interfaces) and update the doc-comment now that the standalone .mjs is wired. - apps/daemon/src/finalize-design.ts: switch the type import from the inline declarations introduced in the prior J3 fallback to `import type { ... } from '@open-design/contracts/api/finalize'`. Re-export the names so internal references inside finalize-design.ts keep working without touching every call site. Verified: - node --input-type=module -e "import('@open-design/contracts/api/finalize').then(m=>console.log(JSON.stringify(Object.keys(m))))" prints ["FINALIZE_SCHEMA_VERSION"] — runtime resolution clean. - pnpm --filter @open-design/contracts test: 6/6 (including both package-runtime.test.ts cases on the rebuilt exports map). - pnpm --filter @open-design/daemon typecheck: exits 0. - pnpm --filter @open-design/daemon test: 1089/1089 (no regression vs the prior J3 number). Refs nexu-io/open-design#832 --------- Co-authored-by: DevForgeAI CI/CD Engineer <devforge-ai@development.ai> |
||
|
|
1e8926271b |
Harden security scan findings and upgrade dependencies (#806)
* feat: add accent color control and launcher for Open Design * fix: remove launcher binary from PR * test: cover accent appearance edge cases * Harden security scan findings and upgrade deps * Address proxy security review * Pin jsdom for web test stability --------- Co-authored-by: ferasbusiness666 <ferasbusiness666@users.noreply.github.com> Co-authored-by: lefarcen <935902669@qq.com> |
||
|
|
d592f6087f |
feat(mcp): external MCP client with daemon-managed OAuth and 39 design-focused templates (#898)
* feat(mcp): add external MCP client with daemon-managed OAuth and 17 design-focused templates
Open Design now acts as an MCP CLIENT and surfaces tools from third-party
MCP servers to the underlying agent (Claude Code, Hermes, Kimi).
Daemon
- New mcp-config / mcp-oauth / mcp-tokens modules: persist server entries
to .od/mcp-config.json, run the OAuth dance for HTTP/SSE servers
end-to-end on the daemon (so cloud deployments work and tokens
survive across turns), and inject Authorization: Bearer headers into
the per-spawn .mcp.json the daemon writes for Claude Code (or the
ACP mcpServers map for Hermes/Kimi).
- /api/mcp/servers and /api/mcp/oauth/{start,status,disconnect}
endpoints, plus spawn-time wiring in agents that hands the configured
servers to the active agent CLI.
- System-prompt directive for connected external MCPs so the model
does not chase Claude Code's synthetic *_authenticate /
*_complete_authentication tools when the Bearer is already pinned.
Web
- Settings -> External MCP servers panel with per-row OAuth Connect /
Disconnect / Refresh affordances and per-row template hints.
- New "Add server" picker categorized into 7 groups
(image-generation, image-editing, web-capture, ui-components,
data-viz, publishing, utilities) with a search box, sticky close
button, collapsible <details> sections (auto-expand on search),
60vh capped scroll region, and a pinned Custom-server footer.
- ChatComposer /mcp slash and MCP picker button forward to the new
Settings tab; AssistantMessage renders MCP tool calls inline;
markdown autolinker handles bare http(s) URLs (incl. OAuth links)
before italic markers so OAuth callback URLs do not get
italic-fragmented mid-token.
Contracts
- packages/contracts/src/api/mcp.ts owns the wire shapes
(McpServerConfig, McpTemplate with stable McpTemplateCategory
enum, McpServersResponse, OAuth start/status/disconnect bodies, the
postMessage payload from the OAuth callback).
Templates (17 built-in)
- image-generation: Higgsfield (OpenClaw, OAuth HTTP), Pollinations,
Allyson (animated SVG), AWS Bedrock Image (uvx).
- image-editing: Imagician, ImageSorcery.
- web-capture: just-every screenshot-website-fast, ScreenshotOne.
- ui-components: 21st.dev Magic, shadcn/ui, FlyonUI.
- data-viz: AntV Chart, Mermaid.
- publishing: EdgeOne Pages.
- utilities: Filesystem, GitHub, Fetch.
Tests
- apps/daemon/tests/mcp-{config,oauth,tokens,spawn}.test.ts cover
storage round-trip, OAuth helpers, token persistence, spawn-time
wiring, every template's transport / command / args / env-field
invariants, and the canonical category enum.
- apps/web/tests/runtime/markdown.test.tsx covers the new autolinker
ordering rules.
Co-authored-by: Cursor <cursoragent@cursor.com>
* feat(mcp): add 21 more design-focused templates and a `design-systems` category
Expands the built-in MCP picker from 17 to 38 templates so users can compose
the full Open Design craft loop (design-system intake → generate → edit →
audit → publish) without leaving the Settings dialog. Every install spec is
verified live against the upstream README; templates that needed Go binaries,
multi-step `init` ceremonies, or massive runtime stacks (PostgreSQL + Redis
+ Ollama) are intentionally deferred so picking a template still resolves to
a working server in one click.
New `design-systems` category between `web-capture` and `ui-components`
(reflects the upstream-of-components position in the workflow). Mirrored in
`McpTemplateCategory` on both contracts and daemon, and `CATEGORY_ORDER` on
the web side.
New templates by category:
- image-generation (+4): prompt-to-asset (icons / favicons / OG / logos with
free-tier routing across Cloudflare AI / NVIDIA NIM / HF / Stable Horde),
Nano Banana (hosted streamable HTTP, virtual try-on + product placement),
Seedream (hosted streamable HTTP, ByteDance Seedream v3-v5 + SeedEdit),
fal.ai (uvx, 600+ models incl. FLUX / Kling / Hunyuan / MusicGen).
- image-editing (+3): Photopea (34 layered-editor tools — closes the PSD
gap), Topaz Labs (AI upscale / denoise / sharpen), Transloadit (86+ media
pipeline robots).
- web-capture (+1): Pagecast (browser → demo GIF / MP4 with auto-zoom).
- design-systems (+4, NEW category): Figma-Context (Framelink, designs →
code), Design Token Bridge (Tailwind ⇄ CSS ⇄ Figma ⇄ M3 / SwiftUI / W3C
DTCG + WCAG contrast), Design System Extractor (Storybook scrape),
Aesthetics Wiki (cottagecore / dark-academia / y2k / … moodboards).
- data-viz (+2): MCP Dashboards (45+ chart types + KPI dashboards),
Excalidraw Architect (hand-drawn architecture diagrams).
- publishing (+6): PageDrop, PDFSpark, OGForge, QRMint, Slideshot
(HTML → PDF / PPTX / PNG with 7 themes), Deckrun (Markdown → PDF / video,
hosted free tier with no key required).
- utilities (+1): A11y axe-core (WCAG 2.0/2.1/2.2 + color-contrast + ARIA).
Tests cover every new template's wiring (command, args, env / header
required-vs-optional, secret flag), the category enum invariant, and
in-category declaration order for image-generation, design-systems and
publishing buckets where the order is what users see in the picker. 21 new
test cases pass; full mcp-config suite is green.
Templates intentionally deferred (documented in PR body): figma-use
(needs Figma desktop with --remote-debugging-port=9222), m-moire (multi-step
`memi suite init` + daemon ceremony), gemini-media-mcp + trident-mcp (Go
binaries — no npx / uvx path), Pixelle-MCP (full app with web UI + ComfyUI
backend), storybook-addon-mcp (lives inside user's Storybook, not standalone),
primitiv (multi-step init / build / serve), ReftrixMCP (PostgreSQL + Redis +
Ollama + DINOv2), narasimhaponnada/mermaid (overlap with peng-shawn).
Co-authored-by: Cursor <cursoragent@cursor.com>
* feat(mcp): add figma-use template (write designs from chat) under design-systems
figma-use is the natural counterpart to Figma-Context already in this PR:
where Framelink reads Figma designs into the model, figma-use writes back
into the canvas (90+ tools — create frames / text / components / variants,
render JSX into Figma, export PNG/SVG, query nodes via XPath, lint for
WCAG / auto-layout / hardcoded colors, analyze design systems).
Wired as an HTTP MCP template (`http://localhost:38451/mcp`) because
`figma-use mcp serve` only exposes HTTP — there's no stdio mode in the
upstream `serve.ts`. No API key. Two prerequisites the user owns are
spelled out in the description so picking the template still resolves to
a working server: (1) start Figma with `--remote-debugging-port=9222`
(or `figma-use daemon start --pipe` on Figma 126+), and (2) leave
`npx figma-use mcp serve` running in a terminal.
Inserted between `design-system-extractor` and `aesthetics-wiki` so the
design-systems category reads as a workflow: read existing design (Figma
Context) → translate tokens (Token Bridge) → extract from Storybook
(Extractor) → write back to Figma (figma-use) → break creative block
(Aesthetics Wiki).
Tests cover the new template's transport (`http`), endpoint URL, the
empty header-fields invariant (no auth required), and bump the
design-systems group order to include it.
Co-authored-by: Cursor <cursoragent@cursor.com>
* feat(settings): i18n the External MCP / MCP server / Connectors sidebar entries and make the dialog header track the active section
The External MCP sidebar entry this PR introduces was hardcoded English
("External MCP / Add MCP tools (Higgsfield, GitHub…)"). Same for the
adjacent Connectors and MCP server entries. The dialog header was also
pinned to "Execution & model" copy, so opening Settings → External MCP
showed a header that lied about which section the user was on.
Adds six translation keys — `settings.connectorsTitle/Hint`,
`settings.mcpServerTitle/Hint`, `settings.externalMcpTitle/Hint` — and
translates them across all 17 locales (ar, de, en, es-ES, fa, fr, hu, id,
ja, ko, pl, pt-BR, ru, tr, uk, zh-CN, zh-TW).
`SettingsDialog` now derives the header title/subtitle from the active
section (11 sections total) instead of a single hardcoded pair, so each
section renders an honest header.
Co-authored-by: Cursor <cursoragent@cursor.com>
* test(e2e): pin level: 3 on dialog heading lookups for Pets and Connectors
CI's Validate workspace job (#1479) failed two Playwright cases with the
strict-mode violation:
getByRole('dialog').getByRole('heading', { name: 'Pets' })
resolved to 2 elements:
1) <h2>Pets</h2>
2) <h3>Pets</h3>
Same root cause as the unit-test fix already in this PR: the dynamic
dialog `<h2>` now echoes the section's own `<h3>` because the dialog
header tracks the active section. Disambiguate to `level: 3` so each
assertion still pins the section heading specifically (which is what
the test intends to verify).
Audit of the rest of e2e/ for `dialog.getByRole('heading', ...)` —
settings-api-protocol.test.ts looks for "OpenAI API" / "Anthropic API"
section h3s which never appear in the dialog `<h2>` (always
"Execution & model"), so those stay safe.
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(mcp): bind OAuth refresh to the issuing client and skip stale tokens
Persist the OAuth client context (token endpoint, client_id, client_secret,
issuer, redirect_uri, resource) alongside the bearer token so refresh hits
the same client the refresh_token was bound to (RFC 6749 §6). The previous
refresh path re-ran beginAuth with a dummy OOB redirect URI, which kept
getOrRegisterClient from finding the original DCR client and made
providers reject the refresh on the next chat turn. Refreshes now reuse
the persisted endpoint/client pair directly.
Also stop injecting expired access tokens at spawn time when refresh is
unavailable or fails. Pinning a stale Bearer made every Claude MCP call
401 while the prompt still treated the server as connected; on that path
we now skip the entry and let the UI surface a reconnect.
Generated-By: looper 0.6.1 (runner=fixer, agent=claude-code)
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
|
||
|
|
1d1df52f3b |
feat(skills/live-artifact): add 7 example dashboards + contract demo (#716)
* feat(skills/live-artifact): add 7 example dashboards + contract demo
Seven self-contained HTML prototypes under skills/live-artifact/examples/,
each with a distinct visual identity and built-in interactivity for video
demos:
stock-dashboard.html - Bloomberg-style trading floor (dark)
crypto-dashboard.html - DeFi/web3 cyber terminal with on-chain ribbon
crm-table-live.html - multi-dim CRM with Grid/Kanban/Gallery/Calendar
view switcher (light productivity)
monday-operator-live.html - editorial Monday-morning briefing (paper)
competitor-radar-live.html - mission-control radar with rotating sweep
and RGB threat tiers
baby-health-live.html - soft pastel parental panel
stock-portfolio-live/ - full live-artifact contract example: 102
escaped html_template_v1 bindings + 7
data-od-repeat blocks, ready to register
via 'tools live-artifacts create'
Each interactive HTML carries refresh-with-flash, view switching, AI
panel regeneration, clickable rows/cards that mutate state, and toast
notifications. Self-contained - only Google Fonts as external dep.
stock-portfolio-live/ demonstrates the daemon contract: template.html +
data.json + artifact.json + provenance.json. Refresh runners can rewrite
data.json without re-authoring the template.
* fix(skills/live-artifact): address PR #716 review feedback
- Unroll data-od-repeat blocks into indexed data.* bindings so renderHtmlTemplateV1 can interpolate them (it does not expand data-od-repeat or repeat-local aliases like {{t.label}}).
- Rename catalysts[].body to catalysts[].text to satisfy the bounded JSON validator's forbidden-key list (body is rejected case-insensitively); update template binding accordingly.
Generated-By: looper 0.6.1 (runner=fixer, agent=claude-code)
* fix(skills/live-artifact): make stock-portfolio provenance.json contract-compliant
- generatedBy: free-form string -> "agent" (LiveArtifactProvenanceGenerator enum)
- sources[].kind -> sources[].type with LiveArtifactProvenanceSourceType enum values
(connector for brokerage/quotes connectors, derived for AI recommendation)
- Drop non-contract per-source `note` and top-level `summary`/`transformations`/
`refreshContract`/`safetyNotes` fields; preserve their content under the
contract-allowed `notes` field so the example survives schema validation.
Generated-By: looper 0.6.1 (runner=fixer, agent=claude-code)
* fix(skills/live-artifact): use strict ISO-8601 generatedAt in provenance
The daemon's `validateIsoDate` requires `Date.toISOString()` round-trip
equality, so timezone-offset notation like `2026-05-06T14:32:18-05:00`
fails validation even though it parses. Switch to the canonical UTC form
`2026-05-06T19:32:18.000Z` (same instant), which the validator accepts.
* feat(skills): surface examples/*.html as derived skill cards + Live filter
A skill that ships hand-crafted samples under examples/*.html (e.g.
live-artifact's stock dashboard, baby health monitor) now lights up one
gallery card per file instead of a single parent card whose preview can
only ever show one of them. The parent stays in the listing tagged
aggregatesExamples=true so findSkillById and Use this prompt still
resolve back to its SKILL.md body, but the Examples tab hides it so the
derived <parent>:<child> cards aren't shadowed by a duplicate preview.
Subfolder layouts (examples/<name>/template.html + data.json) are
deliberately skipped — their templates still hold {{data.x}}
placeholders that only the daemon-side renderer fills in, so showing
the raw template would render visible braces in the gallery. Ship the
baked output as examples/<name>.html alongside the folder to surface it.
Adds an examples.modeLive filter pill (translated across all 21 locales)
that selects skill.scenario === 'live', so refreshable / connector-backed
samples are easy to find without scrolling through every desktop
prototype. live-artifact's SKILL.md gains scenario: live so it (and
every derived card) lights up there.
Co-authored-by: Cursor <cursoragent@cursor.com>
* perf(web): parallelize entry-view bootstrap so each tab renders independently
Bootstrap used to wait on a single Promise.all behind a global
'Loading workspace…' placeholder, which made the slowest endpoint
(typically /api/agents on cold start, since it probes CLI versions)
gate every tab including the ones that don't need agents at all.
Splits the global bootstrapping flag into per-resource loading flags
(agentsLoading, skillsLoading, dsLoading, projectsLoading,
promptTemplatesLoading) plus a daemonConfigLoaded flag for the merged
daemon config. Each tab now blocks only on the data it actually needs:
Examples renders as soon as skills land, Design Systems on dsList,
Designs on projects+skills+designSystems, etc.
Auto-selecting the first available agent and the default design system
moves into dedicated effects gated on daemonConfigLoaded so they no
longer race ahead of the daemon-stored choice and overwrite it with a
freshly picked first-available pick.
EntryView swaps its single loading prop for skillsLoading,
designSystemsLoading, projectsLoading, promptTemplatesLoading so each
inner tab can pick the right gate without leaking the parent's coarse
state.
Co-authored-by: Cursor <cursoragent@cursor.com>
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
|
||
|
|
e14b8092ea |
feat: add Orbit activity summaries (#681)
* feat: add Orbit activity summaries * fix(orbit): make runs navigable while agent continues * fix(web): widen minimum chat panel * feat: support Orbit template selection * fix(daemon): avoid bogus skill side-file preflight * fix(web): collapse orbit artifact project cards * fix(web): preserve orbit project card titles * fix: improve Orbit run daily briefing * fix: handle Orbit digest data failures * fix: load Orbit templates and connector tools reliably * fix: keep Orbit summary counts consistent Generated-By: looper 0.6.1 (runner=fixer, agent=opencode) * fix: apply Orbit template skill context * fix: cache and curate connector tools for Orbit * fix: align Orbit defaults and connector discovery * fix: simplify Orbit template settings * fix: move connectors into settings * fix: compact connector settings catalog * fix: address Orbit PR feedback Generated-By: looper 0.6.1 (runner=fixer, agent=opencode) * fix: address Orbit PR feedback Generated-By: looper 0.6.1 (runner=fixer, agent=opencode) * fix: address Orbit PR feedback Generated-By: looper 0.6.1 (runner=fixer, agent=opencode) * fix: address Orbit PR feedback Generated-By: looper 0.6.1 (runner=fixer, agent=opencode) * fix: address Orbit PR feedback Generated-By: looper 0.6.1 (runner=fixer, agent=opencode) * fix: address Orbit PR feedback Generated-By: looper 0.6.1 (runner=fixer, agent=opencode) * fix: address Orbit PR feedback Generated-By: looper 0.6.1 (runner=fixer, agent=opencode) * fix: address Orbit PR feedback Generated-By: looper 0.6.1 (runner=fixer, agent=opencode) * fix: prevent connector action button from stretching into pill The icon-only connect/disconnect buttons in the embedded connectors catalog inherited min-width: 92px / 106px from the non-embedded pill rules, overriding the 24px square sizing and causing the buttons to overlap the card head text. Reset min-width to 0 in the embedded icon-only rule so the compact square layout holds. * fix(web): align live artifact file rows * fix: clean up Orbit connector settings lifecycle Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix: address Orbit review regressions Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * feat(web): localize Orbit and connector settings * feat(web): gate Orbit runs without connectors * feat(web): refine connector settings UX * feat(web): safeguard Composio key clearing * fix(web): refresh Composio tool badges * feat(web): show connector logos * feat(daemon): localize Orbit prompt window * fix(daemon): clarify blocked connector callback closes * test(daemon): harden flaky async probes * fix(web): align Indonesian connector locale keys * test(web): align connector browser props * fix(web): preserve explicit credential clears Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(daemon): time out Composio logo proxy fetches Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): localize Indonesian connector settings copy Translate the new connector settings strings in the Indonesian locale and lock them with a regression test so this surface no longer silently falls back to English. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): preserve discovered connector tools Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): preserve onboarding autosave completion Keep settings autosave from clearing onboarding completion after the close gesture, and expose the desktop main types from source so workspace validation can typecheck packaged imports without a prior desktop build. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(daemon): defer Composio catalog cache hydration Load persisted Composio catalog data only after the runtime data directory is configured so startup cannot read another namespace's cache. Add a regression test that exercises the module-load singleton path. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): treat discovery completion independently Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): preserve latest settings draft on close Use the latest persisted settings draft when the dialog closes so onboarding completion does not race a stale daemon sync and overwrite newer Orbit/template selections. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): avoid syncing draft Composio key on Orbit run Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): localize Orbit settings copy Translate the new Indonesian Orbit and autosave strings so the settings UI no longer falls back to English and the locale regression stays covered. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): prefer fresh connector catalog state Keep refetched connector status/auth data authoritative while retaining discovery-only tool metadata so the connectors UI stays consistent after refreshes. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): declare Indonesian locale fallback keys explicitly Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): inline Indonesian fallback strings for CI Replace the Indonesian locale's per-key English lookups with explicit strings so workspace typecheck no longer depends on brittle build-mode resolution in CI. Add a regression test that blocks those per-key English lookups from reappearing in the CI-sensitive fallback sections. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(daemon): restrict proxied connector logos to image MIME types Reject non-image upstream logo responses so the daemon never serves third-party HTML from its localhost origin. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * test(e2e): align settings dialog regressions Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): decouple Orbit runs from media sync failures Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): keep SPA catch-all export-compatible Disable dynamic catch-all params for the exported SPA shell so Next.js static builds can emit the root route again. Add a regression test covering the route config against the web export mode. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): preserve Orbit config and workspace routes Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(daemon): block SVG in connector logo proxy Reject SVG and other unsafe proxied logo responses so third-party logo content cannot execute under the daemon origin, while keeping raster logo fetches working and making rejected responses non-cacheable. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(daemon): fall back to static catalog for empty cache Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): disable Orbit run before connector gate resolves Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(desktop): export shipped desktop types Point the desktop ./main type export at the generated declaration so installed consumers resolve the published file set. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): restore persisted question form selections Render historical submitted answers directly so reloaded question forms keep their locked selections visible. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): retry forced media sync autosave Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(daemon): keep Composio logo timeout through body read Keep the Composio logo fetch timeout active until the response body is fully consumed so stalled body reads abort and clear the inflight cache entry. Add a regression test that proves a delayed body read times out and the next request can recover.\n\nGenerated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): refresh Orbit gate after connector auth Re-check connector availability when the settings window regains focus so Orbit unlocks as soon as a connector finishes authenticating in the same settings session. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(daemon): keep connector detail tool lists intact Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(daemon): ignore malformed Orbit summaries Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(e2e): stabilize design-system multi-select flow Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(daemon): cap Composio logo cache growth Bound the Composio logo cache with LRU eviction and expired-entry pruning so repeated untrusted logo requests cannot grow daemon memory without limit. Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(daemon): bound proxied Composio logo payloads Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): align autosave settings tests Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): remove stray CSS conflict marker Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fixer: address PR #681 follow-up items Generated-By: looper 0.6.2 (runner=fixer, agent=opencode) * fix(web): restore restart routes and connector flows * fix(web): keep SPA export route static * fix(web): stabilize chat scroll tests --------- Co-authored-by: lefarcen <935902669@qq.com> |
||
|
|
e52720aa12 |
feat(daemon): add language boost support for Minimax TTS (#773)
* feat(daemon): add language boost support for Minimax TTS Add --language CLI flag to support language boost parameter for Minimax TTS. This enables better pronunciation for specific languages like Cantonese (Yue). * docs(media): add --language flag to media generation contract Document the language boost parameter for Minimax TTS, enabling better pronunciation for specific languages like Cantonese (Yue). * fix(media): correct Cantonese language_boost value and add input validation - Use correct MiniMax value 'Chinese,Yue' for Cantonese (no space) - Add type guard in server.ts to reject non-string language values - Trim language string before sending to MiniMax API --------- Co-authored-by: root <root@DELLN40.asiacredit.org> |
||
|
kamiandGitHub
|
2eae7da24b |
feat: support Cloudflare Pages custom domains (#851)
* Support Cloudflare Pages custom domains without hiding pages.dev fallback Keep the default Pages preview as the first public link while optional owned-zone binding provisions DNS and Pages custom-domain state in parallel. Constraint: Cloudflare deploys must use the existing direct-upload API path with no Wrangler dependency. Constraint: pages.dev must stay visible even while custom-domain verification is pending. Rejected: Vercel custom-domain support | outside requested Cloudflare-only scope. Rejected: overwriting arbitrary CNAME records | risks taking over user-managed DNS. Confidence: high Scope-risk: moderate Directive: Do not expose providerMetadata through public deploy contracts; keep custom-domain DNS ownership checks conservative. Tested: pnpm --dir apps/daemon exec vitest run -c vitest.config.ts tests/deploy.test.ts tests/deploy-routes.test.ts Tested: pnpm --filter @open-design/contracts build && pnpm --filter @open-design/contracts typecheck && pnpm --filter @open-design/contracts test Tested: pnpm --filter @open-design/web typecheck && pnpm --filter @open-design/web test -- providers/registry.test.ts components/FileViewer.test.tsx i18n/locales.test.ts Tested: pnpm i18n:check && pnpm guard && pnpm typecheck Tested: pnpm --filter @open-design/daemon build && pnpm --filter @open-design/web build && git diff --check Not-tested: real Cloudflare account/token/domain smoke test * Preserve Cloudflare fallback correctness under large accounts and races Constraint: Cloudflare Pages keeps pages.dev as the primary usable fallback while custom domains remain optional typed metadata. Rejected: Treating custom-domain DNS or binding failure as a top-level deployment failure | pages.dev can still be ready and usable. Confidence: high Scope-risk: moderate Directive: Keep custom-domain finality tied to Cloudflare Pages API active status plus URL reachability; do not expose providerMetadata. Tested: pnpm --dir apps/daemon exec vitest run -c vitest.config.ts tests/deploy.test.ts tests/deploy-routes.test.ts; pnpm --filter @open-design/web test -- components/FileViewer.test.tsx i18n/locales.test.ts providers/registry.test.ts; pnpm --filter @open-design/daemon typecheck; pnpm --filter @open-design/web typecheck; pnpm i18n:check; git diff --check; pnpm guard; pnpm typecheck; pnpm --filter @open-design/daemon build; pnpm --filter @open-design/web build Not-tested: Real Cloudflare token/account/zone smoke test. * Keep impeccable design notes local Constraint: .impeccable.md is local assistant/design context and should not be part of the PR diff. Rejected: Keeping the file tracked while adding it to .gitignore | tracked files are not ignored by Git. Confidence: high Scope-risk: narrow Directive: Keep .impeccable.md untracked and ignored; do not rely on it for required project documentation. Tested: git check-ignore -v .impeccable.md; git diff --check Not-tested: Full workspace tests not rerun for ignore-only metadata change. |
||
|
|
56bf6ee1b6 |
feat: agent-callable research command and /search (#615)
* feat: pre-generation research (Tavily) for grounded generation
Adds an optional pre-generation research step so the agent can produce
slides / prototypes / decks grounded in real sources instead of guessing.
User flow:
1. Settings -> Tavily Search -> paste API key (or set TAVILY_API_KEY).
2. Click the new Research button in the chat composer.
3. On send, the daemon runs a Tavily search, prepends the findings
as a <research_context> block ahead of the system prompt, and
spawns the agent. Research progress shows up as status pills in
the chat stream; the agent cites sources inline as [1]/[2]/...
Phase 1 surface:
- Single provider (Tavily), single depth ('shallow'), no LLM
synthesis pass (Tavily's `answer` is the summary).
- Composer toggle only; no popover / depth picker yet.
- Reuses the existing `status` SSE agent payload + StatusPill UI
so no new event variants or renderer code are needed.
Layers touched:
- contracts: ResearchOptions / Source / Findings DTOs;
ChatRequest.research; export from index.
- daemon: apps/daemon/src/research/{index,tavily}.ts orchestrator
+ provider; tavily added to MEDIA_PROVIDERS and ENV_KEYS; hook
in startChatRun before prompt assembly.
- web: ChatComposer toggle + ChatSendMeta; threaded through
ChatPane / ProjectView / streamViaDaemon into ChatRequest.
Side fix (required to land the feature, but useful on its own):
contracts internal relative imports lacked the `.js` suffix that
NodeNext module resolution requires. This was already breaking
`pnpm --filter @open-design/daemon typecheck` on main; without the
fix, none of the new research types were visible to the daemon.
All internal contracts imports now carry `.js`.
Spec: specs/current/research-feature.md (phases 2-4 outlined for
follow-up: composer popover, multi-provider, deep recursion, example
skills with research_recommends).
Verified:
- pnpm --filter @open-design/contracts typecheck/test
- pnpm --filter @open-design/daemon typecheck (the chokidar
project-watchers test is a pre-existing flake, unrelated)
- pnpm --filter @open-design/web typecheck
- node scripts/verify-media-models.mjs
* fix(daemon): clamp Tavily max_results to 20
Tavily's /search endpoint requires `max_results` in [0, 20]; sending a
larger value (e.g. when `research.depth: "deep"` resolves to 30) returns
400 and `runResearch` silently falls back to no-research. Clamp at the
provider boundary so Phase 2 depth tiers above 20 still produce results
instead of failing the request.
Generated-By: looper 0.6.1 (runner=fixer, agent=claude-code)
* Remove stale research merge leftovers
* Add agent-callable research search
* Fix Indonesian locale typecheck
* Fix research command invocation edge cases
* Harden slash search prompt expansion
* Honor research source caps in command contract
* Require search reports in design files
* Add research data provider settings
* Wire web research provider fallback order
* Update research provider fallback wording
* Revert "Update research provider fallback wording"
This reverts commit
|
||
|
|
988fd6db5e |
feat: import existing local folder as project (#597) (#624)
* feat(contracts): types for folder-import endpoint
Add ImportFolderRequest, ImportFolderResponse to the public contract
surface. Extend ProjectMetadata with a baseDir field — when set, the
project's files live at this absolute path instead of .od/projects/<id>/.
Stored as the realpath() result so symlinks cannot redirect later writes.
Refs nexu-io/open-design#597
* feat(daemon): support metadata.baseDir for folder-rooted projects
Add resolveProjectDir() and metadata-aware variants of listFiles,
readProjectFile, writeProjectFile, ensureProject so a project's files
can live under metadata.baseDir (the user's chosen folder) instead of
.od/projects/<id>/. metadata.baseDir is opt-in — projects without it
keep the existing .od/projects/<id>/ behavior unchanged.
When listFiles walks a baseDir-rooted project, it skips conventional
build / install dirs (node_modules, .git, dist, build, .next, .nuxt,
.turbo, .cache, .output, out, coverage, __pycache__, .venv, vendor,
target, .od, .tmp) so the file panel stays focused on design content
instead of being dominated by lockfiles and node_modules.
Add detectEntryFile() — best-effort lookup for index.html or any
.html at the folder root, used by the import endpoint to seed the
initial active tab.
Refs nexu-io/open-design#597
* feat(daemon): add POST /api/import/folder endpoint
Creates a project rooted at the submitted local folder. metadata.baseDir
points at that folder and OD reads / writes there directly — no copy,
no shadow tree, mirroring how Cursor / Claude Code / Aider behave. The
user owns the workspace and is responsible for their own version
control.
Safety:
- baseDir is canonicalized via fs.promises.realpath() at import time so
user-controlled symlinks can't redirect later writes. resolveSafe
enforces the bounds check against the literal stored path; without
realpath, a symlink (e.g. ~/sneaky → /etc) would let writeProjectFile
escape the project tree at every later call because the OS follows
the symlink at open() time.
- Post-realpath lstat ensures the canonical target is itself a real
directory (defense-in-depth).
- The data directory (RUNTIME_DATA_DIR) and its descendants are
refused after symlink resolution so a redirect into the daemon's
own state can't masquerade as a project import.
The web client wires this through state/projects.ts → App.tsx,
landing the user on the auto-detected entry file when present.
Refs nexu-io/open-design#597
* feat(desktop): expose native folder picker to renderer
Adds an Electron preload script that exposes window.electronAPI.pickFolder
via contextBridge. Wires dialog.showOpenDialog through ipcMain so the
web UI can open a native folder selector for project import. Browser-only
users fall back to a text input for the absolute path (handled in the
web layer); the picker stays an optional convenience on the desktop
binary.
ipcMain.handle() registers handlers in an internal map that is not
exposed via eventNames(), so the natural-looking guard
if (!ipcMain.eventNames().includes('dialog:pick-folder')) ipcMain.handle(...)
is always true. On a second createDesktopRuntime() call (dev hot-reload,
packaged-vs-electron mode swap) the body re-runs and ipcMain.handle()
throws 'Attempted to register a second handler'. Use removeHandler()
+ handle() unconditionally — removeHandler() is a documented no-op
when nothing is registered, making the pair idempotent.
Includes *.cts in the apps/desktop tsconfig so the preload script is
typechecked.
Refs nexu-io/open-design#597
* feat(web): add 'From existing folder' option to New Project
UI surface for the import flow:
- A new 'Open folder' affordance in NewProjectPanel that uses the
native picker on Electron (window.electronAPI.pickFolder) and falls
back to an absolute-path text input in the browser.
- importFolderProject() in state/projects.ts: typed wrapper around
POST /api/import/folder using @open-design/contracts types.
- App.tsx wires the response: prepend the new project to the list,
navigate to it, and select the auto-detected entry file as the
active tab.
Skill / design-system pickers from the existing prototype tab are
reused — folder import is a project-creation flow, not a separate
project type.
Refs nexu-io/open-design#597
* docs(architecture): document folder-import endpoint
Adds POST /api/import/folder to the daemon API table and a 'Folder
import' section explaining the single-mode design (direct read/write
in metadata.baseDir, mirroring Cursor / Claude Code / Aider), the
realpath() canonicalization, the RUNTIME_DATA_DIR refusal, and the
SKIP_DIRS list applied to listFiles for baseDir-rooted projects.
Refs nexu-io/open-design#597
* test(daemon): unit + integration tests for folder import
Two new files:
apps/daemon/tests/folder-import-projects.test.ts (13 unit tests):
- resolveProjectDir behavior under all metadata combinations,
including the fallback when baseDir is relative and the
isSafeId-bypass when baseDir is set
- detectEntryFile: index.html priority, .html fallback, null when
no html, no descent into subdirs
- listFiles with metadata.baseDir: walk, SKIP_DIRS hides node_modules
/ .git / dist, back-compat for projects without baseDir
apps/daemon/tests/folder-import-route.test.ts (10 integration tests):
- Happy path: baseDir stored in metadata, importedFrom='folder',
conversation created, entry file detected
- Error paths: missing baseDir, empty, relative, non-existent,
pointing at a file
- Security: realpath canonicalization (the symlink test was the one
that surfaced the original /var vs /private/var mismatch in
RUNTIME_DATA_DIR comparison on macOS)
- Security: a symlink that resolves into RUNTIME_DATA_DIR is rejected
after realpath, not before
Refs nexu-io/open-design#597
* fix(daemon): wire baseDir metadata into chat + deploy reads
Two bugs caught in Codex automated review of #624:
1. chat-route was passing the metadata object directly as the listFiles
opts argument: `listFiles(PROJECTS_DIR, projectId, chatMeta)`. The
listFiles contract reads opts.metadata, not opts itself, so this
silently fell back to .od/projects/<id>/ instead of the imported
folder. existingProjectFiles was empty for baseDir-rooted projects.
Wrap as `{ metadata: chatMeta }`.
2. deploy.ts read project files via readProjectFile without the
metadata third argument, so for baseDir-rooted projects the deploy
and preflight endpoints would look in .od/projects/<id>/ and fail
with file-not-found instead of reading the imported folder. Thread
options.metadata through buildDeployFilePlan → readProjectFile and
pass project?.metadata at the two server.ts callsites
(`POST /api/projects/:id/deploy` and the preflight endpoint).
Add a regression test that locks the listFiles contract: passing a
bare metadata object as opts must NOT scan baseDir — it must fall back
to the standard project dir, otherwise callers can leak the wrong
folder by mistake.
Refs nexu-io/open-design#597, #624 (Codex review)
* fix(daemon): ensure correct metadata handling in folder import
Addressed issues with metadata handling in folder import functionality. Updated the listFiles and readProjectFile methods to correctly utilize the metadata.baseDir, ensuring that project files are read from the intended directory. Added regression tests to verify that passing a bare metadata object does not inadvertently scan the baseDir, maintaining the integrity of project file access.
Refs nexu-io/open-design#597
* fix(daemon): security hardening from Codex review of #624
P1 findings from automated review:
1. POST /api/projects + PATCH /api/projects/:id rejected
client-supplied metadata.baseDir. baseDir is privileged: it lets a
project root inside the user's filesystem, and the realpath() +
RUNTIME_DATA_DIR reentry checks live only on /api/import/folder.
Allowing it on the generic create/patch path lets an attacker
smuggle e.g. /etc through and bypass every import-time guard.
Both endpoints now refuse a baseDir field with 400.
2. resolveSafeReal() helper: realpath()s each candidate path (or its
longest existing prefix for write paths) and re-validates against
realpath(projectRoot). The original resolveSafe() only did a
string-prefix check, which was fooled by symlinks *inside* a
baseDir-rooted project. A repo containing 'assets -> /Users/me/.ssh'
passed the literal prefix check but readFile() followed the link
at open() time. resolveSafeReal() is now used by readProjectFile,
writeProjectFile, and deleteProjectFile.
3. Multer chat-upload destination now resolves to metadata.baseDir for
imported folder projects via a module-level lookup wired to db at
startServer() boot. Previously attachments landed in
.od/projects/<id>/ even for baseDir projects, so the agent (which
runs with cwd=baseDir) couldn't open them.
P2 findings:
4. searchProjectFiles threads metadata through listFiles +
resolveProjectDir so /api/projects/:id/search hits the right tree.
5. buildProjectArchive + buildBatchArchive now accept metadata so
'Download .zip' works for imported folder projects.
6. Watcher subscribe() resolves to baseDir for imported projects so
live-reload SSE actually fires when the user edits files in their
own folder. Registry stays keyed by the canonical directory.
7. Template snapshotting reads source-project files with metadata
so a template can be saved from a baseDir-rooted source.
Tests:
- Regression: POST /api/projects with metadata.baseDir → 400.
- Regression: descendant symlink (assets/leak.txt -> /etc/hosts) is
refused on the raw read endpoint.
Refs nexu-io/open-design#597, #624 (Codex P1+P2 review)
* fix(daemon): close two regressions found in #624 review round 2
@mrcfps caught two more correctness gaps:
1. Archive root symlink escape — buildProjectArchive accepts an optional
?root=<subdir> param to scope the zip to a subdirectory. The path was
resolved with the string-only resolveSafe(), so a directory symlink
inside an imported folder (docs -> /Users/me/.ssh) passed the prefix
check and collectArchiveEntries() then walked outside the project
tree. Switch to the symlink-aware resolveSafeReal() — the same one
that already protects raw read/write/delete paths. The walker itself
already skips dirent symlinks via !isDirectory && !isFile, so
canonicalizing the root is the only missing piece.
2. PATCH metadata wiped baseDir — updateProject() replaces metadata
wholesale. The previous guard only blocked an explicit baseDir
change, but a normal patch that *omits* baseDir (a UI editing
linkedDirs only sends { metadata: { kind, linkedDirs } }) silently
detached imported projects from their folder root. Subsequent
reads/writes/watch/deploy fell back to .od/projects/<id>.
Re-stamp the immutable folder-import fields (baseDir, importedFrom='folder')
from the existing project record onto the incoming patch when the
project is imported. A patch that supplies a *different* baseDir
still gets rejected as before; a patch that supplies the *same*
baseDir is accepted as a no-op. A patch on a non-imported project
that tries to set baseDir is also still rejected (preserves the
POST /api/projects guard from the previous round).
Tests:
- archive endpoint: ?root=<symlink-to-/etc> → 400.
- patch endpoint: PATCH that omits baseDir on an imported project keeps
baseDir intact (project still resolves to the user's folder after).
Refs nexu-io/open-design#597, #624 (Codex P1 round 2)
* fix(web): add Indonesian deploy provider copy
---------
Co-authored-by: INFINITY <valentyn.sotov@trendarena.app>
Co-authored-by: Siri-Ray <2667192167@qq.com>
|
||
|
kamiandGitHub
|
09eb88f683 |
Add Cloudflare Pages artifact deployment
Adds Cloudflare Pages artifact deployment support. |
||
|
monshunterandGitHub
|
e6e5928be1 |
feat(web): add connection tests for execution settings (#507)
* feat(settings): add connection test for providers and CLI agents Adds a "Test" action in the Settings dialog that verifies the configured provider (Anthropic/OpenAI/Azure/Google) or CLI agent without sending a real chat. Backed by a new daemon endpoint and shared contracts, with categorized inline statuses and i18n strings across all supported locales. * fix(settings): address connection test review feedback * fix(daemon): pass empty MCP servers for connection probes * fix(connection-test): address review blockers * fix(daemon): fail json stream runs on structured errors * fix(contracts): build connection test subpath export * Use draft CLI env in agent connection tests * fix(i18n): add fallback ids for new curated content |
||
|
SidandGitHub
|
33255a8fdf |
Fix agent CLI config and workspace focus mode (#604)
* fix agent CLI config and workspace focus mode * address CLI env review follow-ups |
||
|
Justin GaoandGitHub
|
cbe2baf596 |
feat(web): add skills & design systems management page in settings (#535)
* feat(web): add skills & design systems management page in settings Add a new "Library" section in Settings that lets users browse, search, preview, and enable/disable skills and design systems. Disabled items are excluded from the create-project picker. Phase 1 — browse/toggle only. Closes #497 * fix(web): persist empty disabled lists and deduplicate DS preview Use empty array instead of undefined when all items are re-enabled so the daemon merge clears the key. Move DS preview panel outside the category group loop so it renders once, not per group. * fix(web): address review feedback on library settings Clear disabled lists on invalid daemon writes, memoize enabled item filters in App.tsx, and guard preview fetch against rapid-click race conditions. * fix(web): hydrate disabled lists from daemon and keep full lists in ProjectView Merge daemonConfig.disabledSkills/disabledDesignSystems during bootstrap so the values survive localStorage resets. Pass unfiltered skills and design systems to ProjectView so existing project metadata resolves correctly. |
||
|
|
79fcaef129 |
Add Tweaks mode for HTML previews with picker, pod selection, and batched chat attachments (#513)
* Add tweaks mode for HTML preview comments * Fix tweaks geometry and restore critique migration * Harden tweaks mode reload sync * Guard tweaks batch sends during active runs --------- Co-authored-by: puma <puma@pumas-MacBook-Air.local> |