mirror of
https://github.com/nexu-io/open-design.git
synced 2026-06-01 03:14:35 +07:00
593 commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
lefarcen
|
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.
|
||
|
lefarcen
|
7b191b5f85
|
fix: load Orbit templates from design templates (#1442)
(cherry picked from commit
|
||
|
lefarcen
|
4d8d233ce0
|
Fix Langfuse report finalization hook (#1402) | ||
PerishFire
|
e6c5560884
|
Fix appearance accent color persistence (#1439) | ||
|
lefarcen
|
2a0ebea50b |
release: Open Design 0.7.0
- bump 14 monorepo package.json files to 0.7.0 (root + apps/{web,daemon,desktop,packaged,landing-page} + packages/{contracts,platform,sidecar,sidecar-proto} + tools/{dev,pack,pr} + e2e); apps/packaged was already at 0.6.1 from beta lane, all others at 0.6.0
- add CHANGELOG.md [0.7.0] - 2026-05-12 entry covering 97 merged PRs since 0.6.0:
- Critique Theater: Phase 7 web client state machine (#1307) + Phase 6.2 daemon artifact extraction (#1085)
- Web/UI: thumbs-up/down feedback widget (#1308), Cmd+, opens Settings (#1173), Finalize design package + Continue in CLI (#974), fetch models button for BYOK (#1034), provider models alphabetical sort (#1097), collapsible MCP JSON field-mapping (#1136), design file rename (#894)
- Daemon: auto-memory store with chat-protocol-aware extraction (#999), install/uninstall skills & design systems (#1003), HTTP 206 range requests for video/audio (#1105), scheduled routines (#1033), agent runtime + route registration refactor (#1063, #1043)
- HyperFrames: HTML-in-Canvas across web + skills (#866)
- Skills/design systems: generic skills + design-templates split + finalize-design API (#955), agent-browser skill (#1284), WeChat design system + login-flow skill (#1083), hud/loom/trading-terminal design systems (#1069), release-notes-one-pager skill (#873), tokens.css schema (#1231)
- Packaging: macOS Intel (x64) build (#759), official Nix flake (#402), beta packaging cache (#1095)
- Maintainer ops: tools-pr PR-duty workspace (#1259), MAINTAINERS.md (#1290), contributor card bot (#932), PR→issue linking discipline (#1263)
- Changed: conversation run isolation (#1271), default English i18n fallback (#1270), Codex CLI exit diagnostics / empty-response handling / path fallback (#1267, #1244, #1205)
- Fixed: ~30 web + desktop + daemon + packaging bugfixes
- Internal: nightly UI/desktop regression coverage (#1256), e2e/release report hardening (#1140), entry/settings automation (#954)
- catch up [Unreleased] compare link to v0.7.0 and add missing [0.6.0] release link
- add 97 PR footnote refs ([#402]..[#1330])
Verified locally: pnpm install + pre-build contracts/daemon/desktop dist + pnpm typecheck (exit 0 across all 14 packages on Node 22.22 with engine-warning).
Release workflow validation runs after merge via release-stable.
|
||
shangxinyu1
|
5d674410f2
|
test: stabilize extended Playwright coverage (#1341)
* test: align extended Playwright coverage with current UI behavior * test: address extended suite review feedback * test: restore Codex path hydration assertion |
||
Eli
|
9c489aa045
|
feat(web): redesign Designs tab cards — covers, tags, overflow menu, multi-select (#1161)
* feat(web): redesign Designs tab cards — covers, tags, overflow menu, multi-select - Render real previews on project cards: HTML iframe / image / video / hashed gradient fallback with project initial; lazily fetches the project's primary file when metadata.entryFile is unset, prefers index.html → newest html → image → video. - Live artifact card thumbnails embed the rendered artifact URL via sandboxed iframe. - Replace the per-card close button with a `…` overflow menu (Rename, Delete) that opens on hover/click; click-outside and Esc close it. - Add multi-select mode (toolbar toggle → checkbox per card → "N selected · Delete · Cancel" pill) with batch delete via the existing onDelete prop. - Add a category tag to every card (Prototype / Live Artifact / Slide / Media) derived from project.metadata.intent / kind / skillId. - Replace browser prompt() and confirm() with custom modals (rename input + danger-confirm) reusing the existing .modal shell. - Add `more-horizontal` icon and 16 new i18n keys across all 18 locales (zh-CN/zh-TW localized; others fall back to English). * test(e2e): update home delete flow for overflow menu + custom confirm modal The previous flow targeted a per-card X button labelled "delete project <name>" and asserted on a native `dialog` event. The card UI now exposes a `…` overflow menu and a styled confirm modal, so reach delete via the menu and assert against the modal's Cancel / Delete buttons instead. * fix(web): harden Designs tab preview sandbox * fix(web): hide Designs select mode in kanban |
||
|
Eli
|
77f69257a7
|
feat(web): in-context comment thread for the artifact preview (#1276)
* feat(web): free-pin fallback in comment mode for unannotated artifacts When the artifact has no data-od-id annotations, clicking in Comment mode now posts a synthetic position-based target so the host opens a popover at the click location. Daemon upsert validation requires a non-empty selector/label, so the pin uses [data-od-pin=ID] and label 'pin'. Coordinates are document-space (viewport + scrollY) so pins stay anchored after scroll/reload. Clicks on interactive elements (a/button/input/textarea/select/label/contenteditable) keep their native behavior and are not pinned. * feat(web): tighten comment popover layout for free-pin and element targets The popover header used to dump the raw elementId verbatim — fine for data-od-id targets like 'hero-cta' but jarring for free-pins where elementId is a synthetic 'pin-...' string. Branch on the prefix and show 'Pin · at X, Y' for free-pins; keep the label + selection kind for real element / pod targets. Replace the text 'Close' button with an icon-only close affordance to match the popover-as-card visual. Action row is now two right-aligned buttons (Comment + Send to Claude) for element targets and (Add note + Send to Claude) for pod targets, eliminating the three-button row that wrapped onto two lines at narrow widths. The 'Remove' affordance for existing comments stays left-aligned. * feat(web): drop comments tab from chat sidebar The chat sidebar's 'Comments' tab listed saved/attached preview comments but duplicates the per-element popover already shown in the artifact viewer. Hide the tab and its content while the right-side comment thread panel takes over the same surface in-context. The CommentsPanel / CommentSection components stay defined as dead code for the moment so callers and translation keys remain valid; a later pass can delete them. * feat(web): right-side comment thread panel in board mode Render a 320px CommentSidePanel anchored to the right of the artifact preview whenever board (comment) mode is on. The panel lists every saved preview comment for the current file with an avatar initial, the element label (or 'Pin' for free-pin synthetic ids), an Xd/Xh/Xm-ago timestamp, the note body, a Reply link, and a checkbox. Reply focuses the comment's element via liveSnapshotForComment so the popover opens at the right anchor. Selecting one or more comments via the checkboxes surfaces a 'N selected · Clear · Send to Claude' action bar above the list; Send to Claude reuses the existing onSendBoardCommentAttachments pipeline via commentsToAttachments. The panel takes the place of the chat sidebar's removed Comments tab so the thread lives next to the artifact instead of behind a tab switch. * feat(web): styles for right-side comment thread panel Floating 320px panel anchored to the right edge of the artifact preview with a scrollable comment list and a coral selection bar that appears when one or more comments are checked. Selected items get a coral tint; the reply / check / send-to-claude controls match the popover's coral primary tone. * feat(web): toast confirmation on comment save, close popover After savePersistentComment succeeds, close the popover via clearBoardComposer and surface a transient 'Comment saved' (or 'Pin saved' for free-pin targets) toast for 2.2s. Replaces the previous behavior where the popover stayed open with an empty draft after save, which left users uncertain whether the save landed and forced an extra click to dismiss. * feat(web): position the comment-save toast at the top of the preview * feat(web): allow editing saved comment notes via the side panel Rename the per-item 'Reply' affordance to 'Edit' (no thread model exists yet, so reply was misleading) and pre-fill the popover with the existing note when clicked. The save path goes through onSavePreviewComment which the daemon implements as an upsert keyed on (project, conversation, filePath, elementId), so the edit overwrites the existing row's note without spawning a duplicate. Also fall back to a snapshot synthesized from the saved comment's own fields when the corresponding live target is no longer in the iframe DOM (e.g. free-pin parents that were re-rendered), so the edit path still works after artifact reloads. * feat(web): hide already-sent comments from the side panel After Send to Claude, the daemon flips the comment status from 'open' to 'applying' (and then 'needs_review' / 'resolved' / 'failed' depending on the run). Filter the side panel to status === 'open' so sent comments visibly leave the list — the user gets clear feedback that the send landed and the panel stays focused on actionable, un-sent items. * feat(web): drop single-tab bar and conversation count badge After the Comments tab was removed the chat header still rendered a one-tab 'tablist' just for the Chat tab, which read as visual noise without a sibling to switch between. Drop the tabs wrapper entirely; the chat content stays mounted and the header now hosts only the conversation-history affordance. Also drop the numeric badge that overlaid the conversation history button: counting open conversations next to a generic history icon was easy to mistake for an unread / notification count. The dropdown itself remains the canonical place to see and switch between past conversations. * feat(web): right-align chat header actions after tab bar removal With the tabs wrapper gone, chat-header-actions sat flush left because nothing was pushing it across the header. Add margin-left: auto so the history / new-conversation / collapse buttons land at the right edge, matching the design files / index.html tab row's own right-aligned controls. * feat(web): rename board-mode toggle to Comment with comment icon The artifact preview toolbar's board-mode entry was labeled 'Tweaks' with the tweaks icon, which collided with the palette Tweaks button next to it and hid the comment capability behind a generic label. Rename to 'Comment' with the comment icon and switch to the viewer-action class so the button matches the surrounding toolbar items (Edit/Draw) and the coral active state lands on the right surface. * fix(web): pass designTemplates to ProjectView in api-empty-response test The test props for ProjectView were missing the designTemplates prop that was added to Props in #955 (generic skills split). CI's strict typecheck (tsc -b --noEmit) caught it; local runs that hit project references differently did not. Pass an empty SkillSummary array — matches the empty skills fixture for the same reason. |
||
|
Eli
|
928079daf5
|
feat(web): consolidate Image/Video/Audio entries into a Media tab (#1167)
Reduces the New Project panel's top-level tab count by collapsing the three media surfaces into a single Media tab with an inner segmented control, and polishes the controls inside that tab so they stop dominating the panel: - Media tab + segmented (Image / Video / Audio) inside the panel body. Underlying ProjectKind branches and submission contract unchanged — the daemon still receives kind=image/video/audio. - Model picker rewritten as a combobox: one trigger row + searchable, provider-grouped popover with Recommended badges. Replaces the flat grid of provider-grouped cards that scrolled past the fold once the fourth provider landed. - Aspect picker compressed from a 5-card grid to a single row of segmented pills with mini ratio glyphs. - Image surface no longer carries a free-form Style notes field; it was redundant with the prompt template + main prompt input. - Live artifact tab locks fidelity to high-fidelity (the wireframe option is now hidden) — a wireframe live artifact doesn't make sense and the picker added noise. i18n: adds tabMedia / titleMedia / model* keys across all 18 locales, removes imageStyleLabel / imageStylePlaceholder. Tests + e2e selectors updated to drive the new Media tab + segmented surface flow. |
||
|
Eli
|
1b307bf17f
|
feat(web): tweaks palette popover with HSL hue-shift recoloring (#1292)
* feat(web): tweaks palette popover with HSL hue-shift recoloring Adds a Tweaks color-palette popover to the HTML preview toolbar. Selecting a palette re-skins the iframe in place via a srcDoc-side bridge that walks the DOM and shifts every chromatic paint to the target hue while preserving each color's saturation and lightness — pale tints stay pale, bold CTAs stay bold, just in the new color family. Mono-noir desaturates instead of shifting. - runtime/srcdoc: new injectPaletteBridge + paletteBridge / initialPalette options - file-viewer-render-mode: paletteActive flips URL-load back to srcDoc so the bridge can be injected - FileViewer: state, popover, postMessage wiring, srcDoc + useUrlLoadPreview integration - PaletteTweaks: popover UI with Original + Coral / Electric / Acid forest / Risograph / Mono noir - PreviewDrawOverlay: stub pass-through until the draw branch lands * feat(web): hide finalize-design toolbar from project header * test(e2e): skip project actions toolbar flow after toolbar removal |
||
nettee
|
03da01a56f
|
ci: use open-design bot for contributors wall refresh (#1349) | ||
Nicholas-Xiong
|
c0b679ecbc
|
fix: restore custom dropdown chevron for timezone selector in dark mode (#1368)
Fixes #1359 The timezone selector in the Routines form was showing repeated dropdown icons and poor text readability in dark mode because: 1. set to remove the native chevron, but didn't restore a custom one via background-image 2. Missing caused text to overlap with any chevron 3. No dark-mode-specific chevron color was defined This commit adds the custom dropdown chevron styling (matching the global select behavior) with proper padding and dark-mode color variants, ensuring: - Single, correctly-positioned chevron icon - Sufficient padding to prevent text overlap - Proper contrast in both light and dark themes - Consistent visual behavior with other form controls |
||
Sid
|
fb47d0ae51
|
style(web): polish EntryView UI — sidebar layout, folder tabs, slim form, blue selected token (#1360)
* chore(web): upgrade radius scale + introduce blue --selected token
UI polish pass — design tokens for follow-up commits.
Radius scale was visually too square at the small end. Bump up so
buttons / inputs / cards feel rounded rather than boxy:
- `--radius-sm: 6px → 8px` (buttons, inputs, small chips)
- `--radius: 10px → 12px` (medium containers, Recent filter pill)
- `--radius-lg: 14px → 16px` (project cards)
- `--radius-pill: 999px` unchanged (status chips)
Introduce a separate "selected" colour so selection indicators
(card borders, focus rings) read as blue instead of fighting with
the orange brand accent that drives primary CTAs:
- `--selected: #2563eb` (Tailwind blue-600)
- `--selected-soft: rgba(37, 99, 235, 0.16)` (soft tint for shadows)
No selectors are migrated to `--selected` in this commit — that
happens in a later "selected state" commit so the diff stays scoped.
* refactor(web): replace entry global header with sidebar brand + reorder bottom chips
Pre-existing layout: a global \`AppChromeHeader\` strip sat across the
whole top of EntryView (logo + settings gear), then a 2-column body
below it. Visual mass concentrated in a thin horizontal bar that did
not relate to the page's column structure, and the settings gear
duplicated the bottom Local-CLI chip.
New layout matches the two-column "brand-in-sidebar + tabs-in-main"
pattern: the brand block lives at the top of \`.entry-side\` (left
column), the right tabs live at the top of \`.entry-main\`, and the
vertical divider between them is the only horizontal seam.
EntryView:
- Drop \`<AppChromeHeader actions={avatarMenu} />\` from EntryView's
render — the home page no longer renders the global chrome strip.
(ProjectView still uses AppChromeHeader for back-nav / file
actions, so the component itself stays in the codebase.)
- Add a sidebar brand block inside \`.entry-side\` using the
already-defined \`.entry-brand\` / \`.entry-brand-mark\` /
\`.entry-brand-title\` classes that were sitting dead in
index.css.
- Reorder \`.entry-side-foot\` chips so that the env-critical
Local CLI row sits on top of the row, with the secondary
toggles (language picker, pet adoption, X follow icon) compact
on a second row. The Follow @nexudotio chip drops its text
label and becomes icon-only — pure marketing content, so it no
longer earns a full-width pill.
- Settings access moves entirely to the Local CLI chip's existing
click handler; the top-right gear is gone (it was a duplicate).
CSS:
- \`.entry-shell\` grid: \`auto 1fr\` → \`1fr\` (no header row).
- \`.entry-side\` background: \`var(--bg-panel)\` → \`transparent\`,
so the sidebar shares the page beige and only the New-prototype
card reads as white. Removes the "everything on the left is on
one big white sheet" feeling.
- \`.entry-brand\` gets \`padding: 24px 20px 18px\` so the logo +
title block has breathing room at the top of the sidebar.
- \`.entry-brand-mark\` width/height \`44 → 34\`. The previous
44px gradient ring was visually heavier than the title text it
sat next to.
- \`.entry-brand-title\` weight \`600 → 450\`, color
\`var(--text-strong)\` → \`var(--text)\`. Serif title still
reads as the page anchor without the chunky "bold black" stamp.
- \`.entry-brand-actions\` added for future right-aligned actions
(carries no actual content in this commit — kept so re-adding a
settings/avatar entry point doesn't need new CSS).
- \`.entry-side-foot .foot-pill\` slim pass: padding
\`4px 10px → 3px 8px\`, font \`11.5px → 10.5px\`, gap \`6 → 5\`,
plus \`justify-content: center\` and \`min-height: 24px\` so the
icon-only Follow pill stays the same height as the text pills
next to it.
* style(web): align right tabs row with brand row + strip hover/focus noise
Right column's tabs row ("Designs / Templates / Design systems /
Image templates / Video templates") needed three things:
1. Vertical center of tab text aligned with the brand logo on the
left (both rows feel like one row, separated by the vertical
divider only).
2. Active tab's underline sitting flush on the horizontal divider
below the tabs (not floating mid-row).
3. No hover background, no focus outline, no transition — tabs are
a navigation strip, not action buttons.
Changes:
- `.entry-header` padding `0 28px` → `24px 28px 0`, drop the
`min-height: 52px`. Padding-top mirrors the brand block's
padding-top (24px) so left logo top and right tabs top land on
the same Y. Header height now content-driven; underline meets
the `border-bottom` divider naturally.
- `.entry-tabs` gets `align-self: stretch` + `align-items: center`
+ `gap: 2px → 24px`. The stretch lets the tabs container fill
header height; the bigger gap matches Claude Design's tab
rhythm.
- `.entry-tab` becomes a "plain underline tab":
- `border-radius: 6px 6px 0 0 → 0` (no folder-tab look — that's
on the left tabs).
- `padding: 14px 11px → 6px 4px 8px` so text + underline form a
tight group, with the underline sitting at the bottom of the
tab box right above the header divider.
- `font-size: 14px → 12px` matches the left newproj tabs (set
in commit 4) — both columns share the same tab type-size.
- `transition: none` removes the inherited 120ms background /
border / color transition.
- Hover / focus / active states explicitly zero out background,
border-color, outline. Hover keeps a subtle color change
(`text-muted → text`) so the tab still feels interactive
without flashing a chip behind it.
- Active state colors are duplicated across `.active`,
`.active:hover`, `.active:focus`, `.active:focus-visible` so the
black underline never gets overwritten by the inactive-state
rules above.
* style(web): folder-tab merge on left newproj tabs + flat card top corners
The left "Prototype / Live artifact / Slide deck / …" tabs sat as
plain underline tabs above a fully-rounded card. The active tab and
card looked like two stacked rectangles with a gap.
Folder-tab pattern:
- Active tab gets a white background + 12px top corners + a 1px
border on top / left / right.
- Active tab's bottom border matches the card's background color
(effectively invisible) — so where the tab sits, the card's top
border is "broken" and tab + card read as one merged shape.
- Card top corners are square (`border-radius: 0 0 12px 12px`),
bottom corners stay 12px. With the active tab's square bottom
edge, the merge line at the tab/card seam is a clean horizontal,
not a curve mismatch.
Implementation:
- `.newproj-tabs-shell`:
- `overflow: hidden → visible` so the tab's overlap with the
card below isn't clipped at the shell's bottom edge.
- `margin-bottom: -1px` + `z-index: 2` so the shell renders on
top of the card and the 1px tab/card overlap actually paints.
- The `.can-left { padding-left: 40 }` / `.can-right` overrides
used to reserve room for scroll arrows are removed (arrows
are hidden, no extra padding needed).
- `.newproj-tabs` keeps its horizontal `overflow-x: auto` so the
8 project-type tabs can still scroll inside the sidebar width.
- `.newproj-tabs-arrow` becomes `display: none`. The two
chevron-circle buttons added clutter without much benefit —
users with touchpads / wheels / keyboard already scroll the
tabs row natively, and the `::before` / `::after` linear-
gradient fades (now using `--bg` instead of `--bg-panel` so
they fade into the page beige, not the sidebar panel that no
longer exists) signal there are more tabs to the right.
- `.newproj-tab`:
- Replace the plain bottom-underline (`border-bottom: 2px
solid transparent`) with a full transparent 1px border so
the active state can flip just the colors without changing
layout.
- `border-radius: 0 → 12px 12px 0 0`.
- `position: relative` for z-index stacking.
- `padding: 10px 6px → 7px 14px` (less vertical, more
horizontal — tabs read as "labels" rather than chunky
buttons).
- Symmetric top/bottom padding (`7px`) so the text + folder-
tab top corners stack cleanly.
- `transition: none` — no animation between active/inactive
states (tabs are nav, not action buttons).
- All hover / focus / focus-visible / active states zeroed out
background and border-color so the inherited `button { … }`
base style (which adds bg-subtle on hover) does not bleed in.
Subtle color change on hover (`text-muted → text`) is the
only affordance.
- `.newproj-tab.active` (+ active hover/focus combos so the
base rules don't override): white bg, full var(--border) on
three sides, bottom border = var(--bg-panel) (invisible
against card), z-index 3 (above non-active tabs and shell
pseudo-elements).
- `.newproj-body`:
- `margin: 0 24px` so the card breathes inside the sidebar
(and the active tab's left edge aligns with the card's left
edge).
- `padding: 18px 24px 28px → 16px 18px 18px` — tighter.
- `border-radius: (full 12) → 0 0 12px 12px` for the
flat-top merge with the active tab.
- Adds explicit `border` + `background: var(--bg-panel)` +
`box-shadow: var(--shadow-xs)` so the form reads as a card
floating on the transparent sidebar.
- `flex: 1 → 0 0 auto` (and `min-height: 0` / `overflow-y:
auto` removed) — the card is content-sized, not stretched
to fill the sidebar. Empty space below the card is now
page beige, not a giant white sheet.
- `gap: 14px → 12px` between form sections.
* style(web): slim NewProjectPanel form (title, fidelity, buttons, ds-picker)
The form inside the new white card felt overweight against the
compacted layout from the previous commits — fidelity cards were
~133px tall, the Create button + Open-folder secondary button both
had ~11px symmetric padding, the design-system trigger had a 32px
avatar in a 55px-tall row. Slim every element so the card reads as a
focused form, not a stack of beefy buttons.
Title:
- \`.newproj-title\` font \`14px / 600 → 13px / 550\`. Still
visibly the section heading but no longer competing with the
serif brand title above.
Fidelity:
- \`.fidelity-thumb { aspect-ratio: 12/7 → 16/7 }\`. The previous
aspect made cards taller than they needed to be in the narrow
sidebar column.
- \`.fidelity-card { gap: 8 → 6, padding: 10/10/12 → 8/8/10 }\`.
Combined with the thumb aspect change, card height drops from
~133px → ~102px (visually close to the Claude Design reference
while keeping the same content).
Primary / secondary buttons:
- \`.newproj-create\` padding \`11px (symmetric) → 8px 11px\`,
margin-top \`4 → 2\` — primary CTA no longer towers over the
fidelity cards above it.
- \`.newproj-import\` padding \`10px → 6px 10px\` — the secondary
"Import Claude Design ZIP" button feels like an alt option, not
a peer of Create.
Design system trigger:
- \`.ds-picker-trigger\` gap \`10 → 8\`, padding \`8/10 → 6/10\`.
- \`.ds-picker-title\` font \`13 → 12.5\` so name + subtitle stay
legible in the slimmer row without overflowing the column.
- \`.ds-avatar\` width/height \`32 → 26\`, border-radius \`6 → 5\`.
The thumbnail was the dominant element in the row; shrinking it
pulls the row height from ~55px → ~50px.
Footer disclaimer:
- \`.newproj-footer\` padding-top \`0 → 12px\`. The "Only you can
see your project by default." line was butting against the card
bottom; 12px of air separates the disclaimer (page-bg context)
from the card (panel-bg context) cleanly.
* style(web): blue selected indicators + Recent filter rounded + neutral input focus
Three small "selection state" tweaks driven by the new
\`--selected\` token introduced earlier in this branch:
1. Fidelity card selected border is now blue, not the brand
accent. The orange Create button + the orange selected card
border were fighting for the same visual role (primary
action vs primary selection). Blue clearly says "this is
the one that is selected" without competing with the CTA.
- \`.fidelity-card.active\` border-color
\`var(--accent) → var(--selected)\`.
- Box-shadow ring + soft 0.04 drop swapped from the orange
\`180/90/59\` rgba tuple to the blue \`37/99/235\` tuple.
- \`.fidelity-card.active .fidelity-thumb\` border swapped
from \`var(--accent-soft) → var(--selected-soft)\`.
2. Recent / Your designs filter is no longer a fully-rounded
pill. The bottom-left settings chips deserve to be the only
"999px pill" shape — those are tertiary status indicators.
The Recent/Your designs toggle is a higher-importance
inline filter, so it gets the medium radius instead.
- \`.subtab-pill\` wrapper border-radius
\`var(--radius-pill) → var(--radius)\` (12px).
- Inner button border-radius
\`var(--radius-pill) → var(--radius-sm)\` (8px).
- Active state background \`var(--text) → var(--bg-panel)\`,
color \`var(--bg) → var(--text)\`. The "black filled pill"
read as a status badge; white-on-faint-gray reads as
"selected toggle" — same shape as Claude Design's Recent
pill.
3. Input focus is neutralised. The base \`input:focus\` rule
added an orange border + a 3px orange-soft ring around the
focused field — way too much visual weight for a quiet form
("Project name" → focus made it scream).
- \`input:focus / textarea:focus / select:focus\` border-color
\`var(--accent) → var(--border-strong)\` (light grey).
- Box-shadow ring removed (\`none\`). Focused inputs now only
darken their border by one step — barely visible but enough
to confirm focus.
These three changes are grouped because they all migrate selection-
state styling off the brand accent and onto neutral / blue tokens.
The next pass (if any) can sweep the remaining \`var(--accent)\`
selection sites (\`.ds-row.active\`, \`.ds-picker-trigger.open\`,
\`.conv-pill.open\`, …) to use \`--selected\` too, but each of those
lives in a different surface and felt out of scope for the entry
view polish.
* refactor(web): pet rail toggle moves inside pet pill as split button
WHAT
- Convert the pet pill from a single `<button>` to a `<div>` containing
two buttons separated by a 1px divider:
* `.pet-pill-main` keeps the existing "Adopt a pet" / "Change pet"
glyph + label + unadopted dot, still wired to `onAdoptPet`.
* `.pet-pill-toggle` is a small icon-only button that flips
`petRailHidden` — eye icon when the rail is hidden ("click to show"),
eye-off when visible ("click to hide").
- Drop the old avatar-menu popover from EntryView entirely:
`avatarMenuOpen` state, the outside-click / Escape effect, and the
cog-popover trigger are all removed. The `Settings` entry of that
popover was already redundant with the `Local CLI` chip; the
`Hide/Show pet picker` entry now lives directly on the pet pill.
- CSS in the `.pet-pill` block:
* `height: 24px` + `padding: 0` so the outer pill matches every other
chip in the row vertically.
* `.pet-pill-glyph` reduced from 14px to 12px and constrained to a
14x14 inline-flex box so the unicorn / paw glyph stops pushing the
chip taller than 24px.
* Per-region hover (`.pet-pill-main:hover`, `.pet-pill-toggle:hover`)
so each side of the split lights up independently, with the divider
inheriting the accent tint while the chip is in `pet-pill-fresh`.
WHY
- After commit
|
||
huyhoangnhh98
|
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> |
||
|
PerishFire
|
93865f71e7
|
fix(daemon): remove opencode stdin dash sentinel (#1365) | ||
Krishna shakula
|
1ce7d6e8c5
|
fix: use ACP config options for model selection (#1208) | ||
ashleyashli
|
a4649dacb3
|
fix: check contributor tiers on review and comment events (#1248)
* fix: check contributor tiers on review and comment events Expand the contributor card workflow to run tier checks for PR reviews, issue comments, PR review comments, and discussion activity. The bot now understands pull_request_target directly, so remove the event-name shim. Co-authored-by: Cursor <cursoragent@cursor.com> * fix: drop fork-unsafe triggers (review, issue_comment, review_comment) Per @mrcfps and @chatgpt-codex-connector review: GitHub withholds repository secrets on pull_request_review, pull_request_review_comment, and issue_comment events when they originate on forked PRs, so wiring those events here would fail-closed exactly for external contributors. Keep the fork-safe triggers (pull_request_target.closed, issues.opened, discussion.*, discussion_comment.*) and document why the three are excluded. They can be re-added later via a workflow_run handoff. --------- Co-authored-by: ashley li <ashleyli@ashleydeMacBook-Air-2.local> Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: qiongyu1999 <qiongyu1999@gmail.com> |
||
|
lefarcen
|
43f7fc536a
|
Add Langfuse telemetry relay (#1296)
* Add Langfuse telemetry relay * Configure telemetry worker custom domain * Add telemetry relay health check * Harden telemetry relay config |
||
|
nettee
|
71b4a331ab
|
spec(web): Token first tailwind (#1201) | ||
Nagendhra Madishetti
|
dbc94b83ed
|
feat(web): add thumbs-up/down feedback widget under completed assistant turns (#1288) (#1308)
* feat(web): add thumbs-up/down feedback widget under completed assistant turns (#1288) Adds a lightweight feedback widget that surfaces under each assistant turn whose run succeeded. Users can submit positive or negative feedback in one click; the negative path opens an optional free-text comment area. The widget never blocks the message composer and only mounts after the run has produced its final artifact, matching the acceptance criteria. What ships - `<MessageFeedback>` (apps/web/src/components/MessageFeedback.tsx) renders the three states: idle (prompt + thumbs), submitted positive (confirmation + Change), submitted negative (confirmation + optional comment textarea + Send + Change). - AssistantMessage.tsx slots the widget under AssistantFooter, gated on `runSucceeded && !hasEmptyResponse`, so failed and empty-response turns don't ask the user to rate something that never finished. - The full record shape leaves room for the future analytics metadata the issue calls out (rating, comment, submittedAt; artifactRef / runId derivable from the surrounding message whenever the analytics pipeline lands). Persistence (v1 = localStorage) Lefarcen's clarifying comment on the issue asked whether v1 should be daemon-persisted or in-memory while the analytics pipeline is defined. The daemon's messages table is column-strict, so daemon persistence would require a SQLite migration plus a contract bump on `ChatMessage`; locking that shape in before the analytics pipeline is designed risks reworking it twice. localStorage is the middle ground: feedback survives reload (so the "feedback state is visually clear after submission" criterion holds across tabs and sessions) without committing the wire shape. The hook surface is just `(value, setter)`, so a future PR can swap the storage layer for a daemon mirror or an analytics shipper without touching the React surface. The store handles corrupted JSON, unknown future rating values, disabled storage (private-mode browsers), and broadcasts changes across listeners in the same tab via a CustomEvent so two mounts of the hook for the same messageId stay in sync. i18n 11 new keys under `feedback.*` (prompt, thumbsUp/Down, two confirmation chips, comment label/placeholder/submit/saved, change). English source values authored alongside the keys; zh-CN translations added in the same pass so the locale alignment test stays green and Chinese users see Chinese strings from day one. The other 16 locales pick up English fallbacks via their existing `...en` spread. Test coverage - `tests/state/message-feedback.test.ts` (8 jsdom cases) — round-trip, null-clear, corrupted JSON, missing rating, unknown rating, key collision across messages. - `tests/components/MessageFeedback.test.tsx` (7 jsdom cases) — idle state, positive submit, negative submit, comment save, blank-comment Send disabled, Change unsticks the rating, rehydration from pre-populated storage. The locale alignment test continues to enforce that every locale declares the new keys (5/5 across 18 locales). Validated - pnpm guard clean - pnpm --filter @open-design/web typecheck clean - tests/i18n/locales.test.ts 5/5 - tests/state/message-feedback.test.ts 8/8 - tests/components/MessageFeedback.test.tsx 7/7 - Full web suite: 98 files, 903 tests * fix(web): tighten feedback widget gate + storage sync + textarea, add styles (PR #1308 review) Addresses every P2/P3 from the codex + Siri-Ray + lefarcen reviews on PR #1308, plus a couple of polish items the review surfaced indirectly. Visibility gate (lefarcen P2) The gate was `runSucceeded && !hasEmptyResponse`, which also matched text-only acknowledgements and question-form replies. The issue scopes feedback to turns that produced a final artifact, so the gate now also requires `produced.length > 0`. New AssistantMessage suite (5 jsdom cases) pins: artifact -> shown, no-artifact -> hidden, streaming -> hidden, failed run -> hidden, empty_response -> hidden. Storage sync (codex P2 + lefarcen P2) The previous broadcast contract was: write storage, dispatch a bare CustomEvent, listeners re-read storage. That had two failure modes: - setItem throwing (private mode / quota / disabled storage) left the listener seeing null and clobbering the in-memory state the user just confirmed. - The clear path early-returned after removeItem and never dispatched, so a second mount of the same messageId stayed in the submitted state when the user clicked Change. New contract: every successful OR failed write dispatches a CustomEvent whose `detail.value` carries the new feedback record (or null). Listeners apply the value directly without re-reading. Same- tab sync survives storage failures and the clear path no longer early-returns. Cross-tab still re-reads on the platform `storage` event since that event has no detail. Two new storage tests pin the new broadcast contract (positive + null) and the failed-setItem path; two new component tests pin in-session confirmation under setItem failure and two-mount Submit + Change synchrony. Textarea draft fix (lefarcen P3) The textarea used `draftComment || feedback.comment || ''` as its controlled value, so erasing a saved comment snapped it back. The draft is now exclusively the source of truth; a ref-backed effect re-seeds the draft from feedback.comment whenever the rating transitions (mount, idle -> negative, cross-mount sync). Send is now enabled when `draftComment !== savedComment`, which lets the user both edit and clear a saved comment. New component test pins erase+ Send actually removing a previously-saved comment. Accessibility The confirmation chip and "Comment saved" tag both gain `role="status"` + `aria-live="polite"` so screen readers announce the state transition. The thumb buttons keep their `aria-label`. CSS (lefarcen P3) The widget's `.message-feedback*` class set had no rules in index.css, so it rendered with default browser controls. Added a ~130-line block that mirrors the surrounding chat pill/chip vocabulary: bg-subtle background, border-pill confirmation chip, accent-tinted positive state and amber-tinted negative state to match the assistant-footer's data-unfinished pattern. Comment area sits below the chip and wraps on narrow widths so the composer isn't pushed off-screen on small panes. Validated - pnpm guard clean - pnpm --filter @open-design/web typecheck clean - tests/state/message-feedback.test.ts 10/10 (was 8, +2 broadcast) - tests/components/MessageFeedback.test.tsx 10/10 (was 7, +3 sync / storage-failure / clear-saved-comment) - tests/components/AssistantMessage.test.tsx 5/5 (new file) - tests/i18n/locales.test.ts 5/5 - Full web suite: 866 tests --------- Co-authored-by: Nagendhra <nagendhra405@gmail.com> |
||
github-actions[bot]
|
c9d3358af4
|
docs(readme): refresh contributors wall (#1330)
Co-authored-by: mrcfps <23410977+mrcfps@users.noreply.github.com> |
||
|
Nagendhra Madishetti
|
1df3eca161
|
feat(web): Critique Theater Phase 7 — reducer + useCritiqueStream + useCritiqueReplay (#1307)
* 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.
---------
Co-authored-by: Nagendhra <nagendhra405@gmail.com>
|
||
|
Nagendhra Madishetti
|
64510b790b
|
fix(web): translate Design Files refresh strings instead of hardcoding English (#1254) (#1300)
* fix(web): translate Design Files / live artifact refresh strings instead of hardcoding English When the app language was set to Chinese, the Design Files refresh flow showed Chinese for the surrounding chrome but kept English for every label and message originating in describeRefreshStatus, describeEventPhase, and the refresh-event timeline body of LiveArtifactRefreshHistoryPanel. Same-screen mixed-language UX, the exact symptom reported in #1254. Root cause: those three sites bypassed i18n entirely. describeRefreshStatus returned hardcoded English label + description strings for the running / succeeded / failed / idle / never statuses; describeEventPhase returned hardcoded Started / Succeeded / Failed labels; the timeline body inlined "Refresh started…", "<n> source(s) updated", and "Refresh failed." string literals; and the empty-timeline copy ("No refresh activity yet in this session. Trigger Refresh to record a timeline…") was hardcoded too. Fix: thread the existing TranslateFn through both helpers, swap every hardcoded string for a t() lookup, and pull the empty-timeline copy and the failure-fallback through the same path. Added 13 new keys under liveArtifact.refresh.* — statusRunning, the five *Description keys, three event-phase labels (eventStarted/Succeeded/Failed), eventStartedDetail, sourcesUpdatedOne/Many with an {n} placeholder, and timelineEmpty. Status labels for succeeded / failed / ready / never already had keys (statusSucceeded / statusFailed / statusReady / statusNever) so those are reused unchanged. Locales: full Chinese translations added to zh-CN.ts (the locale directly named in the issue). The other 16 locales pick up English fallbacks through their existing ...en spread, so the locale-key alignment test stays green; native translations for those locales can land via the usual locale-team passes without re-touching the source code. * fix(web): cover the rest of the refresh panel under i18n + add a zh-CN render test Lefarcen's review on #1254 / PR #1300 surfaced that the first pass only translated three helpers (describeRefreshStatus, describeEventPhase, session timeline body) and left the rest of the panel in English. Under a Chinese UI the panel still mixed languages, which was exactly the regression the issue was filed for. This commit threads t() through every user-visible refresh-panel string the user would see in the Chinese flow: - Hero block: "Last refreshed" label + "Never" empty state. - Created / Last updated facts + their "Unknown" empty label. - Persisted refresh history header, hint, empty-state copy. - Persisted timeline status badge: succeeded / running / failed / cancelled / skipped now resolve through describePersistedStatus, which uses an exhaustive switch off LiveArtifactRefreshLogEntry's status union so a future contract addition trips tsc. - Session activity header, hint. - Document source header, hint, Type / Tool / Connector field labels. - Advanced debug metadata summary + note line. - "just now" relative-time fallback in the persisted timeline. 22 new i18n keys total (23 with the new heroLastRefreshedNever distinct from statusNever); zh-CN strings authored alongside the English source, every other locale picks them up via its existing ...en spread and the locale-key alignment test stays green. Intentionally untranslated surfaces: raw daemon payloads inside the <details> debug panel (event.step / refreshId / error.message and the JSON.stringify dump), since those are agent / connector identifiers and stack-trace style strings, not localised copy. The debug summary heading itself is translated; if the debug section should be hidden in localised primary flows, that is a separate UX call worth its own issue. Test coverage: new render test wraps LiveArtifactRefreshHistoryPanel in I18nProvider initial="zh-CN" and pins the Chinese rendering of every translated label, plus negative assertions that the formerly hardcoded English literals are NOT present in the markup. With the no-provider fallback returning English, the existing static-markup tests can't observe the regression this PR is meant to fix; the zh-CN render test is the only one that would have caught the original gap and will catch the next one. Validated: pnpm guard, pnpm --filter @open-design/web typecheck, locales.test.ts (5/5), FileViewer.test.tsx (69/69, +1 new zh-CN test), full web suite (92 files, 841 tests). * fix(web): route formatRelativeTime through Intl.RelativeTimeFormat so units localise Lefarcen's second pass on PR #1300 caught the remaining hardcoded English path: formatRelativeTime() still emitted units like `5s ago` and `45m ago`, so Chinese users would see those strings inside the otherwise-translated refresh panel. The function now takes the active locale + TranslateFn and routes through Intl.RelativeTimeFormat with style: 'narrow', numeric: 'always'. That preserves the historical `5s ago` shape for English while producing locale-correct output for every other locale (zh-CN gets `5秒前` / `45分前`, with the right past / future suffix and word order). The `just now` carve-out (abs < 5s) keeps using t('liveArtifact.refresh.justNow') since Intl's narrow output for zero-delta reads awkwardly. A try/catch around the RTF constructor falls back to 'en' if the runtime rejects the locale, so the function is safe on engines with limited ICU data. Callsites threaded through: - LiveArtifactRefreshHistoryPanel hero metric (`lastRefreshedAt`) - Session timeline event row (`event.startedAt`) - Session timeline event time (`event.at`) - LiveArtifactRefreshFact for the created / last-updated facts; the component now accepts optional `locale` + `t` props and the panel passes them in. Test coverage extension: - The existing zh-CN render test sets a real lastRefreshedAt (now - 45s) and real session-event timestamps, then asserts the Chinese past-tense suffix `前` appears AND the legacy English `Xs ago` / `Xm ago` shapes do NOT. That was the gap lefarcen pointed at: setting `lastRefreshedAt: undefined` couldn't see the regression because no relative-time formatting ran. - Added a small second test for the lastRefreshedAt-undefined empty hero so the original `从未` coverage still pins. Validated: pnpm guard, pnpm --filter @open-design/web typecheck, FileViewer.test.tsx (70/70, +1 new test), locales.test.ts (5/5), full web suite (92 files, 842 tests). --------- Co-authored-by: Nagendhra <nagendhra405@gmail.com> |
||
|
github-actions[bot]
|
5fa861137d
|
Update docs/assets/github-metrics.svg - [Skip GitHub Action] (#1328)
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> |
||
|
PerishFire
|
819c34fd8f
|
fix(tools-pr): fall back on reviewDecision for unresolved-changes-requested (#1287)
* fix(tools-pr): fall back on reviewDecision for unresolved-changes-requested Patrol classify on the live 102-PR queue missed three PRs (#1101, #1127, #1163) where GitHub's reviewDecision is CHANGES_REQUESTED but the classify tag did not fire. Root cause is a divergence between two notions of "latest review state per reviewer": - GitHub's reviewDecision keeps a reviewer's CHANGES_REQUESTED in effect until that same reviewer submits APPROVED or DISMISSED. A subsequent COMMENTED review by the same reviewer does NOT supersede it. - Our `reduceLatestReviewsByAuthor` collapses every reviewer to their latest review with no special-casing of state, so a CHANGES_REQUESTED followed by COMMENTED disappears from the reduced view. `tagUnresolvedChangesRequested` filtered the reduced view for `state === "CHANGES_REQUESTED"`, so the three PRs above (each had a reviewer write CHANGES_REQUESTED → COMMENTED) escaped the rule even though the PR-level reviewDecision was still CHANGES_REQUESTED. Add a narrow fallback: when the first path returns no per-reviewer reviewers, trust `facts.reviewDecision === "CHANGES_REQUESTED"` as the source of truth. The fallback reason and source token differ from the first path so report consumers can tell which signal fired. Reducer semantics left alone on purpose — flipping COMMENTED handling there would cascade to `bot-only-approval`, `stale-approval`, and `humanReviewerSignalAt`, each of which has its own correctness story. * fix(tools-pr): keep fallback reason strictly factual Codex flagged that the fallback path's reason text asserted a specific review sequence ("CHANGES_REQUESTED then COMMENTED") that the condition alone does not prove. The condition only observes: - `facts.reviewDecision === "CHANGES_REQUESTED"`, and - after `reduceLatestReviewsByAuthor`, no review carries `state === "CHANGES_REQUESTED"`. Multiple GitHub configurations satisfy that pair — a reviewer's CR followed by COMMENTED, a CR that sits outside the `reviews(last: 30)` fetch window, etc. Per `tools/pr/AGENTS.md`'s strictly-factual rule, the reason must report only what is directly observed, not the most likely upstream cause. Drop the inferred-cause clause from `reason`; move the explanation of possible upstream causes into the code comment above the branch where it does not show up in classify output. * docs(tools-pr): document fallback data source for unresolved-changes-requested Siri-Ray and lefarcen both flagged that the tag dictionary row for `unresolved-changes-requested` only describes the primary per-reviewer path. The fallback added earlier in this PR emits the same tag with a different `source` token (`gh.reviewDecision` vs the original `gh.latestReviews[].state`), so report consumers need the dictionary to list both paths to interpret which one fired. Update the row to call out both: the primary per-reviewer rule, and the PR-level reviewDecision fallback that fires when no per-reviewer CR survives the latest-per-author reduction. The two-token source column mirrors the actual `Tag.source` strings emitted at runtime. * test(tools-pr): pin both emission paths of unresolved-changes-requested lefarcen flagged the fallback was validated only by live-PR examples in the PR body, so a refactor could silently regress the coverage. Add a deterministic test file `tests/tags-unresolved-cr.test.ts` that exercises `classifyPr` against crafted `PrFacts` fixtures: - primary path (per-reviewer CR after reduction) fires with source=gh.latestReviews[].state and surfaces the reviewer login - fallback path fires with source=gh.reviewDecision when no per-reviewer CR survives reduction (covers both the COMMENTED-follow-up shape and the empty-reviews shape — the latter pins the `reviews(last: 30)` out-of-window concern from the factual-reason fix) - primary wins over fallback when both signals are present (single tag emitted, source=gh.latestReviews[].state) - two negative cases: empty reviewDecision and APPROVED — neither emits Also extend the fallback's code comment with the observed scale (3 of 102 open PRs hit this gap: #1101, #1127, #1163) so future maintainers can tell this is a recurring queue pattern, not a theoretical edge case. This is the first test under `tools/pr/tests/`; the package test script already ran `node --import tsx --test tests/*.test.ts` against an empty glob, so no scaffolding changes are needed. |
||
Caprika
|
5bd9763181
|
[codex] Improve Claude Code exit diagnostics (#1267)
* fix daemon claude diagnostics * fix claude custom endpoint auth diagnostics * fix project view api empty response test props * fix claude diagnostic review gaps * fix silent custom endpoint claude diagnostics * fix claude diagnostic credential redaction * fix quoted api key redaction * fix claude diagnostic tail redaction * fix silent claude configured profile diagnostics |
||
chaoxiaoche
|
a75d9938c7
|
feat(design-systems): add structured tokens.css schema (default + kami) (#1231)
* feat(design-systems): add structured tokens.css schema (default + kami) Compile each brand's DESIGN.md prose into a machine-readable :root block agents paste verbatim, removing the "Primary → --accent" translation step where most token misuse happens. Daemon prompt injection lands in a follow-up; lint-artifact already enforces the shared token vocabulary so no rule changes needed. Schema validated across two contrasting aesthetics: - default (sans-serif, cobalt, B2B utility) — stress test the shallow form, 2-level fg / 2-level surface - kami (serif, parchment, ink-blue, print-first) — stress test the rich form, 4-level fg ramp, 3-level surface, ring elevation, i18n font stacks, and solid-hex tag tints (print renderers double-paint alpha) Schema growth from kami's stress test (5 new optional slots, all backward-compatible — default aliases via var() to existing tokens): - --fg-2 / --meta (4-level fg ramp) - --surface-warm (3-level surface) - --border-soft (2-level border) - --elev-ring (ring elevation as first-class level) Brand-specific extensions live in tokens.css with explicit "NOT in shared schema" labels and a documented promotion path (≥2 brands need it → promote to schema slot). components.html in each brand is a self-contained reference fixture that exercises every token through real layouts. Both fixtures lint clean against apps/daemon/src/lint-artifact.ts. Co-authored-by: Cursor <cursoragent@cursor.com> * feat(design-systems): add token-fixture drift guard Each design system in design-systems/<brand>/ ships two files agents consume in tandem: tokens.css (canonical token bindings) and components.html (a self-contained fixture whose first <style> embeds the same :root paste so the file renders standalone). The fixture's :root block is a copy of tokens.css's :root block, kept in sync only by an inline comment. This adds scripts/check-tokens-fixture-sync.ts and registers it in pnpm guard. The check pairs each brand's tokens.css with its components.html and asserts the unscoped :root block is byte-equivalent after canonical normalization (CSS comments stripped, whitespace collapsed, separator spacing normalized). Brands missing one half of the pair, or with no :root rule in either file, fail the guard. Scoped overrides like :root[lang="zh-CN"] are not required to appear in the fixture (per the kami fixture's inline comment they are pasted only when an artifact's <html lang> matches), so the check only compares the unscoped :root block. Verified: pnpm guard passes for default + kami, fails on intentional value drift, fails on missing token, tolerates whitespace-only formatting differences. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(design-systems): point fixture CTAs to real files Both default and kami components.html advertised in-page anchors (#tokens, #spec, #surface, #accent, #type, #components) but defined no matching ids, so every CTA was a no-op when the fixture was opened locally — flagged by mrcfps in #1231. Re-point each link to a real artifact in the same brand directory: - "View tokens" / "Inspect tokens" / "Inspect typography" → ./tokens.css - "Read the spec" / "Read the rule" → ./DESIGN.md Browsers render these as raw source views, which is the desired UX for a reference fixture: clicking the CTA shows the underlying contract instead of jumping to nothing. Agents copying the fixture also learn the pattern of "buttons link to actual sibling resources". The :root token block is unchanged, so the token-fixture drift guard still passes for both brands. Co-authored-by: Cursor <cursoragent@cursor.com> * feat(design-systems): codify token schema (A1/A2/B/C layers) The two-brand pilot (default + kami) settled the shape of the shared token schema; this commit codifies it as a machine-readable contract and enforces it in pnpm guard, addressing lefarcen's review on #1231: > the optional-vs-required split won't generalize cleanly when brand > #3 needs different Layer A tokens or when multiple brands converge > on the same extension (promoting C→B→A). Consider surfacing that > limitation in the PR narrative or in a future SCHEMA.md. Schema lives under design-systems/_schema/ as three files: - tokens.schema.ts — TypeScript declaration of every shared token with its layer (A1-identity / A1-structure / A2 / B-slot), plus per-brand C-extension allowlists and a global C-prefix allowlist - defaults.css — CSS mirror of A2 fallback values, used as the human-readable contract reviewer's-eye copy and the future input to the derive script - AGENTS.md — schema layer model, C → B-slot → A2 promotion rules, when-not-to-add-a-token guidance Layer model: A1-identity 8 tokens — bg/surface/fg/muted/border/accent + font-display/font-body. The brand IS these values; no fallback is defensible. A1-structure 18 tokens — type scale (8), leading (2), tracking (1), section-y (3), container (4). Structural decisions vary per brand by design and have no cross-brand default. A2 26 tokens — accent states, semantic colors, motion, base spacing scale, radius, elevation, focus, font-mono. Required in every tokens.css; fallback lives in defaults.css for the future derive script to inline when DESIGN.md does not specify the value. B-slot 4 tokens — fg-2 / meta / surface-warm / border-soft. Brand may bind independently or alias the named sibling via var(...) for components that target the richer ramp. C-extension n tokens — brand-specific names (kami's tag-bg-*, leading-display, accent-light, etc.). Allowlisted per-brand in BRAND_EXTENSIONS or globally by prefix in BRAND_EXTENSION_PREFIXES. Promote when a second brand adopts the same name. Why A2 fails the guard today: Artifacts are generated by agents pasting one brand's :root block into a single <style>; there is no global stylesheet that supplies fallbacks at runtime. A tokens.css missing an A2 declaration would silently break any var() reference in the fixture. Until the derive script (PR-B) lands and inlines defaults, every brand's tokens.css must declare every A2 token directly. The guard enforces this strictly. Why --font-mono lands in A2 (not A1): 149 brands' DESIGN.md files were surveyed: 87 (58%) declare a monospace stack, 62 (42%) do not — including major brands like bmw / nike / apple / notion / mastercard / meta. Agent paste cannot rely on the brand author having written it down; a defaultable A2 fallback (with CJK brands like kami overriding) is safer than forcing every brand author to add a field they may not realize their kbd / code-block components need. Five guard checks, each registered as its own entry in scripts/guard.ts so failures attribute to a specific contract: 1. token-fixture sync — components.html :root ↔ tokens.css :root byte-equivalent (existing) 2. A1 required tokens — every brand declares every A1 token 3. A2 required tokens — every brand declares every A2 token 4. unknown token allowlist — every declared token is in schema or brand-extension allowlist 5. A2 defaults parity — defaults.css ↔ tokens.schema.ts fallback byte-equivalent Verified on default + kami: - 26 A1 tokens declared in both brands - 26 A2 tokens declared in both brands - 129 total declarations, all match shared schema or brand extensions - defaults.css ↔ tokens.schema.ts parity holds - sanity test: drifting --motion-fast in defaults.css fails check 5 with a clear divergence message The PR description originally listed "Dedicated SCHEMA.md" as explicitly NOT in this PR ("Once 3+ brands ship, extracting a single source of truth becomes worthwhile"). That boundary moves: lefarcen's review surfaced the schema-generalization risk, and the schema must exist as a machine-enforced contract before the derive script can read it. The TS file replaces the markdown that was deferred. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(web/tests): pass missing designTemplates prop to ProjectView Pre-existing typecheck regression on main: PR #955 ( |
||
|
nettee
|
87a95b7fb4
|
Fix conversation run isolation (#1271) | ||
Kaelz31
|
3524a43d18
|
fix: pretty-print JSON file previews (#1206)
* fix: pretty-print JSON file previews * fix: avoid formatting JSON with unsafe numbers * fix: preserve precision-sensitive JSON previews * fix: preserve signed zero in JSON previews * fix: scan JSON numbers without repeated slicing --------- Co-authored-by: Kael S <YOUR_GITHUB_EMAIL_HERE> |
||
eggward han
|
a0316d2599
|
fix(web): suppress autosave indicator for draft-only Connector key edits (#1232)
When the user typed a replacement Composio API key, the global Settings autosave loop persisted `buildPersistedConfig(cfg)` — which intentionally strips the in-flight secret — and then advanced the indicator through 'saving' -> 'saved' despite the key never actually being written. The "All changes saved" status then contradicted the section-local "Save key" gesture and eroded trust in the saved-state badge for a sensitive field. The autosave effect now tracks the snapshot at the last successful save (or the initial cfg on mount) and compares the next snapshot's persisted shape against it via a new `isAutosaveDraftOnlyChange` helper. When the only diffs since last save are fields that `buildPersistedConfig` strips (today the Composio API key, generalizing to any future save-on-explicit-confirm secret), the persist call is skipped and the indicator settles to 'idle' instead of flashing 'saved'. The forced media-provider sync path still runs because that is a real outbound effect even when the persisted shape hasn't changed. Refs #1187 |
||
donglrd
|
19f1ff7995
|
Reject filesystem root folder imports (#1266) | ||
이용진
|
bbd14bd6fb
|
replace time-specific Orbit greetings with neutral defaults (#1291)
* replace time-specific Orbit greetings with neutral defaults Orbit default greeting is hard-coded to a morning-specific phrase and is not suitable as generic copy issue * fix skill trigger mistake 每日简报 -> 早安简报 |
||
|
PerishFire
|
c3d41c7d45
|
fix(tools-pr): chunk stats fetch through cursor-paginated GraphQL (#1285)
`fetchOpenPrs` was reading the stats chunk via `gh pr list --limit 1000 --json mergeStateStatus,...`. With the default limit raised to 1000 in #1259, this 502s reliably on the live open queue (107 PRs): GitHub's GraphQL gateway has to recompute mergeStateStatus for every PR up front, and the resulting query exceeds the gateway budget once the requested page passes ~60 PRs. Switch the stats chunk to `fetchPaginatedPrList`, the same cursor- paginated GraphQL helper that already drives reviews / comments / commits / assignment-timelines. Page size stays at PR_LIST_PAGE_SIZE (30), well within the gateway budget, and the heavy stats fetch is now consistent with the other heavy chunks. Verified locally: `pnpm tools-pr list` now completes against the live 107-PR queue without a 502. |
||
|
nettee
|
be77dc0394
|
Default English resource i18n fallback (#1270) | ||
Joey-nexu
|
12ac2e988e
|
docs: add Maintainer rules (MAINTAINERS.md + CONTRIBUTING entry-point) (#1290)
Adds a public set of rules for the External Maintainer role: who qualifies, how nominations work, what permissions Maintainers gain, what's expected of them, and how step-down works. The Core Team's individual roster is intentionally not enumerated. What's public is the rules everyone plays by. - New file: MAINTAINERS.md (English authoritative version) + 5 locale variants matching the existing CONTRIBUTING.md i18n surface (de, fr, ja-JP, pt-BR, zh-CN). Non-EN/non-zh-CN variants are machine-translated drafts marked at the top — native-speaker review is welcome via follow-up PRs. - CONTRIBUTING.md (and its 5 locale variants): adds a short Becoming a Maintainer section that points at MAINTAINERS.md, so the rules live in one place and translation drift is bounded. Decisions not in this PR (intentional): - No internal Core Team roster. - No internal observability dashboards. - No nomination PR / public voting flow (Core-Team-consensus-driven for now; to be revisited once External Maintainers exceed 5). Co-authored-by: Joey Li <lijinwei@open-design.ai> |
||
|
Caprika
|
fb079d8115
|
Add reliable agent-browser skill (#1284)
* Add reliable agent browser skill * Fix ProjectView delete conversation test props |
||
|
PerishFire
|
1eb20e3807
|
fix(web): keep tweaks selection usable without annotations (#1268) | ||
Sebastian Westberg
|
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
|
||
初晨
|
0f0d214298
|
fix(web): render static previews for sketch json files (#1060)
* fix(web): render static previews for sketch json files * fix(web): tolerate malformed sketch text items * fix(web): harden sketch preview parsing * fix(web): preserve sketch items on round-trip * fix(web): clear sketch files destructively * fix(web): unblock unsupported sketch saves |
||
Dongsen
|
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. |
||
|
Dongsen
|
12ce5ad38b
|
fix(web): ignore <artifact> tags inside markdown code spans and fences (#1132)
* test(web): add failing parser cases for <artifact> recitation in markdown code Cover the three real-world prose contexts where the model legitimately quotes the artifact tag without intending to emit one: - inside an inline backtick span - inside a fenced code block - spread across streaming chunks crossing the fence boundary Establishes the RED baseline before parser code-fence awareness lands. * fix(web): ignore <artifact> tags inside markdown code spans and fences The streaming artifact parser scanned the buffer with a raw indexOf, guarded only by 'next char must be whitespace'. That meant any literal <artifact ...> the model recited while documenting the protocol — even inside backticks or a ```html fence — flipped the parser into artifact mode, swallowed the rest of the reply from the chat UI, and (when a matching </artifact> appeared in the recitation) silently wrote a spurious file to disk via persistArtifact. Replace findOpenTag with a linear scan that tracks fenced code blocks (```) and inline code spans (`), skipping any <artifact prefix found inside either. If the buffer ends mid-fence, return a partial match anchored at the fence start so the next streaming chunk can resolve the boundary without losing fence context. Closes #1130. * fix(web): match renderer fence/inline-code rules in artifact parser Codex review on PR #1132 caught that the previous fix toggled inFence on any triple-backtick run anywhere in the buffer, including mid-line, while the chat renderer (apps/web/src/runtime/markdown.tsx) only treats ``` as a fence when it occupies a whole line matching /^[ ]{0,3}```(\w[\w+-]*)?\s*$/. That asymmetry would suppress a real <artifact> tag emitted after a prose sentence like "the opening marker is ```html and the response then writes:". Rework findOpenTag in three passes that mirror the renderer: 1. Walk \n-terminated lines; only a line that matches FENCE_LINE_RE toggles fence state. Open fences without a close (or with an unterminated tail line) return partial so the next chunk can resolve. 2. Collect inline code spans with /`[^`]+`/g — the same regex used by renderInline — so what the parser skips matches what the user sees as code. Unmatched trailing backticks after the last \n hold back. 3. Find the first <artifact …> outside any skip range; preserve the existing partial-prefix tail handling. Adds a regression test covering the exact case Codex reported. * test(web): pin parser behavior on double-backtick and in-fence string literal recitation Two cases raised in PR #1132 review: - a real artifact tag wrapped in '``<artifact …>``' (double-backtick inline code span) should not be treated as a real artifact - a fenced JS example whose body contains a string literal like 'const fence = "```";' should not pop fence state early and let a later literal <artifact> be parsed as real Both already pass on 96e88ca because the line-anchored fence regex and the renderer-aligned inline regex handle them correctly. Pinning the behavior so future regressions surface as test failures. * fix(web): make stripArtifact markdown-aware to stop truncating literal recitations The streaming artifact parser was hardened in 96e88ca to skip <artifact> recitations inside backticks and fences, but the post-stream stripper at AssistantMessage.tsx still ran a naive 'content.indexOf("<artifact")' over the same text events. As reported by lefarcen on PR #1132, that meant chat replies with literal protocol recitations could still get silently truncated mid-explanation — even though the parser preserved them in the text stream and the file panel was no longer polluted with ghost files. Extract the renderer-aligned classification (FENCE_LINE_RE, INLINE_CODE_RE, computeSkipRanges, rangeContains) into a single source of truth at apps/web/src/artifacts/markdown-context.ts so the parser and the stripper agree on what counts as code. Add apps/web/src/artifacts/strip.ts with a markdown-aware stripArtifact that: - ignores any <artifact open inside a fenced block or inline code span - looks for </artifact> with the same skip-range filter, so a real open paired with a literal close inside backticks does not strip a literal body that is meant to render - returns content unchanged when an open exists with no matching real close (the previous implementation sliced to end-of-string, which would nuke trailing prose on a malformed or still-streaming tag) Refactor parser.ts to import the shared helpers; behavior preserved (all seven existing parser tests still pass). New strip.test.ts covers six cases including the empirically-verified inline-backtick regression. * fix(web): align artifact stripper/parser fence rules with renderer exactly Two gaps surfaced in review at a0bf05f: - markdown-context.ts used a single FENCE_LINE_RE that allowed 0-3 leading spaces and reused the same pattern for opening and closing fences. The chat renderer (runtime/markdown.tsx:44 and :49) is asymmetric — opens with /^```(\w[\w+-]*)?\s*$/, closes with /^```\s*$/, and rejects any leading indentation on either side. Indented " ```html" was being treated as a code fence even though the renderer keeps it as a paragraph, and a literal "```html" line inside an open fenced example was closing the skip range early — both could expose a real or literal <artifact …> to the wrong handler. - stripArtifact discarded computeSkipRanges' unclosedFenceStart, so a fenced literal that ends at EOF without a trailing newline (very common for chat output) leaked the inner <artifact …> recitation to the stripper, reproducing the original #1130 truncation symptom on a narrower input shape. Split FENCE_LINE_RE into FENCE_OPEN_RE / FENCE_CLOSE_RE with no leading indentation, gate the fence state machine on the right side of the toggle, and have stripArtifact extend skip ranges to end-of-content when a fence is left open. Also tightened the parser's tail-line hold-back regex to match the renderer's no-leading-space rule. Added regression tests for the EOF-unclosed-fence case, the indented pseudo-fence (renderer treats as paragraph, stripper must strip the real artifact), and a "```html" line inside an open fence. Refs nexu-io/open-design#1130 * refactor(web): align streaming tail-line fence guard with FENCE_OPEN_RE The streaming parser's tail-line hold-back used a stricter local regex (/^```\w*$/) than the renderer's FENCE_OPEN_RE (/^```(\w[\w+-]*)?\s*$/), missing valid opener tails like ```c++, ```ts-, or ``` (trailing space). In practice these tails are still held back by the unmatched-backtick parity scan that runs immediately after — three backticks in a tail line are odd, so firstUnmatched stays set and the parser holds from that position. So this wasn't a runtime correctness bug, just a regex divergence that future readers could trip on. Drop the local regex and reuse FENCE_OPEN_RE so the tail check matches the same shape the rest of the pipeline already uses. Pinned the behavior with three new parser tests (`+`/`-` info-string suffix and trailing-space tails arriving as the first chunk) — they pass at HEAD, proving the parity scan was already covering these cases. Refs nexu-io/open-design#1132 (lefarcen polish P2) * fix(web): scope inline-code skip ranges per block and reject <artifact prefix-shared opens INLINE_CODE_RE previously ran over the whole buffer, so an unmatched backtick in one paragraph could pair with a backtick in a later paragraph and create a phantom inline span that swallowed any real <artifact …> between them. Mirror runtime/markdown.tsx by splitting the buffer on fence / blank / heading / list / hr boundaries and running INLINE_CODE_RE per block region instead. stripArtifact accepted any unskipped `<artifact` substring as a real open, while the streaming parser already required a following whitespace character — so prose like `<artifactual>demo</artifact>` was being truncated to `prefix suffix`. Extract the parser's real-open guard into isRealArtifactOpenAt and reuse it from both sides. While reordering findOpenTag for the shared guard, also fix the related hold-back ordering issue tracked at #1141: a stray tail-line backtick or fence-opener prefix used to suppress an artifact already complete earlier in the buffer. Scan for the earliest complete real open first, then pick the earliest hold-back position only when no complete tag was found. Regressions pinned in parser.test.ts and strip.test.ts for both new finding shapes. * fix(web): keep HR-shaped lines inside paragraph regions for inline-code scanning The previous walker closed inline-scan regions on lines matching the HR regex, but `parseBlocks()` in runtime/markdown.tsx does not break a paragraph on HR — its inner accumulation loop only breaks on blank / fence / heading / ul / ol (runtime/markdown.tsx:95-104). HR is only an HR block in the outer loop's first-look, never mid-paragraph. So inputs like `intro \`\n---\n<artifact …>…</artifact>\n---\nclosing \`` are one paragraph in the renderer, whose two stray backticks pair to cover the literal artifact recitation — but the walker was splitting on the `---` lines, leaving the recitation outside skip ranges, and the parser/stripper would treat it as a real tag. Drop HR from the paragraph-break list (HR-shaped lines carry no backticks of their own, so keeping them inside the surrounding region is benign either way) and document the renderer-mirror rationale. Regressions pinned on both sides. |
||
|
Sid
|
156bf5a34e
|
fix(web): refresh home projects after deleting a conversation (#1202) (#1219)
The home design cards render their `Needs input` badge from the cached `/api/projects` payload — App.tsx owns the `projects` state and exposes a `refreshProjects` callback that ProjectView already fires from every other state-changing branch (run end, live-artifact events, project rename, etc.). The conversation-delete branch silently skipped it: deleting a conversation that owned an unanswered `<question-form>` flips the daemon-side flag, but the home view kept showing the stale badge until the next manual reload. Call `onProjectsRefresh()` immediately after a successful `deleteConversation` API response (and only then — if the request fails the cached state is still the truth and we must not pretend otherwise). Adds `onProjectsRefresh` to the useCallback deps for exhaustive-deps correctness; matches the pattern at the four existing call sites in this file. New regression coverage in `apps/web/tests/components/ProjectView.deleteConversation.test.tsx`: - triggers onProjectsRefresh after deleting a conversation (verified RED before this fix, GREEN after) - does not trigger onProjectsRefresh when the delete request fails (defensive complement so a future "always refresh" refactor doesn't paper over a real failure with a stale-but-confident UI) |
||
|
shangxinyu1
|
10802bb0b0
|
test: expand nightly UI and desktop regression coverage (#1256)
* e2e(ui): cover examples preview flows * e2e(ui): cover Codex local CLI fallback UX * test: expand desktop and connector regression coverage * e2e(ui): cover workspace restoration flows * e2e(ui): cover retry recovery workspace flow * test: cover artifact and connector recovery flows * e2e(ui): cover Continue in CLI stale provenance flow * e2e(ui): cover BYOK model fetch caching * test: expand Orbit and desktop connector coverage * e2e(ui): cover workspace quick switcher recovery flows * e2e(ui): cover connector pending authorization recovery * e2e(ui): cover workspace and conversation restoration routes * e2e(ui): cover conversation draft and attachment restoration * e2e(ui): cover conversation history selection recovery * e2e(ui): cover workspace surface conversation selection * test: cover artifact presentation and orbit link behavior * test: cover artifact external link restoration * e2e(ui): cover root-route deep-link restoration * e2e(specs): cover Orbit open-artifact desktop click * e2e(specs): cover desktop artifact open link * test: fix Orbit settings fixture type drift * test: split Playwright critical and extended suites * test: fix ProjectView design template fixtures * ci: split workspace test stages * guard: allow split Playwright suite scripts * test: shrink Playwright critical suite * test: restore omitted Playwright suites |
||
|
PerishFire
|
8c0fb8dc01
|
feat(tools-pr): add maintainer PR-duty workspace (#1259)
* feat(tools-pr): add maintainer PR-duty workspace
Adds `tools/pr` as the maintainer-only control plane for PR-duty work on
this repo. Thin `gh` wrapper that encodes repo-specific knowledge:
review lanes, forbidden surfaces, lane-specific checklists, validation
command derivation from touched packages.
Subcommands:
- `list` — triage open queue by lane and review-state bucket.
- `view <num>` — agent-friendly review brief for a single PR.
- `classify [num]` — emit script-level tags for one PR or the whole
open queue; full-queue JSON output lands under `.tmp/tools-pr/classify/`
with rate-limit telemetry per run.
- `assignment` — assigner-perspective view of PR ownership, idle time,
and blockers (derived from existing tags; no new judgments).
Tag dictionary (13 tags) covers: bot-only-approval, needs-rebase,
forbidden-surface, unlabeled, duplicate-title, non-ascii-slug,
maintainer-edits-disabled, org-member, unresolved-changes-requested,
stale-approval, and three awaiting-* timing tags. Each rule is
expressible as one factual sentence over `gh` data + repo paths — see
`tools/pr/AGENTS.md` for the full dictionary plus precision rules.
Templates in `tools/pr/templates/*.md` are aesthetic references for
recurring maintainer comments (duplicate-title ask, awaiting-author
nudge, agent-review brief shape). `templates/examples/` holds
frozen-in-time agent-review snapshots for three PR shapes.
Infrastructure:
- `gh()` wraps `execFile` with minimum-touch retry (2 attempts at 1s + 2s
backoff) on transient 5xx / network errors. Persistent failures still
surface — retry is anti-jitter, not an exponential-backoff resilience
layer.
- Heavy chunks (`reviews`, `comments`, `commits`, assignment timelines)
use cursor-paginated `gh api graphql` via `fetchPaginatedPrList` to
stay under GitHub's GraphQL server-side timeout. Light chunks stay on
`gh pr list --json`.
- `fetchOrgMembers` cached per process via `gh api orgs/<owner>/members
--paginate`.
Wiring:
- Root `package.json` adds `pnpm tools-pr` to the allowed root entry
points.
- `scripts/postinstall.mjs` builds `tools/pr` alongside other workspace
packages.
- `scripts/guard.ts` allowlists `tools/pr/bin/tools-pr.mjs` and
`tools/pr/esbuild.config.mjs`, and adds `pr/` to the `tools/` top-level
layout allowlist.
- Root `AGENTS.md` and `tools/AGENTS.md` document the new command
surface, root-command-boundary update, and per-tool ownership.
* docs(agents): brief tools-pr in root AGENTS.md, link to tools/pr/AGENTS.md
Adds a `PR-duty tooling` section to the root AGENTS.md summarising what
`pnpm tools-pr` is, listing the four common subcommands (list / view /
classify / assignment), and pointing readers to `tools/pr/AGENTS.md` for
the full tag dictionary, operational playbook, templates, and design
rules. The section keeps root-level guidance to high-level orientation
while details stay local to the tool's own AGENTS.md.
* fix(tools-pr): drop overly broad touches-root-package.json forbidden hit
`deriveForbidden` was flagging any change to root `package.json` as a
forbidden-surface hit, but AGENTS.md §Root command boundary only forbids
specific *lifecycle* aliases (pnpm dev / test / build / daemon / preview
/ start) — tools-control-plane entrypoints like `pnpm tools-pr` are
explicitly allowed. Distinguishing "forbidden alias" from "allowed
entry" requires reading the diff content, which is `pnpm guard`'s job
rather than a path-derived classify tag.
Dogfooded on this branch's own PR (#1259), which added the `pnpm
tools-pr` script and was incorrectly flagged. Removing the hit aligns
the `forbidden-surface` tag with what tools-pr can mechanically detect
from file paths alone (apps/nextjs/, packages/shared/).
* fix(tools-pr): paginate commits fetch, recognise ready-to-merge, escape title-index separator
Three review follow-ups on #1259, all factual fixes:
- `fetchOpenPrCommits` now uses `fetchPaginatedPrList` instead of a
one-shot `pullRequests(first: $first)` query. GitHub GraphQL caps
connection page size at 100, so the previous implementation would
fail at runtime when callers passed `--limit > 100`. The paginated
path makes the commits fetch consistent with the other heavy chunks
(reviews, comments, assignment timelines) and removes the artificial
ceiling entirely. The `limit` parameter is dropped from
`fetchOpenPrCommits`; the CLI `--limit` continues to bound the
`gh pr list --json` chunks.
- `deriveStatus` in `assignment.ts` now reads `facts.reviewDecision`
and `facts.mergeStateStatus`. When the PR is `APPROVED` with merge
state `CLEAN` or `UNSTABLE` and carries no blockers, status renders
as `ready to merge` instead of falling through to `in review`. The
assignment view loses its main triage signal without this — a clean
human-approved PR rendered identical to a REVIEW_REQUIRED one.
- `tags.ts:tagDuplicateTitle` and `tags.ts:buildContext` both
constructed the title-index key with a literal NUL byte between
author and title, which made the file appear as binary in `git diff`
/ review tooling. Replaced the literal byte with a Unicode escape
sequence in source; the runtime string value is identical, the
source stays plain text and round-trips through review tooling
cleanly.
* fix(tools-pr): raise default --limit to 1000 to cover the live open queue
mrcfps flagged that `tools-pr list` (and `classify --all`, `assignment`)
defaults to `--limit 100`, which silently drops every PR past the first
100 in the open queue. The repo currently sits at 104 open PRs, so the
out-of-the-box run was already omitting four PRs.
Raise the default to 1000 in `list.ts`, `classify.ts`, and `assignment.ts`,
and remove the now-pointless 200 ceiling — `gh pr list --limit N` paginates
internally, so a high cap is cheap. Users can still pass `--limit <small>`
for a truncated preview. CLI help text on the three subcommands updated to
match.
* fix(web): pass designTemplates to ProjectView render helper
#955 made `designTemplates` a required Prop on ProjectView, but the test
helper added in #1244 (`renderProjectView` in
`ProjectView.api-empty-response.test.tsx`) was never updated. The two
PRs landed on main without conflicting, leaving `apps/web` typecheck red
for every PR that rebases past
|
||
Tom Huang
|
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
(
|
||
|
PerishFire
|
f2db5a749c
|
chore: enforce PR→issue linking discipline (#1263)
PRs that omit Fixes #N break the release-time reverse lookup (issue → closing PR → merge sha → first containing tag), since the auto-link only fires on the explicit closing keywords. We've been doing this by hand on recent fixes; codify it so future PRs don't drift. - Add .github/pull_request_template.md with a Fixes # placeholder so the link surface is in front of the author by default. - Add a corresponding bullet to the Bug follow-up workflow in the root AGENTS.md so the discipline lives next to the methodology that produces issue-linked work. |
||
|
PerishFire
|
a797e079b1
|
fix(desktop): exit fullscreen before hiding window on macOS close (#1249)
* fix(desktop): exit fullscreen before hiding window on macOS close (#1215) When a preview is in 演示 → 全屏 mode, the macOS close handler called window.hide() directly, leaving the OS fullscreen Space orphaned as a black screen — the window vanished but the Space stayed up. Extract hideWindowExitingFullscreen as the named invariant ("hide, but first leave fullscreen so the OS Space tears down with the window") and route the darwin close handler through it. The hide is deferred until 'leave-full-screen' fires so we don't race the OS Space teardown. Bootstraps Vitest on apps/desktop with a single test under tests/main/hide-window-exiting-fullscreen.test.ts that exercises the helper through a structural mock — the bug shape is pure logic, no real Electron window required. Spec was red against a hide-only helper and green after the leave-full-screen sequencing. * docs(agents): codify bug follow-up workflow Distill the spec-first / cheapest-layer / scope-discipline / invariant-shaped-fix / baseline-diff playbook used recently on #135 and #1215 into a top-level subsection of root AGENTS.md, framed as a default action shape with explicit room for case-by-case judgment rather than a hard rule. Includes a single pointer back to the worked example spec. * docs(agents): require staged human verification for visible bugs Add the human-verification gate as a sixth bullet in the Bug follow-up workflow. UI / platform-native / animation symptoms can pass green specs and still ship the visible regression — proven by #1215, where the desktop unit test green-lighted the helper logic but only a side-by-side buggy-vs-fix run on a real macOS Space proved the black-screen actually went away. Reinforces the production-API-only seed constraint while we're there: source-level backdoors prove a fake flow, not the real one, so they invalidate the verification. * fix(desktop): defer hide across the fullscreen-enter transition (#1215) mrcfps observed on PR #1249 that the close handler only catches windows already in fullscreen — Electron's enter-full-screen event is async on macOS, so isFullScreen() can still read false during the OS Space transition triggered by requestFullscreen(). A close in that window took the plain hide() path and stranded the same black Space the fix was meant to eliminate. Track in-flight fullscreen entry from webContents.enter-html-full-screen (set) and BrowserWindow.{enter,leave}-full-screen (clear), and surface it through WindowFullscreenSurface.isEnteringFullscreen. The helper now parks on enter-full-screen until the OS confirms the Space, then runs the existing exit-then-hide path. Adds a regression test ("waits out a fullscreen-enter transition before exiting and hiding") that goes red against the previous helper. |
||
|
Caprika
|
f7f2661bda
|
[codex] Handle empty API responses as no output (#1244)
* Handle empty API responses as no output * Fix empty API response comment cleanup * Stabilize API empty response detection |
||
|
PerishFire
|
421ddf553c
|
fix(pack/win): close running app before silent reinstall (#1238) | ||
|
nettee
|
e859c31574
|
fix(web): complete finished tool calls missing results (#1240) |