Ready to release v0.3.4 by Keith-CY · Pull Request #162 · lay2dev/clawpal

Keith-CY · 2026-04-01T08:54:00Z

No description provided.

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com> Closes #123 (Phase 1)

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

Ensures app_logs, recipe_cmds, upgrade, util, and watchdog_cmds can access shared types from the parent mod.rs module.

refactor: extract commands/mod.rs into domain submodules (Phase 3)

- docs/architecture/overview.md: system overview, tech stack, layer diagram, directory guide, key data flows, constraints - docs/architecture/commands.md: command layer structure, module organization, new command workflow, remote_* proxy explanation, prohibited patterns - .github/CODEOWNERS: ownership rules for architecture docs, Tauri capabilities, core domain, CI workflows - Update harness plan: mark completed items Ref #123

- Upgrade business-flow-test-matrix.md to standard gate doc (6 gate levels) - Add failure-diagnosis.md runbook (diagnosis flow, rollback procedures) - Add entropy-governance.md (weekly checklist, metrics template) - Update harness plan: mark all phases complete, list deferred items Ref #123

chore: complete harness engineering standard (Phase 3b + Phase 4)

…lling, render probes (#132) * perf: optimize Home page data loading (P0 + P1) P0 fixes: - Parallelize initial data requests: remove artificial setTimeout delays (350ms model profiles, 250/400ms runtime snapshot). All requests now fire simultaneously on mount. - Merge 3 separate polling intervals (queuedCommandsCount, runtimeSnapshot, statusExtra) into a single unified poll loop with shouldPollResource() controlling per-resource frequency. Reduces backend call volume ~60%. - Skip redundant ConfigSnapshot fetch when RuntimeSnapshot is already in persisted cache (they share agents, globalDefaultModel, fallbackModels). P1 fixes: - Fix unhealthy status flash: config-only initial state now sets healthy=null instead of healthy=false, rendering 'Checking...' badge instead of a false red 'Unhealthy' badge during SSH connection setup. - Instant model profile population: add listModelProfiles to persisted read cache methods, initialize state from cache on mount. Infrastructure: - Add RenderProbe class (src/lib/render-probe.ts): production-grade performance.mark() instrumentation for measuring first-render latency of each UI section. Exposes data via window.__RENDER_PROBES__. - Add Home Perf E2E workflow: Docker OpenClaw container + Playwright + IPC mock collects render timings, posts/updates a single PR comment via peter-evans/create-or-update-comment (comment-tag: home-perf-e2e). Tests: 12 new unit tests across 3 files (all pass). * fix: Dockerfile heredoc → COPY, switch to sticky-pull-request-comment - Dockerfile: replace heredoc RUN (unsupported) with COPY from seed/ dir - Workflow: replace peter-evans/create-or-update-comment (no comment-tag) with marocchino/sticky-pull-request-comment (header-based upsert) - Add fallback report.md generation for failed runs * fix: add git to Docker image (required by openclaw npm install) * fix: install @playwright/test package before running perf E2E * fix: comprehensive IPC mock + click-through StartPage in perf E2E - Add mock handlers for StartPage flow (list_registered_instances, discover_local_instances, list_ssh_hosts, precheck_*, connect_*, etc.) - Playwright spec now waits for app init, then clicks local instance card to navigate into Home where probes fire - Handle partial probe collection gracefully (timeout → skip run) * fix: remote SSH default model not loading (JSON pointer + skip logic) Bug 1 (pre-existing): remote_instance_runtime_snapshot_impl runs `openclaw config get agents --json` which returns the agents subtree, but extract_default_model_and_fallbacks uses /agents/defaults/model JSON pointer expecting the full config. The subtree has no /agents prefix, so globalDefaultModel and fallbackModels are always null for remote instances. Fix: wrap config_json in { "agents": config_json } before calling extract_default_model_and_fallbacks. Bug 2 (introduced in this PR): shouldSkipConfigSnapshot only checked persistedRuntimeSnapshot != null, not whether globalDefaultModel had a value. When the cached RuntimeSnapshot had null model data (from Bug 1), ConfigSnapshot was still skipped — the only path that could correctly read the model via SFTP. Fix: shouldSkipConfigSnapshot now checks globalDefaultModel != null. * style: cargo fmt overview.rs --------- Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

#139) Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

…l contributors) - screenshot.yml: checkout from fork repo, skip push/comment for fork PRs - coverage.yml: skip PR comment for fork PRs (read-only GITHUB_TOKEN) - metrics.yml: skip PR comment for fork PRs (read-only GITHUB_TOKEN) Fork PRs still run all checks/builds/tests — only write-back steps (pushing refs, posting PR comments) are skipped since the fork token lacks write permissions.

- coverage.yml: save comment body as artifact for fork PR pickup - metrics.yml: save comment body as artifact for fork PR pickup - fork-pr-comment.yml: new workflow_run handler that downloads comment artifacts and posts them with write permissions Same-repo PRs still post comments directly (no change). Fork PRs now get full metrics/coverage visibility via the workflow_run relay.

Download step uses continue-on-error so the workflow doesn't fail when the upstream workflow failed before producing the comment artifact. Subsequent steps gated on download success.

…enshot.yml When triggered via workflow_dispatch, pull_request context fields are empty. Use '|| github.repository' and '|| github.ref' fallbacks so manual runs still check out the correct code. Addresses Keith-CY's review feedback on PR #143.

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com> Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Co-authored-by: cjs <cjs@cjs-mini.local> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Jason Chai <jingsongcai@126.com> Co-authored-by: zhixian <zhixian@bunker.local> Co-authored-by: dev01lay2 <dev@lay2.dev> Co-authored-by: lay2OcBot <bot@openclaw.ai> fix: write model registrations to correct config path (#148) fix: resolve Channels page infinite refresh and data loading failure (#149)

…sis (#155) Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

#157) - Add guildResolutionWarning and channelResolutionWarning fields to DiscordGuildChannel (backend + frontend types) - Frontend merge (channel-cache.ts) now clears warnings when cached names resolve the guild/channel — only shows warnings when both network request AND cache miss - Channels.tsx uses specific warning fields: guild-level header shows only guild warnings, channel cards show only channel warnings - Backend discovery.rs populates specific fields with clearer messages distinguishing Guild (server) vs Channel resolution failures Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

…on remote host (#159) Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

* fix: prevent Apply stuck state on gateway restart timeout Backend: fire-and-forget gateway restart via tokio::spawn with 15s timeout so the apply function returns immediately instead of hanging on SSH exec. Frontend: 30s safety timeout in handleApply to recover UI state even if the Tauri IPC never resolves. * fix: add 30s timeout to run_openclaw_remote_with_env The real hang occurs during command execution (openclaw config set), not just at gateway restart. When config changes trigger an automatic gateway restart, the SSH exec channel hangs indefinitely because session.exec() has no timeout. This causes queues.clear() to never execute, so after page refresh the pending changes bar still shows '1 pending change' with an empty diff (config was already written before the hang). * fix: scope remote command timeouts by class + prevent double-apply - Remote commands now use per-class timeouts: install/bootstrap/upgrade/setup get 180s, all others keep 30s. This prevents false timeouts during network I/O heavy operations (addresses review feedback #1). - Add applyInFlightRef guard to handleApply so the safety timeout unblocking the UI after 30s cannot trigger a second concurrent apply. The ref stays true until the request settles, preventing racing applies even after the button re-enables (addresses review feedback #2). --------- Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

github-actions · 2026-04-01T08:55:58Z

📊 Test Coverage Report

Metric	Base (`main`)	PR (`develop`)	Delta
Lines	74.34% (6134/8251)	74.00% (6333/8558)	🔴 -0.34%
Functions	68.88% (704/1022)	68.89% (724/1051)	🟢 +0.01%
Regions	75.86% (10156/13388)	75.38% (10393/13788)	🔴 -0.48%

Coverage measured by cargo llvm-cov (clawpal-core + clawpal-cli).

dev01lay2

Reviewed for release-blocking (P1) issues. I did not find any P1 blockers in this release PR.

github-actions · 2026-04-01T09:03:40Z

📸 UI Screenshots

Commit: b9b364568589f29447b1f43bcb7a9b6684c6565a | Screenshots: Download artifact

Light Mode — Core Pages

Start Page	Home	Channels

Recipes	Cron	Doctor

Context	History	Chat Panel

Settings (4 scroll positions)

Main	Appearance	Advanced	Bottom

Start Page Sections

Overview	Profiles	Settings

Dark Mode

Start	Home	Channels	Doctor

Dark mode — more pages

Recipes	Cron	Settings

Responsive + Dialogs

Home 1024×680	Chat 1024×680	Create Agent

🔧 Harness: Docker + Xvfb + tauri-driver + Selenium | 28 screenshots, 13 flows

github-actions · 2026-04-01T09:04:44Z

📏 Metrics Gate Report

Status: ✅ All gates passed

Commit Size ❌

Metric	Value	Limit	Status
Commits checked	29	—	—
All within limit	16/29	≤ 500 lines	❌
Largest commit	46681 lines	≤ 500	❌

Bundle Size ✅

Metric	Value	Limit	Status
JS bundle (raw)	1077 KB	—	—
JS bundle (gzip)	329 KB	≤ 350 KB	✅
JS initial load (gzip)	173 KB	≤ 180 KB	✅

Perf Metrics E2E ✅

Metric	Value	Limit	Status
Tests	10 passed, 0 failed	0 failures	✅
RSS (test process)	3.2 MB	≤ 20 MB	✅
VMS (test process)	270.2 MB	—	ℹ️
Command P50 latency	28 µs	≤ 1000 µs	✅
Command P95 latency	43 µs	≤ 5000 µs	✅
Command max latency	62 µs	≤ 50000 µs	✅

Command Perf (local) ✅

Metric	Value	Status
Tests	4 passed, 0 failed	✅
Commands measured	5	ℹ️
RSS (test process)	4.4 MB	ℹ️

Local command timings

Command	P50 (µs)	P95 (µs)	Max (µs)
local_openclaw_config_exists	1	9	9
list_ssh_hosts	2	10	10
get_app_preferences	16	37	37
read_app_log	60	95	95
read_error_log	5	10	10

Command Perf (remote SSH) ✅

Metric	Value	Status
SSH transport	OK	✅
Command failures	9/15 runs	ℹ️ Docker (no gateway)

Remote command timings (via Docker SSH)

Command	Median	Max
openclaw_status	2258 ms	2275 ms
cat__root_.openclaw_openclaw.json	240 ms	240 ms
openclaw_gateway	1597 ms	1598 ms
openclaw_cron	1266 ms	1305 ms
openclaw_agent	2225 ms	2232 ms

Home Page Render Probes (real IPC) ✅

Probe	Value	Limit	Status
status	61 ms	≤ 500 ms	✅
version	179 ms	≤ 500 ms	✅
agents	9 ms	≤ 500 ms	✅
models	9 ms	≤ 500 ms	✅
settled	61 ms	≤ 500 ms	✅

Code Readability

File	Lines	Target	Status
`commands/mod.rs`	11863	≤ 300	⚠️
`commands/doctor_assistant.rs`	5637	≤ 3000	⚠️
`commands/rescue.rs`	3402	≤ 2000	⚠️
`cli_runner.rs`	3295	≤ 1200	⚠️
`commands/profiles.rs`	1923	≤ 1500	⚠️
`commands/sessions.rs`	1828	≤ 1100	⚠️
`commands/discovery.rs`	1702	≤ 1000	⚠️
`commands/credentials.rs`	1629	≤ 1000	⚠️
`openclaw_doc_resolver.rs`	1362	≤ 800	⚠️
`commands/ssh.rs`	1216	≤ 700	⚠️
`commands/doctor.rs`	1193	≤ 700	⚠️
`pages/Settings.tsx`	1100	≤ 700	⚠️
`recipe_adapter_tests.rs`	1100	≤ 700	⚠️
`ssh.rs`	980	≤ 600	⚠️
`agent_identity.rs`	937	≤ 600	⚠️
`pages/StartPage.tsx`	898	≤ 500	⚠️
`recipe_library.rs`	884	≤ 500	⚠️
`pages/Cook.tsx`	881	≤ 500	⚠️
`pages/Recipes.tsx`	875	≤ 500	⚠️
`pages/Home.tsx`	867	≤ 500	⚠️
`recipe_library_tests.rs`	861	≤ 500	⚠️
`lib/types.ts`	856	≤ 500	⚠️
`install/commands.rs`	839	≤ 500	⚠️
`commands/backup.rs`	788	≤ 500	⚠️
`lib/use-api.ts`	774	≤ 500	⚠️
`recipe_adapter.rs`	757	≤ 500	⚠️
`pages/RecipeStudio.tsx`	680	≤ 500	⚠️
`components/SessionAnalysisPanel.tsx`	679	≤ 500	⚠️
`commands/model.rs`	645	≤ 500	⚠️
`bridge_client.rs`	645	≤ 500	⚠️
`recipe_action_catalog.rs`	631	≤ 500	⚠️
`App.tsx`	625	≤ 500	⚠️
`lib/actions.ts`	622	≤ 500	⚠️
`components/InstallHub.tsx`	619	≤ 500	⚠️
`recipe_workspace.rs`	613	≤ 500	⚠️
`agent_fallback.rs`	609	≤ 500	⚠️
`recipe.rs`	596	≤ 500	⚠️
`commands/config.rs`	549	≤ 500	⚠️
`commands/discord.rs`	545	≤ 500	⚠️
`recipe_runtime/systemd.rs`	537	≤ 500	⚠️
`commands/overview.rs`	533	≤ 500	⚠️
`lib/api.ts`	527	≤ 500	⚠️
`install/runners/docker.rs`	525	≤ 500	⚠️
`commands/types.rs`	524	≤ 500	⚠️
`commands/instance.rs`	501	≤ 500	⚠️
`markdown_document.rs`	497	≤ 500	✅
`pages/Channels.tsx`	480	≤ 500	✅
`pages/Doctor.tsx`	479	≤ 500	✅
`node_client.rs`	452	≤ 500	✅
`commands/discover_local.rs`	441	≤ 500	✅
`lib/__tests__/guidance.test.ts`	439	≤ 500	✅
`recipe_executor.rs`	437	≤ 500	✅
`components/__tests__/DoctorRecoveryOverview.test.tsx`	429	≤ 500	✅
`pages/Cron.tsx`	428	≤ 500	✅
`recipe_executor_tests.rs`	422	≤ 500	✅
`hooks/useInstanceDataStore.ts`	421	≤ 500	✅
`bug_report/queue.rs`	409	≤ 500	✅
`lib.rs`	408	≤ 500	✅
`lib/api-read-cache.ts`	405	≤ 500	✅
`commands/channels.rs`	405	≤ 500	✅
`components/__tests__/InstanceCard.test.tsx`	400	≤ 500	✅
`components/DoctorRecoveryOverview.tsx`	383	≤ 500	✅
`pages/Orchestrator.tsx`	372	≤ 500	✅
`recipe_tests.rs`	360	≤ 500	✅
`lib/use-cached-query.ts`	353	≤ 500	✅
`hooks/useSshConnection.ts`	353	≤ 500	✅
`components/DoctorTempProviderDialog.tsx`	349	≤ 500	✅
`commands/precheck.rs`	346	≤ 500	✅
`bug_report/collector.rs`	335	≤ 500	✅
`components/InstanceCard.tsx`	334	≤ 500	✅
`hooks/useWorkspaceTabs.ts`	331	≤ 500	✅
`lib/doctor-report-i18n.ts`	328	≤ 500	✅
`components/ParamForm.tsx`	324	≤ 500	✅
`pages/__tests__/cook-plan-context.test.ts`	319	≤ 500	✅
`pages/__tests__/overview-loading.test.ts`	318	≤ 500	✅
`hooks/useChannelCache.ts`	317	≤ 500	✅
`components/__tests__/SshFormWidget.test.tsx`	312	≤ 500	✅
`components/PendingChangesBar.tsx`	308	≤ 500	✅
`hooks/useInstancePersistence.ts`	303	≤ 500	✅
`doctor.rs`	303	≤ 500	✅
`recipe_planner_tests.rs`	302	≤ 500	✅
Files > 500 lines	45	trend ↓	⚠️
Files over target	45	0	⚠️

📊 Metrics defined in docs/architecture/metrics.md

github-actions · 2026-04-01T09:06:50Z

📦 PR Build Artifacts

Platform	Download	Size
Windows-x64	📥 clawpal-Windows-x64	27.5 MB
Linux-x64	📥 clawpal-Linux-x64	105.8 MB
macOS-x64	📥 clawpal-macOS-x64	14.0 MB
macOS-ARM64	📥 clawpal-macOS-ARM64	13.2 MB

🔨 Built from 47acbe5 · View workflow run
⚠️ Unsigned development builds — for testing only

github-actions · 2026-04-01T09:09:42Z

Recipe GUI E2E Perf

Artifacts: perf report

Recipe	Page Load (ms)	Form Fill (ms)	Execution (ms)	Verification (ms)	Total (ms)
dedicated-agent	—	—	—	—	⚠️ Skipped: Discord not configured in E2E environment
channel-persona-pack	—	—	—	—	⚠️ Skipped: Discord not configured in E2E environment
agent-persona-pack	—	—	—	—	⚠️ Skipped: Discord not configured in E2E environment

Harness: Docker + Xvfb + tauri-driver + Selenium

dev01lay2 and others added 30 commits March 16, 2026 15:23

chore: establish harness engineering standard (Phase 1 — 仓库入口归一) (#124)

e1f1116

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com> Closes #123 (Phase 1)

chore: harness engineering Phase 2 — 验证与流程归一 (#125)

3965880

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

refactor: extract commands/mod.rs into domain submodules

da7e3d2

style: cargo fmt --all

8f17a07

fix: add missing use super::* imports to extracted modules

af051e0

Ensures app_logs, recipe_cmds, upgrade, util, and watchdog_cmds can access shared types from the parent mod.rs module.

Merge pull request #126 from lay2dev/chore/harness-phase3

c6ce937

refactor: extract commands/mod.rs into domain submodules (Phase 3)

chore: remove CODEOWNERS (unnecessary for current team size)

382eb9f

Merge pull request #127 from lay2dev/chore/harness-phase3b

9f81447

chore: complete harness engineering standard (Phase 3b + Phase 4)

chore: add metrics framework, instrumentation, and CI gates (#136)

05c92fb

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

refactor: reduce commands/mod.rs from 8,869 to 230 lines (#137)

1f039de

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

feat: UI screenshot harness — automated visual acceptance for every PR (

5d6b568

#139) Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

chore: optimize metrics framework and codebase readability (#140)

50db11b

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

fix: replace mock IPC with real SSH bridge for Home render probes (#141)

e912222

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

fix(ci): handle missing artifacts in fork-pr-comment workflow

242e262

Download step uses continue-on-error so the workflow doesn't fail when the upstream workflow failed before producing the comment artifact. Subsequent steps gated on download success.

fix(ci): save screenshots before orphan checkout clears worktree

45463b1

chore: add local CI scripts, metrics gate, and pre-commit hook (#144)

8528c0e

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

feat: reimplement stream-based session and backup loading (#145)

6eb41b5

chore: add lefthook for automatic pre-commit hook installation (#146)

d3d26ad

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

feat(ssh): add exec_streaming for bounded-memory remote session analy…

28fec55

…sis (#155) Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

fix: flatten Context page nesting and fix session loading state (#156)

379ada8

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

fix(security): proxy Discord API calls through SSH to keep bot token …

a188a51

…on remote host (#159) Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

feat: move profile sync to Profiles page with selectable devices (#161)

47acbe5

Co-authored-by: dev01lay2 <dev01lay2@users.noreply.github.com>

Keith-CY temporarily deployed to production April 1, 2026 08:54 — with GitHub Actions Inactive

dev01lay2 approved these changes Apr 1, 2026

View reviewed changes

Keith-CY merged commit 5847639 into main Apr 1, 2026
26 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Ready to release v0.3.4#162

Ready to release v0.3.4#162
Keith-CY merged 31 commits intomainfrom
develop

Keith-CY commented Apr 1, 2026

Uh oh!

github-actions bot commented Apr 1, 2026

Uh oh!

dev01lay2 left a comment

Uh oh!

Uh oh!

github-actions bot commented Apr 1, 2026

Uh oh!

github-actions bot commented Apr 1, 2026

Uh oh!

github-actions bot commented Apr 1, 2026

Uh oh!

github-actions bot commented Apr 1, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

Keith-CY commented Apr 1, 2026

Uh oh!

github-actions bot commented Apr 1, 2026

📊 Test Coverage Report

Uh oh!

dev01lay2 left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

github-actions bot commented Apr 1, 2026

📸 UI Screenshots

Light Mode — Core Pages

Dark Mode

Responsive + Dialogs

Uh oh!

github-actions bot commented Apr 1, 2026

📏 Metrics Gate Report

Commit Size ❌

Bundle Size ✅

Perf Metrics E2E ✅

Command Perf (local) ✅

Command Perf (remote SSH) ✅

Home Page Render Probes (real IPC) ✅

Code Readability

Uh oh!

github-actions bot commented Apr 1, 2026

📦 PR Build Artifacts

Uh oh!

github-actions bot commented Apr 1, 2026

Recipe GUI E2E Perf

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants