codeflow-maintainer
Skill75 starsChanged 4 months agoMITView on GitHub
OpenClaw-only maintainer PR workflow modes and skills. For OpenCoven/coven PR creation, redirect to skills/pr-agent, the Coven PR Readiness Agent.
In this repository59 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- skills/heygen-skills/CLAUDE.mdCLAUDE.md
- 1passwordSkill
- ask-curlSkill
- boardSkill
- cleanup-unused-filesSkill
- codex-session-managerSkill
- coven-astraSkill
- coven-board-entrySkill
- coven-charmSkill
- coven-cloud-maintainerSkill
- coven-codySkill
- coven-echoSkill
- coven-identity-canonSkill
- coven-kittySkill
- coven-parallel-workSkill
- coven-sageSkill
- coven-task-managerSkill
- daily-dev-agenticSkill
- discrawlSkill
- disk-space-optimizerSkill
- epubSkill
- familiar-board-stewardshipSkill
- heygen-avatarSkill
- heygen-videoSkill
- heygen-skillsSkill
- higgsfield-generateSkill
- higgsfieldSkill
- linear-issue-managementSkill
- lit-ui-designerSkill
- lobsterSkill
- maintainer-barSkill
- memory-recallSkill
- memory-timeline-managerSkill
- ml-engineerSkill
- nodriverSkill
- ocrSkill
- openclaw-devSkill
- openclaw-trustSkill
- opencoven-designSkill
- OpenCoven Role Creation ProcessSkill
- pr-agentSkill
- project-managerSkill
- prompt-engineerSkill
- prompt-vaultSkill
- research-ingestionSkill
- ritual-dapp-frontendSkill
- rollcallSkill
- security-agentSkill
- skill-scannerSkill
- slidev-deck-pipelineSkill
- tauri-apple-releaseSkill
- threejs-animationSkill
- tinyfish-agent-runSkill
- tinyfish-browserSkill
- tinyfish-fetchSkill
- tinyfish-searchSkill
- tinyfishSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
14 rules. About 0 are common, 9 are distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈9
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 52
Do not trust PR code by default
- Very specific: line 53
Keep types strict (no any)
- Distinctive: line 69
Do we have any questions?
- Very specific: line 89
Add changelog entry with (#PR) and thanks @author
- Very specific: line 91
Use scripts/committer for commits
- Distinctive: line 99
Follow-up refactors to defer?
- Distinctive: line 113
All BLOCKER and IMPORTANT findings resolved
- Distinctive: line 120
Post-merge: Verify PR reaches MERGED state.
- Very specific: line 135
Update clawtributors for new contributors (bun scripts/update-clawtributors.ts)
- Distinctive: line 136
Create follow-up issues for deferred refactors
- Distinctive: line 156
OpenClaw: use the OpenClaw maintainer skill bundle.
- Very specific: line 157
OpenCoven/coven: use skills/pr-agent, the Coven PR Readiness Agent.
- Distinctive: line 174
Follow the mode progression strictly
- Distinctive: line 175
Use the checkpoint questions before advancing
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
CodeFlow Maintainer Skill
Source of Truth
PR_WORKFLOW.md: https://github.com/openclaw/maintainers/blob/main/.agents/skills/PR_WORKFLOW.md
This skill is OpenClaw-only. It implements the workflow defined in
openclaw/maintainers for OpenClaw repositories. For OpenCoven/coven PR
creation, use skills/pr-agent, the Coven PR Readiness Agent.
Working Rule
Skills execute workflow. Maintainers provide judgment. Always pause between skills to evaluate technical direction, not just command success.
Workflow Modes
1. Triage
Purpose: Assess whether PR is worth reviewing.
Focus:
- Is the problem statement clear?
- Does a realistic verification path exist?
- Is this PR worth our time?
- Are there duplicates?
Recommendation: Proceed to Review, Request Info, or Close.
Agent prompt:
[Mode: TRIAGE] Assess whether this PR is worth reviewing.
Focus on problem clarity, scope match, duplicate detection.
DO NOT suggest code changes, approvals, or merges.2. Review (review-pr)
Purpose: Review only — produce findings and recommendation. DO NOT modify code.
Quality bar:
- Do not trust PR code by default
- Keep types strict (no
any) - Evaluate security impact and abuse paths
- Understand the system before changing it
Finding severities:
BLOCKER— must fix before prepareIMPORTANT— must fix before prepareMINOR— should fix (optional)NOTE— informational only
Artifacts: .local/review.md, .local/review.json
Checkpoint before advancing:
- What problem are they trying to solve?
- What is the most optimal implementation?
- Can we fix up everything?
- Do we have any questions?
Agent prompt:
[Mode: REVIEW] Produce findings, DO NOT modify code.
Focus: correctness, security, tests, types, scope, hardening.
Severity: BLOCKER > IMPORTANT > MINOR > NOTE.
Recommendation: READY FOR PREPARE, NEEDS WORK, NEEDS DISCUSSION, or CLOSE.3. Prepare (prepare-pr)
Purpose: Make PR merge-ready on its head branch.
Order: Rebase → Fix BLOCKER/IMPORTANT → Changelog → Gates → Push
Gates: pnpm build, pnpm check, pnpm test
Rules:
- Rebase onto main FIRST before any fixes
- Add changelog entry with
(#PR)andthanks @author - Push to PR head branch only, NEVER to main
- Use
scripts/committerfor commits
Checkpoint before advancing:
- Most optimal implementation?
- Properly scoped and reusing existing logic?
- Properly typed and hardened?
- Enough meaningful tests? (fake timers where appropriate)
- Any security vulnerabilities introduced?
- Follow-up refactors to defer?
Agent prompt:
[Mode: PREPARE] Make PR merge-ready. Rebase first, then fix, then gates.
Push to PR head branch only, never to main.
Gates: pnpm build, pnpm check, pnpm test.4. Merge (merge-pr)
Purpose: Final verification and squash-merge.
Go/no-go checklist:
- All BLOCKER and IMPORTANT findings resolved
- Meaningful verification, low regression risk
- Changelog updated with PR # and thanks @author
- CI green, branch not behind main
Merge method: Squash with Co-authored-by: trailers for PR author + maintainer.
Post-merge: Verify PR reaches MERGED state. Post comment with merge SHA.
Agent prompt:
[Mode: MERGE] All gates must be green. Squash merge with Co-authored-by.
Verify MERGED state. Post comment with SHA.
DO NOT modify code — go back to Prepare if needed.5. Post-Merge
Purpose: Cleanup, attribution, follow-ups.
Tasks:
- Delete branch
- Update clawtributors for new contributors (
bun scripts/update-clawtributors.ts) - Create follow-up issues for deferred refactors
- Clean up worktree
Mode Transitions
Triage → Review → Prepare → Merge → Post-Merge
↑___________↓ (back to Prepare if changes needed)- Each transition requires checkpoint questions answered
- Modes cannot be skipped (triage → prepare is not valid)
read-onlyoverrides everything when active
Related Skills
PR Agent Orchestrator
For concrete workflow execution, see the pr-agent skill:
- OpenClaw: use the OpenClaw maintainer skill bundle.
- OpenCoven/coven: use
skills/pr-agent, the Coven PR Readiness Agent.
The pr-agent skill provides:
- Automatic artifact validation (
.local/*.json,.local/*.md) - Script wrapper invocation (
scripts/pr-review,scripts/pr-prepare,scripts/pr-merge) - Checkpoint enforcement between phases
- Escalation trigger detection
Inheritable Skills
This skill is designed to be inherited by other agents:
- review-pr: https://github.com/openclaw/maintainers/tree/main/.agents/skills/review-pr
- prepare-pr: https://github.com/openclaw/maintainers/tree/main/.agents/skills/prepare-pr
- merge-pr: https://github.com/openclaw/maintainers/tree/main/.agents/skills/merge-pr
Agents inheriting this skill should:
- Follow the mode progression strictly
- Use the checkpoint questions before advancing
- Inject the mode-specific prompt into their system context
- Respect the quality bar (types, security, tests, scope)
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

