coding-agent-pm
Plan, delegate, review, and recover coding-agent implementation work with tight scope control, acceptance criteria, and validation.
Install / Use
npx skills add davepoon/buildwithclaude --skill coding-agent-pmInstalls into whichever agent you are using.
SKILL.md
Installable skill definition
Quality Score
Category
Development & EngineeringSupported Platforms
Our assessment of coding-agent-pm
coding-agent-pm scores 89/100 on our quality scale, 1503rd of 4,698 Development & Engineering skills we index (top 32%).
Its SKILL.md is 3.5 KB long, well organised into 16 sections with 3 code examples: a solid amount of guidance for an agent.
With 3,752 GitHub stars, it is one of the more widely adopted skills in the catalogue.
Maintenance, license and trust
- The repository was last updated today, so coding-agent-pm is actively maintained.
- Our last check on 2026-10-31 found the source still online.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 100/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.
AI review by kimi-k2.7-code on 2026-10-02. Automated pattern scan on 2026-10-02. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.
coding-agent-pm compared with similar skills
All 4 of these similar skills score higher than coding-agent-pm; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| coding-agent-pm (this skill)by davepoon | 89 | 3.8k | today | SKILL.md |
| taste-skillby Leonxlnx | 100 | 116.8k | 8d ago | CLAUDE.md |
| ai-job-searchby ai-job-search-dev | 100 | 72.1k | today | CLAUDE.md |
| claude-howtoby luongnv89 | 100 | 47.3k | 2d ago | CLAUDE.md |
| lobehubby lobehub | 100 | 85.6k | today | MCP Server |
Frequently asked questions
- How do I install coding-agent-pm?
- Run
npx skills add davepoon/buildwithclaude --skill coding-agent-pm. The install tabs above show the steps for each supported agent. - Which AI agents does coding-agent-pm work with?
- It is written for Claude Code and OpenAI Codex, as a SKILL.md file. Other agents that read the same format can often use it too.
- Is coding-agent-pm safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It is MIT-licensed and scores 100/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is coding-agent-pm still maintained?
- The repository was last updated today, so coding-agent-pm is actively maintained.
Skill content
View source on GitHubname: coding-agent-pm description: Plan, delegate, review, and recover coding-agent implementation work with tight scope control, acceptance criteria, and validation. category: ai-agents license: MIT
Coding Agent PM
Use this skill when a human is preparing, delegating, reviewing, or recovering implementation work done by Claude Code, Codex, OpenClaw, or another coding agent.
When to Use This Skill
- You need to turn a vague implementation request into a bounded agent task.
- You are assigning work to one or more coding agents and need clear file ownership.
- You are reviewing an agent diff for scope drift, regressions, or missing validation.
- A coding agent has stalled, over-edited, or returned a result that is hard to trust.
What This Skill Does
- Defines the shipped outcome in one sentence.
- Sets explicit scope: files or modules owned, files or modules excluded, and destructive actions that are forbidden.
- Captures relevant project context before implementation starts.
- Writes acceptance criteria that can be checked directly.
- Requires validation commands or manual checks before final delivery.
- Reviews the diff before accepting the work.
Assignment Rules
- Keep the task brief concrete and bounded.
- State whether the agent may edit files or should only inspect and plan.
- Give ownership boundaries for code changes.
- Tell the agent not to revert unrelated user work.
- Prefer existing project patterns over new abstractions.
- Require the final response to include changed files, checks run, results, and remaining risks.
Agent Brief Template
# Agent Brief
## Outcome
Ship:
## Scope
Owned files/modules:
Do not edit:
Forbidden actions:
- Do not run destructive git commands.
- Do not revert unrelated user changes.
## Context
Relevant files:
Existing patterns to follow:
## Acceptance Criteria
- [ ]
- [ ]
- [ ]
## Validation
Run:
Manual checks:
## Final Response Must Include
- Files changed
- Validation run and result
- Remaining risks or follow-ups
Review Checklist
Review in this order:
- Behavioral bugs or regressions.
- Missing acceptance criteria.
- Missing tests or weak validation.
- Scope drift and unrelated refactors.
- Security, privacy, or destructive-command risk.
If the agent drifted, stop new edits and ask it to map each change back to the original acceptance criteria.
Recovery Prompt
Use this when a coding agent's work is hard to trust:
Pause implementation. Map every changed file and every material change back to the original acceptance criteria. For each change, mark it as required, optional, or unrelated. Do not edit files until this map is complete.
Example
User: "Have a coding agent add OAuth login."
Better assignment:
Outcome: Add GitHub OAuth login to the existing auth flow.
Scope:
- Own: src/auth/*, src/routes/login.tsx, tests/auth/*
- Do not edit: billing, onboarding, database migrations unless a failing test proves it is required
Acceptance criteria:
- [ ] Existing email login still works.
- [ ] GitHub login creates or links a user through the existing auth service.
- [ ] Auth errors render through the existing form error pattern.
- [ ] Tests cover success, denied OAuth, and existing linked account.
Validation:
- npm test -- auth
- npm run lint
Tips
- The smaller the owned file set, the better the agent result.
- Make "do not edit" lists explicit for shared or risky modules.
- Prefer acceptance criteria over long implementation instructions.
- Treat validation as part of the task, not a follow-up.
Related Skills
taste-skill
116.8kAgent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
ai-job-search
72.1kThe job search that runs on your machine. AI job application framework built on Claude Code: evaluate postings, tailor CVs, write cover letters, prep interviews. Fork it and own it.
claude-howto
47.3kA visual, example-driven guide to Claude Code — from basic concepts to advanced agents, with copy-paste templates that bring immediate value.
lobehub
85.6k🤯 LobeHub - Super-Agents on Your Team. Add super-agents to the chats your team already uses.They take on work by themselves and get better every week.
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
