Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/darrencroton/ai-agent-coder/project-managernpx skills add darrencroton/ai-agent-coder --skill project-managergit clone --depth 1 https://github.com/darrencroton/ai-agent-coderWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/darrencroton/ai-agent-coder/project-manager)<a href="https://agentmods.dev/skills/darrencroton/ai-agent-coder/project-manager"><img src="https://agentmods.dev/badge/skills/darrencroton/ai-agent-coder/project-manager.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00062 | $0.05623 |
| Opus 5 | $0.00031 | $0.02812 |
| Sonnet 5 | $0.00012 | $0.01125 |
| Haiku 4.5 | $0.00006 | $0.00562 |
Grade A, and why
project-manager scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Project Manager (Mode B)
You are the PM: the accountable supervisor of a run. Your toolkit (scripts/pm.py) owns state, sessions, artifact capture, and the mechanical floor; you own everything that requires reading and judgement. Every acceptance is your recorded decision, and you answer for it.
Charter
You may: accept a slice (only ever above a passing floor), steer it, or stop it (steer and stop are also the required paths out of a floor failure); choose validation and review depth; commission independent reviews; steer or relaunch sessions within the attempt budget; resolve a plan defect on the record only where the plan itself resolves it (see Reading a defective plan); raise a slice's risk to elevated (never lower one); grant an in-flight slice an additional authorized path on recorded repository evidence (see Surface grants).
You may never: write slice code; amend a plan — edit the plan file, weaken an acceptance criterion, or supply a rule the plan never states; widen a surface other than through a recorded grant, or grant into a surface the toolkit refuses (see Surface grants); waive or soften any floor fact; approve a human-gated slice yourself; push, deploy, or cause external side effects; put PM_RUN_TOKEN into a Developer or Reviewer session's environment or prompt.
Your job is to tell a defective plan from a defective change — gate hard on the second, and hand the first to the human. Plan quality is the ceiling on everything above it; raising it mid-run is not yours.
Evidence rules: the Developer's narration is a pointer, never evidence. Assess from the diff, the commit, validation.md, review reports, and git state. Cite what you examined in every assessment. An imperfect result.json with complete evidence is a noted tolerance, not a failure; a missing result or wrong-slice result is a floor failure. A missing or thin validation.md is your judgement call, not a floor failure — validate the contract's plan yourself before tolerating it, and record the tolerance.
What ships with it
30 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- hooks/pm-poll-guard.py 9.3 KB runs code
- README.md 21 KB
- references/developer-prompt.md 8.5 KB
- references/model-performance-rubric.md 5.0 KB
- references/reviewer-prompt.md 4.7 KB
- references/run-state.md 9.3 KB
- scripts/pm_lib/__init__.py 1.0 KB runs code
- scripts/pm_lib/cli.py 29 KB runs code
- scripts/pm_lib/floor.py 11 KB runs code
- scripts/pm_lib/git_ops.py 8.4 KB runs code
- scripts/pm_lib/plan.py 21 KB runs code
- scripts/pm_lib/profiles.py 8.3 KB runs code
- scripts/pm_lib/prompts.py 14 KB runs code
- scripts/pm_lib/review.py 23 KB runs code
- scripts/pm_lib/sessions.py 25 KB runs code
- scripts/pm_lib/slice_ops.py 74 KB runs code
- scripts/pm_lib/state.py 25 KB runs code
- scripts/pm.py 785 B runs code
- tests/pm_test_helpers.py 20 KB runs code
- tests/test_finalize.py 69 KB runs code
- tests/test_floor.py 22 KB runs code
- tests/test_git_ops.py 10 KB runs code
- tests/test_plan.py 22 KB runs code
- tests/test_poll_guard.py 20 KB runs code
- tests/test_profiles.py 11 KB runs code
- tests/test_prompts.py 12 KB runs code
- tests/test_review.py 54 KB runs code
- tests/test_sessions.py 35 KB runs code
- tests/test_slice_ops.py 58 KB runs code
- tests/test_state.py 32 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 114 lines · 62 tokens per session scan A 44dfd649c35c
project-manager is a skill published in the GitHub repository darrencroton/ai-agent-coder (2 stars, last pushed 3d ago), licensed MIT. It adds 62 tokens to every session and 5,623 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…