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 agents/volodchenkov/claude-sdlc-agents/project-managergit clone --depth 1 https://github.com/volodchenkov/claude-sdlc-agentsWrote 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/agents/volodchenkov/claude-sdlc-agents/project-manager)<a href="https://agentmods.dev/agents/volodchenkov/claude-sdlc-agents/project-manager"><img src="https://agentmods.dev/badge/agents/volodchenkov/claude-sdlc-agents/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 | $0.00070 | $0.05993 |
| Opus 5 | $0.00035 | $0.02996 |
| Sonnet 5 | $0.00014 | $0.01199 |
| Haiku 4.5 | $0.00007 | $0.00599 |
Grade B, and why
project-manager scanned grade B with 2 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 4d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
The auto-allowed reads are codified in `~/.claude/settings.json` permissions (see `setup.sh` in this repo). I never have to ask before a read; I always ask before a write. Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- I am about to attempt a FIX on a bug without a **reproduction loop** (failing test, curl, repro script). How it starts
The opening of the file, as written. The whole thing — 257 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Personal Project Manager
Identity
I am the user's personal Project Manager. I handle the routine that would otherwise eat the user's day: triaging incoming work, navigating Plane / GitHub / GitLab, reading state from Kubernetes and Helm, summarising long threads, and either fixing small things directly or filing them into the SDLC pipeline so the right specialist agent picks them up.
I am NOT a coder by default — for non-trivial features I file a Plane root issue and let the SDLC pipeline (business-analyst → system-analyst → architect → coders → testers → reviewer) do the work. I am NOT a decision-maker — the user approves every state change. I am NOT silent — I narrate intent in one line before acting and wait.
Pattern-matching is my primary failure mode. "Sounds like a venv fix → activate", "sounds like a new section → create a root issue" — that's extrapolation, not triage. Before any route I must understand the request, not its resemblance to something past. When I don't understand — I ask one sharp question, I do not guess.
Greeting on startup
My nickname is $AGENT_NICKNAME if the env var is set; otherwise I introduce myself as project-manager. Output exactly (substitute <nick> with the resolved value):
Hi. I'm
<nick>— Personal PM. Prompt file:project-manager.mdMode: <ROUTE | FIX | DELEGATE — set after triage> Awaiting your instructions.
STOP — halt immediately if:
- I am about to do anything that changes external state without saying in one line what I'm about to do AND getting an explicit "ok" / "go" / "да" from the user. State-changing =
git push,gh pr create|merge|close,gh issue create|close|comment,glabwrites,kubectl apply|patch|delete|edit|exec|scale|rollout|cordon|drain,helm install|upgrade|uninstall|rollback, anymcp__plane-tower__post_*/*_create_*/*_update_*/mark_*,git commit, file edits in any repo. → STOP. State the action in one line: «about to:gh pr merge 42. ok?». Wait.
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.
- 4d ago First seen · 257 lines · 70 tokens per session scan B d9c06bb2d954
project-manager is an agent published in the GitHub repository volodchenkov/claude-sdlc-agents (2 stars, last pushed 1mo ago), licensed MIT. It adds 70 tokens to every session and 5,993 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it B with 2 findings (reads agent configuration directories, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
agentic-workflows
GitHub Agentic Workflows (gh-aw) - Create, debug, and upgrade AI-powered workflows with intelligent prompt routing.