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/hunterg325/claude-code-pm/pm-startnpx skills add hunterg325/claude-code-pm --skill pm-startgit clone --depth 1 https://github.com/hunterg325/claude-code-pmWrote 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/hunterg325/claude-code-pm/pm-start)<a href="https://agentmods.dev/skills/hunterg325/claude-code-pm/pm-start"><img src="https://agentmods.dev/badge/skills/hunterg325/claude-code-pm/pm-start.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 | $0.00032 | $0.00836 |
| Opus 5 | $0.00016 | $0.00418 |
| Sonnet 5 | $0.00006 | $0.00167 |
| Haiku 4.5 | $0.00003 | $0.00084 |
Grade A, and why
pm-start scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/pm:start — Launch PM Orchestration
Pre-flight Checks
Run these checks before starting. Surface clear errors for any failures:
- claude-peers broker: Call
list_peers— if it fails, tell user to start the claude-peers broker (cd ~/claude-peers && bun run src/index.ts) - Linear MCP: Call
list_issueswith a limit of 1 — if it fails, tell user to configure Linear MCP and link to https://docs.claude.ai/integrations/linear - gh CLI: Run
gh auth status— if it fails, tell user to rungh auth login - pm-config.yaml: Read
.claude/pm-config.yaml— if missing, offer to create it from${CLAUDE_PLUGIN_ROOT}/templates/pm-config.example.yaml - target_repo: Validate that
target_repois set in pm-config.yaml and the path exists and is a git repo
If any check fails, stop and report all failures at once. Do not proceed.
Startup Sequence
- Read
.claude/pm-config.yamland validate required fields (target_repo,linear.project) - Set your peer summary via
set_summary: "PM Orchestrator — managing {project_name}" - Check for
.claude/pm-state.json— if exists, enter crash recovery mode:- Reconcile against
list_peers,git worktree list, Linear statuses - Report recovered state to user
- Reconcile against
- If
--dry-runargument provided: fetch tasks, show what would be assigned, then stop
Spawn Worker Pool
The PM does NOT spawn workers itself. Workers are interactive Claude Code sessions running in visible Terminal.app tabs so the user can observe them.
Run the spawn script:
bash ${CLAUDE_PLUGIN_ROOT}/scripts/spawn-workers.sh "{target_repo}" {max_concurrent}
This opens max_concurrent Terminal tabs, each running:
cd {target_repo} && claude --dangerously-skip-permissions --dangerously-load-development-channels server:claude-peers
Workers start idle and wait for task assignments via claude-peers messages.
After spawning, wait ~15 seconds for workers to register, then verify with list_peers.
Enter Orchestration Loop
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.
- 3d ago First seen · 75 lines · 32 tokens per session scan A 10e28d6fe965
pm-start is a skill published in the GitHub repository hunterg325/claude-code-pm (2 stars, last pushed 5mo ago), licensed MIT. It adds 32 tokens to every session and 836 once invoked, about $0.0002 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
run
Conduct a task end to end, from a one-line fix to a multi-day programme - a short interview, then autonomous execution to the finish. Classifies the task S/M/L/XL, configures itself to the repo, routes each role to a cost tier, drives the agent team from the kanban board, and stops only at genuine decisions. Use for…
status
Where is the work right now. Regenerates the projections and the kanban board from the journal and answers with the progress line, the agents still running, the open gates and the resume steps. Read-only. Use for /tyran:status or when asked what is going on.
heartbeat
This skill should be used when the user asks to "run a heartbeat", "run the agent loop", "process GitHub issues", "check for work", or runs the /heartbeat command. Executes the WoterClip heartbeat — picks up GitHub issues, resolves personas, does work, and reports back.
woterclip-init
This skill should be used when the user asks to "initialize woterclip", "set up woterclip", "woterclip init", "configure woterclip for this repo", or runs the /woterclip-init command. Scaffolds a repo with WoterClip config, persona directories, and GitHub labels.
coordinator
Coordinates a fleet of autonomous background Claude Code sessions that implement tasks end to end — TDD, docs, PR, adversarial review, production deploy, tagging and post-deploy manual verification — while the human talks only to the coordinator. Use it WHENEVER the user asks to "work several tasks in parallel"…
unleash
Run as a named worker draining the task queue in my private coordination repo, where each task is a GitHub Issue — claim one task at a time, check blocked-by dependencies, plan with explicit assumptions, execute, validate against acceptance criteria, and record results as comments; then stay in ambush behind a…