agentsclimarketplace

Check

Skill christophecapel/claude-mechanisms-tools/skills/check

Session-close audit — session-scoped manifest, two-section open items, decisive close signal. Runs the `/check` audit defined in `check.md`. Use when the user says "check", "audit this session", "close the session", "definition of done", or wants a clean-vs-blocked verdict before ending a session.From its SKILL.md

Install
npx -y skills add christophecapel/claude-mechanisms-tools --skill check

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
  • runs commandsInstructs the agent to run 1 command, including `gh pr view`.

SKILL.md

2.3 KB, 496 tokens by cl100k_base, as published. Nobody here has run it

/check — Session-Close Audit

This skill runs the session-close audit defined in check.md.

What it does

  • Builds a Session Manifest from chat-context (PRs, commits, tickets, memory files, edited files) — derives scope FIRST, before any audit
  • Walks 5 audit sections (Learnings, Changes, Docs, Issues, Memory) with a per-section emoji status (✅ / ⚠️ / ❌) reflecting only Section A items
  • Routes open items into two sections:
    • Section A — In-session (blocks close if any row is tackle now)
    • Section B — Adjacent context (never blocks close, informational only)
  • Pre-render check: validates PR-merge dependencies before flagging rows as blocked by user. If gh pr view shows the PR is already merged, the row auto-flips to tackle now and the dependent action gets executed in the same turn.
  • Emits a decisive verdict: ✅ Ready to close / ❌ Not ready / ⚠️ Close with caveats. Re-running /check for the same state produces the same verdict.

Why session-scoped

Earlier /check versions surfaced every open item across all sessions, projects, and pre-existing tracker tickets. That made every close run feel incomplete and triggered multi-/check loops to walk through items that didn't belong to the session. Session-scoped audit runs in one pass — items from other sessions land in Section B and never block close.

Usage

Type /check at any point you want a clean-vs-blocked verdict on the current session. The skill follows check.md step by step.

Implements mechanisms

What ships with it: 1 file

7.9 KB alongside SKILL.md

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.