AugmentClaude

Skill Stocktake

Audit AGF's internal skills for content quality and relevance.

Installation

  1. Make sure Claude is on your device and in your terminal.

    Skills load from ~/.claude/skills/ when Claude Code starts up — so you need it on your machine first. If you don't have it yet, install it once with the command below, then run claude in any terminal to verify.

    One-time setup
    npm i -g @anthropic-ai/claude-code

    Already have it? Skip ahead.

  2. Paste into Claude Code or into your terminal.

    This copies the whole skill folder into ~/.claude/skills/agf-skill-stocktake-pcliangx/ — the SKILL.md plus any scripts, reference docs, or templates the skill ships with. Safe default: works for every skill.

    Faster alternative (instruction-only skills)

    Skips the clone and grabs only the SKILL.md file. Don't use this if the skill ships Python scripts, reference markdowns, or asset templates — they won't be downloaded and the skill will fail when it tries to load them.

    Quick install (SKILL.md only)
    Sign up to copy
  3. Restart Claude Code.

    Quit and reopen Claude Code (or any other agent that loads from ~/.claude/skills/). New skills are picked up on startup.

  4. Just ask Claude.

    Skills auto-activate when your request matches the skill's description — no slash command needed. Trigger phrases live in the skill's own frontmatter; you can read them in the “What this skill does” section above.

Prefer to read the source first? Open on GitHub.

When Claude uses it

Use when PL / tech-lead audits the content quality of AGF's own skills (`.claude/skills/agf-*/SKILL.md`). Provides the 5-dimension checklist (Actionability / Scope fit / Uniqueness vs CLAUDE.md+standards / Currency of pointers / Gate-binding validity), 5-verdict rubric (Keep/Improve/Update/Retire/Merge), reason-quality bar , and output skeleton. Triggered each MINOR release retro or manually.

What this skill does

Skill Stocktake helps product leads and tech leads audit AGF's internal governance skills for content quality and relevance. It runs during minor release retrospectives or manually when the skill collection shows signs of decay—duplicated scope, stale references, or unclear purpose. The audit applies a five-dimension checklist and a five-verdict rubric to keep the skill inventory actionable, focused, and free of overlap.

What it does

  • Audits five dimensions per skill: Actionability (executable steps, not vague principles), Scope fit (name and trigger aligned), Uniqueness (no duplication with CLAUDE.md, standards, or peer skills), Currency (referenced paths in standards/, docs/adr/ still exist and synced), and Gate binding validity (skill remains tied to a governed process or gate command).
  • Issues five verdicts: Keep (all five dimensions pass), Improve (worth retaining but needs specific fixes), Update (references are stale; verification via rg or search required), Retire (low quality, outdated, or cost-asymmetric; names what replaces it), or Merge into [X] (overlaps with another skill; specifies target and content to integrate).
  • Enforces reason quality: Rejects hollow statements like "unchanged" or "overlaps"; requires self-contained, decidable explanations pointing to concrete defects, coverage gaps, or integration targets.
  • Generates structured output: Writes docs/reviews/skill-stocktake-<YYYY-MM-DD>.md with frontmatter, summary table, grouped details, and action items for Improve/Update/Retire/Merge verdicts.

How to use it

Run this skill when a PL or tech lead manually inventories AGF's governance skills (.claude/skills/agf-*/SKILL.md), or as an optional step in a minor release retrospective. Read each skill's body, apply the five-dimension checklist—paying special attention to Currency by verifying referenced paths exist—fill in the verdict and detailed reason, then output the audit report. Retire and Merge decisions require PL confirmation before deletion or consolidation; Improve and Update are assigned to maintainers for execution.

Related skills