skill-stocktake

Audit Claude skills and commands using structured checklists and holistic judgment.

Updated Mar 16, 2026
One-click install
npx skills add https://github.com/Clean1ines/crm_bot --skill skill-stocktake-clean1ines
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-stocktake
Source: https://github.com/Clean1ines/crm_bot/tree/main/.agents/skills/ecc/skill-stocktake
Command: npx skills add https://github.com/Clean1ines/crm_bot --skill skill-stocktake-clean1ines

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq, and includes scripts (resource) components.

What problem does it solve?

Auditing Claude skills and commands for quality across a growing skill library, ensuring consistency and safety.

Core Features & Use Cases

  • Quick Scan mode re-evaluates only changed skills to deliver rapid quality feedback.
  • Full Stocktake mode performs a comprehensive audit of all skills, including changelog and usage context.
  • Phase-based evaluation helps maintain governance over a large collection of skills.

Quick Start

Analyze from the project root to inventory and evaluate both global and project-level skills.

Frequently Asked Questions about skill-stocktake

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I audit Claude skills and commands for quality at scale?

To audit Claude skills for quality, run a Full Stocktake to evaluate all skills using a structured checklist and holistic judgment, outputting per-skill verdicts to a results.json file.

Can I check only changed Claude skills instead of auditing everything?

Yes, you can check only changed Claude skills by running the Quick Scan mode, which uses diff scripts to re-evaluate modified skills and deliver rapid quality feedback without a full audit.

Do I need jq installed to run a skill stocktake audit?

Yes, you need jq installed to run the skill stocktake audit, as it is a required dependency for processing the JSON outputs and managing the inventory and evaluation phases.

What is the best way to maintain governance over a large collection of Claude skills?

The best way to maintain governance over a large collection of Claude skills is using phase-based evaluation, which inventories global and project-level directories and applies structured quality checklists.

How does the skill audit process read skill metadata during evaluation?

The skill audit process reads SKILL.md frontmatter during evaluation to extract the name and description of each skill, ensuring accurate identification when generating per-skill verdicts in results.json.

Does the audit script work with both global and project-level skill directories?

Yes, the audit script works with both global and project-level skill directories, scanning paths under ~/.claude/skills and the current project’s .claude/skills path to inventory and evaluate all available skills.