skills.rest
product-on-purpose
product-on-purposeOFFICIAL
Explore all skills by this author →
GitHub

utility-pm-skill-auditor

Runs repo-wide governance audits aggregating validators, cross-cutting checks, and counter re-derivation.

640|81|Updated Jan 9, 2026
Search Tags:#pre-release#validators#repo health#governance audit#severity grading#sub-agent dispatch#counter re-derivation
One-click install
npx skills add https://github.com/product-on-purpose/pm-skills --skill utility-pm-skill-auditor
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: utility-pm-skill-auditor
Source: https://github.com/product-on-purpose/pm-skills/tree/main/skills/utility-pm-skill-auditor
Command: npx skills add https://github.com/product-on-purpose/pm-skills --skill utility-pm-skill-auditor

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Individual validators each check one rule, so cross-cutting repository issues like orphaned family contracts, sample gaps, stale declared counters, and skills missing matching commands slip through undetected before a release.

Core Features & Use Cases

  • Aggregated Validator Suite: Invokes the full enforcing validator bundle via scripts/pre-tag-validate.sh and consolidates results into one report.
  • Cross-Cutting Checks: Detects issues no single validator catches, such as dead skill references, sample gaps, and description collisions, graded P0 through P3.
  • Counter Re-Derivation: Re-counts skills, commands, sub-agents, and workflows from the filesystem and compares them against values declared in CONTEXT.md, AGENTS.md, and README.md.
  • Layered Output: Produces a full audit report, a prose Status Summary, and a machine-readable Status YAML block for downstream tooling like release gates.
  • Use Case: Before tagging a release, run a full-scope audit with a P1 severity floor to confirm zero P0 findings and verify that declared catalog counts match the actual repository state.

Quick Start

Ask the AI to run a full pre-release governance audit of the repository with a P1 severity floor and return the findings report with the machine-readable status block.

Frequently Asked Questions about utility-pm-skill-auditor

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

FAQPage Schema
How do I run a pre-release governance audit on a skills repository?▼

Invoke the audit with a scope argument such as full, changed, or since-tag, plus an optional severity floor like P1. It runs the validator bundle via scripts/pre-tag-validate.sh, executes cross-cutting checks, and returns a layered report with P0-P3 findings.

What does a repo-wide audit catch that individual validators miss?▼

Cross-cutting checks catch issues spanning multiple files: skills without matching commands, missing library samples, orphaned family contracts, dead references to deleted skills, and declared counters that drift from actual filesystem counts.

Does this audit work on AI clients other than Claude Code?▼

Yes. On Claude Code it dispatches to the native pm-skill-auditor sub-agent; on other clients like Codex CLI or Cursor it reads the canonical agent definition and executes the four-step audit flow inline, provided the client can read files and run Bash.

Can the auditor fix the issues it finds?▼

No. The auditor is detection-only; it reports findings graded P0 through P3 with recommended fixes, but remediation is left to maintainer judgment or separate tooling. Use a dedicated validation or iteration skill to apply fixes.

When should I use a single-skill validator instead of a repo-wide audit?▼

Use a single-skill validation tool when you want a deep check of one skill's frontmatter and structure against authoring conventions. The auditor goes wide across the entire catalog and does not go deep on any individual skill.

Related Skills

Explore More Product & Management Skills→
langgeniuslanggenius
175

dify-docs-env-vars

Analyze Dify documentation repositories to generate metadata for skill units.

Official
Advanced
Lionad-MorotarLionad-Morotar
7

release-project

Guide software releases from planning to semantic versioning and Git tagging.

Community
Advanced
MuvonMuvon
3

content-audit

Audit drafted content against a canonical rubric with per-axis and overall scores.

Official
Advanced
chiragborse1chiragborse1

kova-testing

Select efficient verification workflows for Kova code changes using typecheck, lint, guard, or Vitest lanes.

Community
Intermediate