vc-audit-context

Audit project context routing, skill discoverability, and Claude/Codex wiring with validator scripts.

Updated Apr 9, 2026
One-click install
npx skills add https://github.com/HyuseCS/project-c --skill vc-audit-context-hyusecs
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: vc-audit-context
Source: https://github.com/HyuseCS/project-c/tree/main/.claude/skills/vc-audit-context
Command: npx skills add https://github.com/HyuseCS/project-c --skill vc-audit-context-hyusecs

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Project context documentation and shared skills drift over time: routers stop indexing files, skills become unreachable, and Claude/Codex agent surfaces fall out of sync. This Skill detects that drift by running a suite of Node.js validators against the context layer and skill inventory. ## Core Features & Use Cases - Context Discovery Validation: Verifies every file under process/context/ is indexed by the root router or a group entrypoint, and that concrete backticked references resolve to real files. - Skill Routing & Catalog Checks: Confirms each shared skill is routed from canonical surfaces or explicitly allowlisted, and regenerates a machine-readable skills catalog. - Cross-Reference & Dependency Analysis: Detects missing local references, skill dependency cycles, and confusable skill names. - Use Case: After reorganizing your project's documentation, run this audit to confirm no context file was orphaned, no skill lost its routing, and Claude and Codex agent definitions remain in parity. ## Quick Start Ask the AI to audit the project context routing and skill discoverability, then fix any validator failures it reports.

Frequently Asked Questions about vc-audit-context

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

FAQPage Schema
How do I audit project context routing and skill discoverability?

Run the validator scripts under .claude/skills/vc-audit-context/scripts/, starting with validate-context-discovery.mjs, then the skill routing, cross-reference, dependency, and keyword validators. Fix reported failures and re-run until they pass.

How do I check that Claude and Codex agents stay in sync?

The context discovery validator compares agent names between .claude/agents markdown files and .codex/agents TOML files, failing on any parity gap. It also checks critical hook files for content drift between the two surfaces.

What does the skill routing validator require for each shared skill?

Every skill under .claude/skills must be mentioned on a canonical routing surface declared in skill-routing-policy.json, such as AGENTS.md or CLAUDE.md, or be explicitly allowlisted in that policy with a written reason.

Why does the context discovery validator skip checks on a fresh clone?

It enters bare-kit mode when process/context/all-context.md is absent, which is expected in an unconfigured kit template. Per-project context-doc checks are skipped until the context layer is initialized.

What frontmatter must each SKILL.md file contain?

Each SKILL.md needs a name and description, plus a non-empty trigger_keywords scalar and a layer value of actor, contract, or helper. The validate-skill-keywords script enforces this and also verifies the generated skills catalog is in sync.