Identify dark-code risks and generate a structured comprehension audit report.
npx skills add https://github.com/az9713/dark-code-skills --skill dark-code-audit
Please help me install this Agent Skill. Skill: dark-code-audit Source: https://github.com/az9713/dark-code-skills/tree/main/skills/dark-code-audit Command: npx skills add https://github.com/az9713/dark-code-skills --skill dark-code-audit
The Dark Code Audit helps teams identify and measure comprehension gaps in codebases where AI-generated changes shipped without clear understanding. It surfaces hotspots, ownership gaps, and documentation deficits to prevent blind trust in automated velocity.
Run the audit from the repository root to generate a full comprehension audit report.
High-intent search queries and answers about installing and using this skill.
To audit comprehension debt, run a structured analysis from your repository root to auto-discover modules, map hotspots, and generate a formal report documenting ownership gaps and missing context layers.
Dark code refers to AI-generated changes shipped without clear understanding or documentation, which you identify by producing a structured comprehension audit that surfaces missing context layers and behavioral contracts.
You generate MODULE_MANIFEST.md and BEHAVIORAL_CONTRACTS.md artifacts by running an automated audit process that auto-discovers modules and creates context layers to document existing code behavior and structure.
Yes, a comprehension audit is specifically tailored for security reviews, compliance checks, architecture decisions, and post-incident investigations by producing governance-ready reports with a formal debt scorecard.
The best way to document risks before a major refactor is conducting a guided interview analysis that produces a structured hotspot map, ownership gap analysis, and a formal comprehension debt scorecard.
Yes, this audit specifically applies to codebases with multi-team modules and absent context layers, generating DECISION_LOG.md files and documenting ownership gaps to prevent blind trust in automated velocity.
spencerbeggsAudit CLAUDE.md files for quality, structure, and design-doc pointers.
bigeasyfreemanInterview architecture, AI tool usage, ownership, and deployment to assess dark-code risks.
eliottbusinessAutomate pre-commit code review with static analysis and security scanning.
dimsourGenerate XML documentation, READMEs, and OpenAPI descriptions from .NET source code.
