The Agent Skills Directory

[SAFE]: The skill implements a structured workflow for auditing AI-generated work. It relies on standard system utilities like ls, grep, find, and git to inspect the local filesystem for accuracy and consistency.
[SAFE]: The Python execution is limited to a one-line script for counting backticks to ensure balanced markdown blocks, which is a benign local operation.
[SAFE]: No external network requests, data exfiltration attempts, or hardcoded credentials were found. The skill operates entirely within the local development environment context.
[SAFE]: The instructions encourage the agent to cross-reference work against project-specific rule files (e.g., CLAUDE.md, AGENTS.md), promoting best practices and safety adherence.

review-qa