The Agent Skills Directory

[SAFE]: The skill is entirely instructional, providing a diagnostic rubric for an AI agent to evaluate code quality, accessibility, and design fidelity. It does not perform any active operations such as file writing, network requests, or command execution.
[SAFE]: References to external repositories from Anthropic and established open-source contributors are used solely for licensing attribution and context, and do not involve any remote code downloads or execution.
[SAFE]: The skill explicitly instructs the agent to remain in a read-only diagnostic mode and forbids fixing issues or making changes during the audit process.

tt-audit