The Agent Skills Directory

[SAFE]: The skill performs static analysis using Python's built-in ast module and tree-sitter via the optional gauntlet plugin. It identifies performance issues (like nested loops over the same iterable or list lookups in loops) without executing the code under analysis.
[SAFE]: Environment interaction is restricted to standard development workflows, such as using git diff to determine which files to scan and executing local verification scripts for citation accuracy.
[SAFE]: The skill mentions external tool installation (cargo install kuva) for the purpose of rendering benchmark charts. This is a documented, legitimate use of a public package registry to support the skill's visualization features.
[SAFE]: The skill uses a modular structure with clear entry points and verification steps, maintaining a human-in-the-loop approach for applying performance fixes.

performance-review