The Agent Skills Directory

[SAFE]: The skill uses a principle of least privilege by restricting shell access to git commands via the Bash tool configuration. It primarily relies on standard file-reading and search operations (Read, Grep, Glob) to analyze codebase structure.
[SAFE]: Mandatory interactive checkpoints are included (AskUserQuestion), requiring the agent to present its understanding and obtain user validation before proceeding to recommendations or code modifications.
[SAFE]: No patterns of prompt injection, data exfiltration, or malicious persistence were detected. The instructions focus entirely on conceptual code analysis and developer-approved refactoring.

concept-review