The Agent Skills Directory

[SAFE]: The skill is designed for codebase interrogation and architectural analysis. All subagents spawned for exploration, synthesis, and critique are explicitly configured with the readonly: true flag, preventing unauthorized file system modifications or process execution.
[SAFE]: No data exfiltration or external network requests were detected. The skill uses local exploration tools (Glob, Grep, Read) to gather architectural information which is then synthesized for the user.
[SAFE]: No obfuscation, persistence mechanisms, or privilege escalation patterns were found. The skill follows established patterns for AI-driven code analysis within the development environment.

how