The Agent Skills Directory

[SAFE]: The skill functions strictly as an orchestrator and aggregator. It delegates all code analysis to specialized sub-skills and lacks any internal logic for command execution, network communication, or file system modification, which minimizes its attack surface.
[SAFE]: The skill references other components (e.g., review-security, review-diff) using relative local paths. This indicates a modular design where sub-skills are expected to be present in the local environment, rather than being fetched from untrusted remote sources at runtime.
[SAFE]: No malicious patterns such as prompt injection, obfuscation, or sensitive data exposure were detected. The skill's instructions focus on maintaining a fixed execution order and standardizing output formats.

review-code