The Agent Skills Directory

[SAFE]: The skill instructions and guidelines are focused on performing static analysis and providing human-readable explanations of code logic. It explicitly discourages code execution for the purpose of explanation ('do not execute the target simply to explain it').
[SAFE]: The skill mentions external files for setup (e.g., _shared/common.md), but these are internal platform references for report formatting and style guidelines, not external code dependencies.
[SAFE]: There are no indicators of prompt injection, data exfiltration, obfuscation, or malicious persistence mechanisms. The skill operates within a 'read-only' posture regarding the user's code.

logic-explain