The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted user code files and possesses modification capabilities.\n
Ingestion points: The Audit workflow in SKILL.md scans files using Read, Glob, and Grep tools.\n
Boundary markers: The instructions lack delimiters or specific directives to ignore instructions found within code comments or string literals.\n
Capability inventory: The skill utilizes the Edit tool in Step 3 (Fix) to modify files based on its analysis.\n
Sanitization: There is no logic for sanitizing or validating file content before processing.

refactor