The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The skill adheres to the principle of least privilege, using only necessary file-reading tools to perform its intended function.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the user's codebase, README, and metadata. While it lacks delimiters (boundary markers) and sanitization for this input, it also lacks dangerous capabilities like network access, shell execution, or file writing that would allow an injection to cause system harm. Ingestion points: Codebase, README, and App Store metadata (via Read, Glob, Grep). Boundary markers: Absent. Capability inventory: Read, Glob, Grep (No network or shell access). Sanitization: Absent.

apple-featuring