The Agent Skills Directory

[SAFE]: The skill establishes a security audit persona with guidelines on identifying exploitable paths in code. The instructions are professional and focused on safety improvements.
[PROMPT_INJECTION]: The skill exhibits an inherent indirect prompt injection surface because it processes untrusted code diffs using capable tools.
Ingestion points: Code diffs and files are ingested via Read, Grep, and Glob tools.
Boundary markers: Absent; the instructions do not specify the use of delimiters to isolate untrusted code from the agent's system instructions.
Capability inventory: The agent is granted access to the Bash tool (shell execution) and file system access via Read, Grep, and Glob tools.
Sanitization: Absent; there are no instructions to sanitize or validate the content of the diffs before processing.

security-reviewer