The Agent Skills Directory

[SAFE]: The skill contains no executable code, scripts, or network exfiltration patterns. It functions as a documentation and template provider for compliance tasks.
[COMMAND_EXECUTION]: The skill specifies the use of Read, Glob, and Grep tools. These are used appropriately to search local project files for legal and regulatory context during a review.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes untrusted user descriptions of initiatives via the $ARGUMENTS parameter.
Ingestion points: User-provided initiative descriptions in SKILL.md.
Boundary markers: None identified.
Capability inventory: Read, Glob, and Grep tools (SKILL.md).
Sanitization: No input sanitization is present.
Risk assessment: Despite the absence of delimiters, the risk is minimal because the toolset is restricted to local file reading and does not include shell execution or network capabilities.

compliance-check