compass
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by scanning project manifest files and documentation to build a local profile.
- Ingestion points: Manifest files and project-specific notes like CLAUDE.md (references/cache-recipes.md Section 4).
- Boundary markers: Absent. The skill does not currently use specific delimiters to isolate ingested file content.
- Capability inventory: Restricted to local file system read and write operations within the project directory.
- Sanitization: Not specified for scanned repository data.
- [COMMAND_EXECUTION]: Standard shell commands such as 'find' and 'wc' are used for repository sizing during the cache initialization process (references/cache-recipes.md Section 7).
Audit Metadata