plan
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests repository context and conversation history which could contain malicious instructions. The risk is mitigated by explicit instructions to stay in a read-only planning mode and only write to the .hermes/plans/ directory.
- Ingestion points: Workspace files and conversation context.
- Boundary markers: Absent.
- Capability inventory: write_file (limited to plans directory) and read-only inspection.
- Sanitization: Absent.
Audit Metadata