reviewing
Pass
Audited by Gen Agent Trust Hub on May 11, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted document content (research, plans, brainstorms, and QA reports) from the filesystem and has the capability to write modifications or errata back to files. This is a standard risk for document-processing skills and is handled according to best practices.
- Ingestion points: Files located in
/research/,/plans/,/brainstorms/, or/qa/directories. - Boundary markers: None explicitly defined in the prompts to distinguish between document content and agent instructions.
- Capability inventory: Writing auto-fixes to documents, appending errata sections, and spawning the
codebase-analyzertool. - Sanitization: No explicit sanitization of the input document content is mentioned.
Audit Metadata