grill-with-docs
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest user plans and designs to produce documentation like ADRs and glossaries. This process creates a surface for indirect prompt injection where adversarial content in the user's plan could potentially influence the documentation generation or the agent's behavior.
- Ingestion points: User-supplied plans or design descriptions mentioned in the description (SKILL.md).
- Boundary markers: No explicit delimiters or instructions are provided to the agent to treat input data as untrusted.
- Capability inventory: The skill description indicates an intent to modify the file system by writing to
CONTEXT.mdand ADR files. - Sanitization: There is no evidence of input validation, escaping, or filtering for content processed during the 'grilling' or 'domain-modeling' phases.
Audit Metadata