grill-with-prd
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from PRD files, which introduces an attack surface for instructions that could be embedded within the processed documents.
- Ingestion points: The skill reads the content of PRD files located at
docs/prd/{feature-name}/prd.md, as described inSKILL.md. - Boundary markers: The instructions in
SKILL.mddo not define delimiters or provide specific instructions to the model to ignore instructions embedded within the documents being reviewed. - Capability inventory: According to
SKILL.md, the skill is permitted to explore the codebase (read access) and update the PRD files (file-write access). - Sanitization: There is no evidence in the skill files that the content from the PRD is sanitized, filtered, or validated before being incorporated into the agent's context.
Audit Metadata