recipe-design
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the processing of external requirements and documentation which may contain untrusted content.
- Ingestion points: Untrusted data is ingested via the
$ARGUMENTSparameter inSKILL.md(labeled as requirements) and via theprd_pathpointing to external PRD files. - Boundary markers: The orchestration logic does not specify the use of clear delimiters or instructions to the subagents to ignore potential instructions embedded within the requirements or codebase evidence.
- Capability inventory: The orchestrated workflow has the capability to write to the repository through the
dev-workflows:technical-designersubagent. - Sanitization: There is no evidence of sanitization or validation of the requirement strings before they are passed to subagent prompts.
Audit Metadata