prd-to-design
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from
plans/<name>/prd.md, which creates an attack surface where hidden instructions within the PRD could influence the agent's design decisions or output. - Ingestion points: The agent is instructed to read
plans/<name>/prd.mdin its entirety during the discovery phase (Phase 1). - Boundary markers: The instructions do not define delimiters or specific warnings to ignore instructions that might be present in the requirement document's text.
- Capability inventory: The skill possesses the capability to write multiple files to the file system (
design.mdand ADR files in theadr/directory). - Sanitization: There is no evidence of sanitization, validation, or filtering of the content read from the PRD before it is used to generate the architecture design.
Audit Metadata