prd-to-plan
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted PRD documents to generate implementation plans that serve as instructions for downstream execution agents.\n
- Ingestion points: The skill reads PRD content from user input or local files (SKILL.md).\n
- Boundary markers: There are no explicit delimiters or instructions to ignore embedded prompts within the processed PRD content.\n
- Capability inventory: The skill has the capability to explore the local codebase and write Markdown files to the
./docs/directory.\n - Sanitization: No sanitization or verification of the PRD content is performed before generating the phased implementation tasks.\n
- Impact: Malicious instructions embedded in a PRD could influence the architectural decisions or specific tasks generated for the execution agent, potentially leading to unauthorized code changes.
Audit Metadata