code-to-prd
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill performs local static analysis using standard library Python scripts. No malicious behavior such as network exfiltration, persistence, or credential theft was detected.
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes user-provided source code.
- Ingestion points: The agent reads files from a project directory during the analysis phases defined in SKILL.md.
- Boundary markers: The instructions lack explicit boundary markers or directions to ignore embedded instructions in the source data.
- Capability inventory: The skill facilitates local file reading and writing (PRD generation) without network access or dynamic execution.
- Sanitization: No sanitization is performed on the ingested code content before it is processed by the agent.
Audit Metadata