li-prd
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input (responses to six product-related questions) and interpolates this content directly into a generated markdown PRD template. This content is then stored and potentially processed by other review personas within the skill, creating a surface for indirect prompt injection.
- Ingestion points: User answers to questions regarding 'Why Now', 'Target User', 'Status Quo', 'Success Criteria', 'Out of Scope', and 'MVP' in
SKILL.md. - Boundary markers: The skill does not use specific delimiters or instructions to ignore embedded commands within the interpolated user content.
- Capability inventory: The skill possesses file-writing capabilities, suggesting saving the PRD to a path derived from user-provided product names in
SKILL.md. - Sanitization: No input validation, escaping, or filtering of user-provided content is performed before it is added to the PRD or used in file paths.
Audit Metadata