paw-ps-research-to-brief
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted research data which could contain malicious instructions designed to influence downstream synthesis or future agent interactions.
- Ingestion points: Reads
product-context.md,market-intelligence.md, andaudience-intelligence.mdfrom the product workspace. - Boundary markers: The synthesis process in
references/03-synthesize-brief.mddoes not use delimiters to isolate research content from the agent's instructions. - Capability inventory: The skill can read and write files within the
.pawbytes/prodig-suites/directory structure. - Sanitization: No content filtering or sanitization is applied to the research data before it is incorporated into the final product brief.
Audit Metadata