user-research
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data which creates an indirect prompt injection attack surface.
- Ingestion points: Untrusted data enters the agent context through the
WebFetchtool when fetching targets for heuristic evaluation or accessibility auditing. - Boundary markers: The skill instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings for content retrieved from external URLs.
- Capability inventory: The skill has access to
Write,Edit, andTaskUpdatetools, which could potentially be abused if the agent follows instructions hidden within fetched research targets. - Sanitization: There is no evidence of content sanitization, filtering, or validation for data ingested from remote interfaces before it is processed for findings and recommendations.
Audit Metadata