review-team
Pass
Audited by Gen Agent Trust Hub on Aug 1, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) by processing untrusted data inputs during its review workflow.
- Ingestion points: The skill reads SDD artifacts and subagent output files (persona JSON records) located in the
.aidoc/review/directory. - Boundary markers: The skill instructs subagents to disregard specific embedded data like self-assessment scores, but does not define strict delimiters for the entire untrusted artifact content.
- Capability inventory: The orchestrator dispatches subagent tasks and performs file system operations within the local project-runtime environment.
- Sanitization: The skill documentation specifies a policy where subagents must treat inputs as untrusted data and strictly conform to a structured findings schema.
Audit Metadata