adversarial-review
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is purely instructional and defines a structured review process without any executable scripts or external dependencies.
- [SAFE]: The instructions explicitly limit the agent to a read-only role, prohibiting file modifications or the creation of new workers, which prevents unauthorized actions.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for analyzing external plans and briefs, creating a data ingestion surface.
- Ingestion points: The skill ingests briefs, plans, investigations, and delivery proposals.
- Boundary markers: No specific boundary markers or delimiters are provided for the ingested content.
- Capability inventory: The skill is strictly read-only and lacks capabilities for file writing, network communication, or command execution.
- Sanitization: No input sanitization or filtering logic is specified.
- Conclusion: The lack of exploitable capabilities mitigates the risks associated with processing untrusted data.
Audit Metadata