validate-answer
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill defines a logical process for agent review and consensus-building. It does not contain executable scripts, network requests, or attempts to access sensitive system files.- [INDIRECT_PROMPT_INJECTION]: The skill processes external artifact data, which represents a potential surface for indirect prompt injection attacks.
- Ingestion points: The skill reads content from artifact paths during the Step 1 dispatch phase.
- Boundary markers: The instructions do not include specific delimiters or 'ignore' instructions for the artifact content.
- Capability inventory: The skill utilizes interactive user questions and the ability to chain to other skills through the Skill tool.
- Sanitization: No specific validation or sanitization routines for the processed data are defined.
Audit Metadata