decide-about-disagreements

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze 'models-consensus' results, which are generated by other AI models. This introduces a surface for indirect prompt injection attacks.
  • Ingestion points: The skill reads council reports, state objects, and raw round files as described in the 'Read the result' section of SKILL.md.
  • Boundary markers: There are no explicit instructions to use delimiters or ignore embedded instructions within the ingested data.
  • Capability inventory: The skill uses file reading capabilities to access consensus artifacts and interactive tool calls (AskUserQuestion, request_user_input) to communicate with the user.
  • Sanitization: No sanitization or validation of the ingested model outputs is mentioned in the instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 12:27 AM
Security Audit — agent-trust-hub — decide-about-disagreements