govern-ai-product-risk

Pass

Audited by Gen Agent Trust Hub on Oct 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external data sources including customer evidence, surveys, project documents, and prior incidents when 'BuildBetter MCP' is active. This creates a surface for indirect prompt injection if the processed content contains hidden instructions intended to override the agent's behavior. The skill includes a mitigation strategy by instructing the agent to 'Respect access boundaries and do not expose sensitive evidence in the risk artifact'.
  • [SAFE]: No instances of prompt injection, data exfiltration, obfuscation, or unauthorized command execution were detected. The external URLs provided in the provenance section are legitimate links to public repositories for educational and reference purposes. The skill's functionality aligns with the stated purpose of AI risk governance.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 27, 2026, 03:14 AM