how-to-write-rulebooks
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- [GENERAL] (SAFE): The content is purely instructional and provides a template for capturing operational knowledge. No malicious logic, obfuscation, or unauthorized access patterns were identified.- [INDIRECT PROMPT INJECTION] (LOW): While the skill encourages agents to incorporate findings from 'web research' into rulebooks (an ingestion point for untrusted data), it does not provide mechanism for automated execution or exfiltration of this data. The risk is limited to the quality and safety of the generated text content. (Evidence: File: SKILL.md, Ingestion: 'web research', Boundary: Absent, Capability: CLI documentation only, Sanitization: Absent).
Audit Metadata