control-audit
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a robust and safe auditing methodology. It provides detailed guidance for mapping evidence to security controls without performing dangerous operations.
- [SAFE]: The skill explicitly prohibits the capture of sensitive data, such as credentials or personal information, in evidence files, recommending redaction as a standard practice.
- [SAFE]: The instructions require the agent to obtain explicit user authorization before interacting with any live systems or APIs (e.g., GitHub API, Cloud CLIs), ensuring user oversight.
- [SAFE]: The included Node.js utility scripts (
validate-findings.mjsandto-sarif.mjs) are transparent tools for data validation and format conversion. They use standard libraries for file I/O and do not perform network operations or execute dynamic code. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest external data (policies, tickets, repo exports). While this presents an attack surface, the skill provides clear boundary instructions, structured output requirements (JSON schema), and sanitization guidelines to mitigate the risk of accidental obedience to instructions embedded in processed data.
Audit Metadata