know-your-unknowns

Pass

Audited by Gen Agent Trust Hub on Aug 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions and reference materials are focused on improving the quality of implementation plans and user requirements. No prompt injection or behavior override patterns were found.
  • [SAFE]: No evidence of data exfiltration or hardcoded credentials was found. The skill directs the agent to scan local codebases for analysis, which is appropriate for its intended purpose.
  • [SAFE]: The skill does not involve any external downloads, remote script execution, or third-party dependencies from untrusted sources.
  • [SAFE]: The project files included in the evaluation directory (scanledger) are standard Go source files for a benign text-scanning utility.
  • [SAFE]: No obfuscation techniques, hidden URLs, or dynamic execution patterns were identified in any of the provided files.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 7, 2026, 03:01 AM
Security Audit — agent-trust-hub — know-your-unknowns