true-seeing

Pass

Audited by Gen Agent Trust Hub on Jun 27, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is comprised entirely of natural language instructions and metadata. It does not include any scripts, binaries, or command-line execution patterns.
  • [SAFE]: No network operations, data exfiltration vectors, or sensitive file access patterns were detected. The skill explicitly states it has no extra runtime dependencies.
  • [SAFE]: There are no instances of obfuscation, hidden content, or malicious prompt injection patterns. The instructions encourage the agent to adhere to evidence-based reporting and acknowledge uncertainties.
  • [SAFE]: The metadata and author information appear consistent with the skill's stated purpose as a metaphorical tool for analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 27, 2026, 07:46 PM
Security Audit — agent-trust-hub — true-seeing