darwin

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill provides conceptual guidance for system evaluation and architectural improvements. Analysis of the instructions reveals no malicious patterns, prompt injections, or attempts to bypass safety filters.\n- [NO_CODE]: The skill is purely instructional and contains no executable scripts, commands, or automation logic.\n- [EXTERNAL_DOWNLOADS]: No external URLs or remote resources are referenced or downloaded in the skill definition.\n- [DATA_EXFILTRATION]: There are no mechanisms for accessing sensitive local files or transmitting data to external servers.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 04:54 AM
Security Audit — agent-trust-hub — darwin