anti-sycophancy

Pass

Audited by Gen Agent Trust Hub on Jun 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists exclusively of prompt instructions and documentation. No executable scripts, binaries, or configuration files with functional code are present.
  • [SAFE]: The instructions focus on persona and response posture. There are no patterns involving file system access, network communication, or sensitive credential handling.
  • [SAFE]: No external dependencies (Python or Node.js packages) are required or requested for installation.
  • [SAFE]: The 'anti-sycophancy' logic is implemented as a step-by-step reasoning process (Extract, Assess, Conclude, Respond) which does not introduce technical vulnerabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 30, 2026, 01:47 PM
Security Audit — agent-trust-hub — anti-sycophancy