anti-sycophancy
Pass
Audited by Gen Agent Trust Hub on Jun 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of natural language instructions designed to modify the agent's response behavior. It does not contain any executable code, scripts, or automated tool invocations.
- [SAFE]: There are no network operations, remote downloads, or external dependencies. The skill operates purely within the logic of the prompt context.
- [SAFE]: No sensitive file access, credential usage, or persistence mechanisms are present. The installation instructions provided in the README are intended for manual user execution and do not involve the agent performing privileged actions.
- [SAFE]: The 'anti-sycophancy' logic, which requires independent assessment of user claims, may actually serve as a protective layer against certain forms of prompt injection or social engineering.
Audit Metadata