affirmations
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill provides instructional guidance for agent self-correction. The affirmations and workflow are designed to help the agent break out of repetitive failure states and do not bypass security controls.
- [NO_CODE]: The skill is composed of markdown and metadata, with no included scripts, binaries, or automated runtime code execution.
- [EXTERNAL_DOWNLOADS]: The documentation mentions a related skill 'inference-sh/skills@prompt-engineering' that can be added via a CLI tool. This is a manual reference for the user or agent and does not occur automatically during the skill's normal operation.
Audit Metadata