skill-judge
Pass
Audited by Gen Agent Trust Hub on Mar 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely composed of Markdown-based instructional content. It does not perform any file system modifications, network requests, or sensitive data access.
- [SAFE]: No obfuscation techniques were identified. All instructions are provided in clear, plain-text Markdown.
- [SAFE]: The skill does not include any external dependencies, package installation commands, or remote script execution patterns.
- [SAFE]: The instructions do not attempt to bypass agent safety filters or extract system prompts. The use of directives like "MANDATORY" and "NEVER" is limited to guiding the agent through its task of evaluating skill quality.
- [SAFE]: Although the skill is designed to process untrusted data (other skill files), it does not define or invoke any high-risk tools (such as shell access or web scraping) that would facilitate an indirect prompt injection attack.
Audit Metadata