llm-evaluation-design
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a set of documentation and configuration files for designing LLM evaluation procedures. It does not contain executable code, scripts, or network-enabled commands.
- [SAFE]: YAML frontmatter in
SKILL.mdcorrectly scopes the skill's identity and metadata without requesting dangerous tool access. - [SAFE]: Evaluation cases (YAML files in
evals/) define testing scenarios for the skill's logic but do not execute malicious payloads or access sensitive system files. - [SAFE]: The prompt instructions in
prompts/llm-evaluation-design.mdemphasize safe practices such as least privilege, data masking, and using isolated environments for production or security-related work.
Audit Metadata