The Agent Skills Directory

[SAFE]: The skill defines a standard instructional workflow for using quality review tools such as list_output_scoring_rules, score_files, and get_file_scores.
[SAFE]: All referenced tools and installation instructions (e.g., npx skills add layerai/skills) belong to the author's official ecosystem and represent expected functionality.
[SAFE]: The content focuses on operational best practices for LLM output evaluation, such as checking constraints, context, and set consistency, without attempting to bypass safety filters or exfiltrate data.

layer-quality