Run evaluations for Hugging Face Hub models using inspect-ai and lighteval on local hardware. Use for backend selection, local GPU evals, and choosing between vLLM / Transformers / accelerate. Not for HF Jobs orchestration, model-card PRs, .eval_results publication, or community-evals automation.
Install
npx skills add huggingface/skills --skill huggingface-community-evalsnpx skills add huggingface/skills --skill huggingface-community-evals -a claude-code -gnpx skills add huggingface/skills --skill huggingface-community-evals -a codex -ggemini skills install https://github.com/huggingface/skills.git --path skills/huggingface-community-evalsgh skill install huggingface/skills huggingface-community-evalsgit clone --depth 1 https://github.com/huggingface/skills /tmp/huggingface__skills
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/huggingface__skills/skills/huggingface-community-evals ~/.claude/skills/huggingface-community-evals # Claude Code
cp -R /tmp/huggingface__skills/skills/huggingface-community-evals ~/.agents/skills/huggingface-community-evals # Codex, Cursor, Gemini CLI, Copilot, OpenCode