Use to test whether an AI agent, prompt, or skill actually works before or after it ships — build a small eval set, score runs, and catch regressions when the prompt or model changes. Trigger when the user asks whether a prompt is good, wants to compare models or versions, sees inconsistent agent output, or is about to put an agent in front of customers.
Install
npx skills add OneWave-AI/claude-skills --skill agent-eval-harnessnpx skills add OneWave-AI/claude-skills --skill agent-eval-harness -a claude-code -gnpx skills add OneWave-AI/claude-skills --skill agent-eval-harness -a codex -ggemini skills install https://github.com/OneWave-AI/claude-skills.git --path agent-eval-harnessgh skill install OneWave-AI/claude-skills agent-eval-harnessgit clone --depth 1 https://github.com/OneWave-AI/claude-skills /tmp/OneWave-AI__claude-skills
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/OneWave-AI__claude-skills/agent-eval-harness ~/.claude/skills/agent-eval-harness # Claude Code
cp -R /tmp/OneWave-AI__claude-skills/agent-eval-harness ~/.agents/skills/agent-eval-harness # Codex, Cursor, Gemini CLI, Copilot, OpenCode