Back to workflow skills
AI and AgentsAgent quantitative evaluationPW
AuthorCommunity author
AI and Agents Skill

Agent quantitative evaluation

Closed-loop quantitative evaluation based on Karpathy AutoResearch and multi-agent review enables automatic task evaluation and continuous optimization

0Uses12Likes15Views

What this skill does

Closed-loop quantitative evaluation based on Karpathy AutoResearch and multi-agent review enables automatic task evaluation and continuous optimization

  • Clarifies goals, inputs, and success criteria
  • Preserves evidence and marks uncertain boundaries
  • Produces structured, editable results
Installationnpx clawhub@latest install agent-evalView source