One-shot LLM eval cases by NAGI STUDIO - same prompt, different agents (model + harness), runnable artifacts side by side.
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).