nagi-studio/nagi-bench

One-shot LLM eval cases by NAGI STUDIO - same prompt, different agents (model + harness), runnable artifacts side by side.

1 file for Codex and OpenCode: nagi-bench AGENTS.md — 2,358 tokens loaded in every session.

Open on GitHub bench.nagi.fun

59Stars on the repository

1Files it configures its agents with

2,358Tokens loaded in every session

2Agents configured

Instructions

nagi-bench AGENTS.md AGENTS.md A 2,358 tok

These files are nagi-studio/nagi-bench's own configuration — they tell Codex and OpenCode how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.