Lego-RL: Harness-Native Reinforcement Learning for Coding Agents
Official Repo for (Self-Improving)^2 Coding Agent: Curating High-Quality Trajectories via Recursive Self-Improvement
EvoPolicyGym is infrastructure for evaluating coding agents and generating training experience through Autonomous Policy Evolution.
SWE-Together: Evaluating Coding Agents in Interactive User Sessions
Official Repository of Harness-R1: Learning to Edit Executable Runtime Harnesses from Agent Failure Trajectories
Local outcome-loop + reward layer for coding agents — remember what worked, ship the next move, measure how it landed. Docs at nimroboai.com
Evidence-first infrastructure for reproducible, isolated executable evaluation and online rewards.
Agentic Research and Evaluation Suite