MiroRL 🐙 is an MCP-first RL framework for deep research agents, with modular tools for policy optimization, environment control, and reproducible experiments.
simulator reinforcement-learning openai-gym gym rl continuous-control robot-simulation multi-agent-rl sample-efficiency mirorl miro-rl miro-env mirorl-benchmark mirorl-tutorial reinforcement-learning-bench
-
Updated
Aug 24, 2025 - Python