AgriManager / Docs
Website GitHub

AgriManager

A multi-simulator LLM-RL framework for agricultural management.

AgriManager integrates large language models with agricultural reinforcement-learning environments. It unifies the DSSAT, WOFOST, and Cycles crop simulators behind a single prompt-based text interface, so one reasoning policy can be trained jointly across simulators, objectives, and interfaces — using GRPO via VeRL.

The AgriManager training loop. DSSAT, WOFOST and Cycles sit behind one unified environment interface. The simulator sends the state s to the LLM policy as text, the policy sends an action a back, and the rewards r feed an RL update from trajectories, which updates the policy.

What is AgriManager#

A crop-management policy faces heterogeneity of two kinds. Distributional shift keeps the interface fixed while weather, crop, or prices change. Schema shift changes the interface itself — different farms expose different observed variables, actions, or objectives.

A fixed-interface neural policy cannot run once a shift changes its input layout or action encoding. A prompt-conditioned LLM policy re-reads the task schema from the prompt of each episode, so a single trained policy can operate across interfaces that would otherwise each need their own architecture. AgriManager is the framework — and benchmark — for studying exactly this.

Highlights#

Start here#

Quick start →

Explore the docs#