Programmable Worlds
for AI Agents
Sandboxed, observable environments to train, evaluate, and verify intelligent systems.
Four Pillars of Programmable Worlds
RL Gym Environments for Agent Training
Train agents on real applications with live state. Collect rollout trajectories, score episodes with layered verifiers, and export directly to policy fine-tuning datasets.
Premade & Custom Apps
Pre-seeded Gmail and Slack replicas, Ubuntu shell, Chromium browser, or custom FastAPI apps.
Computed Ground Truth
Exact state, event sequences, and temporal logic verifiers—no model grading contamination.
Direct Policy Export
Native serialization to GRPO rollouts, DPO preference pairs, and SFT datasets.
env = forge.make("gmail-v1", agent="anthropic:claude-3-7")
obs, info = env.reset(seed=42)
action = agent.predict(obs)
dataset.export("grpo_rollouts.parquet")
Engineered for Fidelity
realistic simulation, verified reward signals, and policy training.
Gymnasium Environment Facade
Shared contracts across all environments for reset, state management, tool schemas, and step execution with zero state leaks.
Layered Verification Suite
Six built-in verifiers compute mathematical verdicts from state transitions and event ordering, enforcing grader independence.
Replay & Branch Engine
Deterministically re-run episodes step-by-step or fork from step T to test alternate action paths and edge cases.
Dataset Exporters
Convert completed trajectories directly into GRPO rollouts, DPO preference pairs, SFT pairs, and failure datasets.
Trace & Loss Taxonomy
Correlates prompts, tool calls, and state changes into unified traces with automated 7-mode failure classification.
Generalization Benchmark
Evaluates trained checkpoints strictly on held-out environments to measure pass rates and reward stability across seeds.
Build Programmable Worlds
Spin up sandboxed environments, evaluate autonomous agents with computed ground truth, and train policies on real-world applications.