Document RULER in README

README.md

@@ -21,43 +21,28 @@ Train multi-step agents for real-world tasks using GRPO.
 
 </div>
 
-## 🦜🔗 LangGraph Integration: Build Smarter Multi-Step Agents
+## 📏 RULER: Zero-Shot Agent Rewards
 
-ART's **LangGraph integration** enables you to train sophisticated ReAct-style agents that improve through reinforcement learning. Build agents that reason, use tools, and adapt their behavior over time without manual prompt engineering.
+**RULER** (Relative Universal LLM-Elicited Rewards) eliminates the need for hand-crafted reward functions by using an LLM-as-judge to automatically score agent trajectories. Simply define your task in the system prompt, and RULER handles the rest—**no labeled data, expert feedback, or reward engineering required**.
 
 ✨ **Key Benefits:**
 
-- **Automatic behavior improvement** - Train agents to get better at multi-step reasoning
-- **Tool usage optimization** - Learn when and how to use tools more effectively
-- **Seamless integration** - Drop-in replacement for LangGraph's LLM initialization
-- **RULER compatibility** - Train without hand-crafted reward functions
+- **2-3x faster development** - Skip reward function engineering entirely
+- **General-purpose** - Works across any task without modification
+- **Strong performance** - Matches or exceeds hand-crafted rewards in 3/4 benchmarks
+- **Easy integration** - Drop-in replacement for manual reward functions
 
 ```python
-import art
-from art.langgraph import init_chat_model, wrap_rollout
-from langgraph.prebuilt import create_react_agent
+# Before: Hours of reward engineering
+def complex_reward_function(trajectory):
+    # 50+ lines of careful scoring logic...
+    pass
 
-async def email_rollout(model: art.Model, scenario: str) -> art.Trajectory:
-    # Create LangGraph agent with ART's chat model
-    chat_model = init_chat_model(model.name)
-    agent = create_react_agent(chat_model, tools)
-
-    await agent.ainvoke({"messages": [("user", scenario)]})
-    return art.Trajectory(reward=1.0, messages_and_choices=[])
-
-# Train your agent
-scenarios = ["Find urgent emails", "Search Q4 budget"]
-
-# Using wrap_rollout (captures interactions automatically)
-groups = await art.gather_trajectory_groups([
-    art.TrajectoryGroup(wrap_rollout(model, email_rollout)(model, s) for _ in range(4))
-    for s in scenarios
-])
-
-await model.train(groups)
+# After: One line with RULER
+judged_group = await ruler_score_group(group, "openai/o3")
 ```
 
-[📖 Learn more about LangGraph integration →](https://art.openpipe.ai/integrations/langgraph-integration) | [🏋️ Try the notebook →](https://colab.research.google.com/github/openpipe/art-notebooks/blob/main/examples/langgraph/art-e-langgraph.ipynb)
+[📖 Learn more about RULER →](https://art.openpipe.ai/fundamentals/ruler)
 
 ## ART Overview