langgraph-omega

LangGraph Integration for OmegaEngine

✨ Features

• 🕸️ Graph-level Governance — Control entire agent workflows
• 🔗 Node Validation — Validate each graph node execution
• 🔄 State Guards — Monitor state transitions
• 📊 Execution Tracing — Full audit of graph execution
• 🛡️ Conditional Overrides — Policy-based routing

📦 Installation

pip install langgraph-omega

🚀 Quick Start

from langgraph.graph import StateGraph
from langgraph_omega import OmegaGraph

# Define your graph
workflow = StateGraph(AgentState)
workflow.add_node("research", research_agent)
workflow.add_node("write", write_agent)
workflow.add_edge("research", "write")

# Wrap with governance
governed_graph = OmegaGraph(
    workflow.compile(),
    omega_api_key="your-omega-key",
    policy_id="agent_workflow",
)

# All node executions are governed
result = governed_graph.invoke({"input": "Write about AI"})

🔗 Node-level Governance

from langgraph_omega import governed_node

@governed_node(policy_id="research_policy")
def research_agent(state):
    # Node execution is validated
    return {"research": "..."}

workflow.add_node("research", research_agent)

📄 License

Licensed under the Apache License 2.0.