langfuse-omega

Langfuse + OmegaEngine Integration

✨ Features

• 📊 Unified Observability — Langfuse tracing + OmegaEngine governance
• 🔗 Bi-directional Sync — Traces flow both ways
• 📈 Policy Metrics — Governance data in Langfuse dashboard
• 🛡️ Alert on Violations — Trigger Langfuse alerts on policy blocks
• ⚡ Zero Config — Auto-detect and integrate

📦 Installation

pip install langfuse-omega

🚀 Quick Start

from langfuse_omega import OmegaLangfuseCallback

callback = OmegaLangfuseCallback(
    langfuse_secret_key="your-langfuse-key",
    langfuse_public_key="your-public-key",
    omega_api_key="your-omega-key",
)

# Use with any LLM framework
from langchain_openai import ChatOpenAI

llm = ChatOpenAI(callbacks=[callback])
response = llm.invoke("Hello!")

# View in Langfuse dashboard with governance metadata

📊 Dashboard Integration

In your Langfuse dashboard, you'll see:

• ✅ Omega Decision (ALLOW/BLOCK/ESCALATE)
• 📊 Risk Score
• 🏷️ Policy Tags
• 🔗 Link to Omega audit trail

🔔 Alert Configuration

callback = OmegaLangfuseCallback(
    alert_on_deny=True,
    alert_on_high_risk=True,
    risk_threshold=70,
)

📄 License

Licensed under the Apache License 2.0.