AgentShield for LangChain

Monitor LangChain agents with full cost tracking, session replay, and guardrails. One line of code.

Quick setup

import agentshield
from langchain.agents import AgentExecutor

shield = agentshield.init(api_key="your-key")

# Wrap your existing LangChain agent
with shield.session(agent_id="my-langchain-agent") as session:
    result = agent_executor.invoke({"input": user_input})
    session.track_langchain(result)