Skip to main content
A single callback handler covers LangChain and every framework built on top of it — including Langflow, LangGraph, and LangServe.

​
Installation

pip install langsight langchain

​
Setup

from langsight.sdk import LangSightClient
from langsight.integrations.langchain import LangSightLangChainCallback

client = LangSightClient(url="http://localhost:8000")

callback = LangSightLangChainCallback(
    client=client,
    server_name="my-tools",        # shown in langsight sessions
    agent_name="my-agent",         # optional
    session_id="sess-abc123",      # optional — groups calls into a session
)

​
LangChain agents

from langchain.agents import initialize_agent, AgentType

agent = initialize_agent(
    tools=tools,
    llm=llm,
    agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION,
    callbacks=[callback],          # ← one line
)

agent.run("What is the weather in Berlin?")

​
LangGraph

from langgraph.graph import StateGraph

graph = StateGraph(...)
# ... define your graph ...

# Pass callback in config
graph.invoke(
    {"input": "your task"},
    config={"callbacks": [callback]},   # ← one line
)

​
Langflow

In Langflow, add the callback to any component that uses tools:
  1. Open your flow in the Langflow UI
  2. Click on the Agent or Tool component
  3. Under Advanced Settings, add to the callbacks list:
from langsight.integrations.langchain import LangSightLangChainCallback
from langsight.sdk import LangSightClient

callback = LangSightLangChainCallback(
    client=LangSightClient(url=os.getenv("LANGSIGHT_URL", "http://localhost:8000")),
    agent_name="langflow-agent",
)

​
What gets traced

Every time a LangChain Tool is called — whether it’s a database lookup, web search, API call, or MCP server tool — LangSight records:
  • Tool name
  • Start time + latency
  • Success / error / timeout status
  • Error message (if failed)
  • Agent name + session ID

​
View traces

langsight sessions

Agent Sessions  (last 24h)
────────────────────────────────────────────────────────────
Session          Agent           Calls   Failed   Duration
sess-abc123      langflow-agent  8       1        3.4s

langsight sessions --id sess-abc123

Trace: sess-abc123  (langflow-agent)
├── 🔧 search-tool/search           340ms  ✓
├── 🔧 postgres-mcp/query            42ms  ✓
├── 🔧 calculator/evaluate            1ms  ✓
└── 🔧 slack-mcp/post_message          —   ✗  timeout

​
Multi-agent (LangGraph)

For LangGraph workflows with multiple agents, pass trace_id to group all agents under one task: 
import uuid
trace_id = str(uuid.uuid4())

orchestrator_callback = LangSightLangChainCallback(
    client=client,
    agent_name="orchestrator",
    trace_id=trace_id,
    session_id=session_id,
)

researcher_callback = LangSightLangChainCallback(
    client=client,
    agent_name="researcher",
    trace_id=trace_id,       # same trace_id — links all agents
    session_id=session_id,
)

​
MCP servers in LangChain

If your LangChain agent uses MCP servers as tools, LangSight gives you full observability:
  • Tool call traces via this callback
  • MCP server health via langsight mcp-health
  • Security scanning via langsight security-scan
No extra setup needed — the callback + CLI work independently and complement each other.