SIGN UPLOG IN
X (Twitter)LinkedInGitHubDiscordYouTube

How to integrate Humanloop MCP with LangChain

Connect LangChain to Humanloop MCP. Create a new project for sentiment analysis, delete your old prototype project permanently, and more using natural language, with authentication handled for you.

Get started for freeGet a demo
Humanloop logoHumanloop
Api Key

Humanloop is a platform for building and refining AI applications through user feedback and data annotation. It helps teams iterate on language model performance for more accurate and effective AI solutions.

4 Tools

Introduction

This guide walks you through connecting Humanloop to LangChain using the Composio tool router. By the end, you'll have a working Humanloop agent that can create and delete projects, list experiments, and review sessions through natural language commands.

This guide will help you understand how to give your LangChain agent real control over a Humanloop account through Composio's Humanloop MCP server.

Before we dive in, let's take a quick look at the key ideas and tools involved.

For exact tool slugs and parameter tables, see the tool catalog in the Composio docs.

Also integrate Humanloop with

Claude CodeClaude CoworkOpenAI Agents SDKClaude Agent SDKCodexOpenClawHermesAtomic AgentGoogle ADKVercel AI SDKMastra AILlamaIndexCrewAIPydantic AIAutoGen

TL;DR

Here's what you'll learn:
  • Get and set up your OpenAI and Composio API keys
  • Connect your Humanloop project to Composio
  • Create a Tool Router MCP session for Humanloop
  • Initialize an MCP client and retrieve Humanloop tools
  • Build a LangChain agent that can interact with Humanloop
  • Set up an interactive chat interface for testing

What is LangChain?

LangChain is a framework for developing applications powered by language models. It provides tools and abstractions for building agents that can reason, use tools, and maintain conversation context.

Key features include:

  • Agent Framework: Build agents that can use tools and make decisions
  • MCP Integration: Connect to external services through Model Context Protocol adapters
  • Memory Management: Maintain conversation history across interactions
  • Multi-Provider Support: Works with OpenAI, Anthropic, and other LLM providers

What is the Humanloop MCP server, and what's possible with it?

The Humanloop MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Humanloop account. It provides structured and secure access to your Humanloop projects, so your agent can perform actions like creating projects, managing experiments, and analyzing session data on your behalf.

  • Project creation and management: Let your agent spin up new Humanloop projects, add descriptions, and organize them for streamlined AI development.
  • Automated project deletion: Direct your agent to permanently remove projects, cleaning up experiments or outdated initiatives with a single command.
  • Experiment tracking and analysis: Easily retrieve and review all experiments within a project, including configurations, status, and results for continuous iteration and improvement.
  • Session history and monitoring: Ask your agent to list and analyze sessions for a given project, giving you insight into user interactions, datapoints, and evaluation history.

What is the Composio tool router, and how does it fit here?

What is Composio SDK?

Composio's Composio SDK helps agents find the right tools for a task at runtime. You can plug in multiple toolkits (like Gmail, HubSpot, and GitHub), and the agent will identify the relevant app and action to complete multi-step workflows. This can reduce token usage and improve the reliability of tool calls. Read more here: Getting started with Composio SDK

The tool router generates a secure MCP URL that your agents can access to perform actions.

How the Composio SDK works

The Composio SDK follows a three-phase workflow:

  1. Discovery: Searches for tools matching your task and returns relevant toolkits with their details.
  2. Authentication: Checks for active connections. If missing, creates an auth config and returns a connection URL via Auth Link.
  3. Execution: Executes the action using the authenticated connection.

Step-by-step Guide

Step by step10 STEPS
1

Prerequisites

Before starting this tutorial, make sure you have:
  • Python 3.10 or higher installed on your system
  • A Composio account with an API key
  • An OpenAI API key
  • Basic familiarity with Python and async programming
2

Getting API Keys for OpenAI and Composio

OpenAI API Key
  • Go to the OpenAI dashboard and create an API key. You'll need credits to use the models, or you can connect to another model provider.
  • Keep the API key safe.
Composio API Key
  • Log in to the Composio dashboard.
  • Navigate to your API settings and generate a new API key.
  • Store this key securely as you'll need it for authentication.
3

Install dependencies

npm install @composio/langchain @langchain/core @langchain/openai @langchain/mcp-adapters dotenv

Install the required packages for LangChain with MCP support.

What's happening:

  • @composio/langchain provides Composio integration for LangChain
  • @langchain/mcp-adapters enables MCP client connections
  • @langchain/core is the core agent framework
  • dotenv/config loads environment variables
4

Set up environment variables

bash
COMPOSIO_API_KEY=your_composio_api_key_here
COMPOSIO_USER_ID=your_composio_user_id_here
OPENAI_API_KEY=your_openai_api_key_here

Create a .env file in your project root.

What's happening:

  • COMPOSIO_API_KEY authenticates your requests to Composio's API
  • COMPOSIO_USER_ID identifies the user for session management
  • OPENAI_API_KEY enables access to OpenAI's language models
5

Import dependencies

import { Composio } from '@composio/core';
import { LangchainProvider } from '@composio/langchain';
import { MultiServerMCPClient } from "@langchain/mcp-adapters";
import { createAgent } from "langchain";
import * as readline from 'readline';
import 'dotenv/config';

dotenv.config();
What's happening:
  • We're importing LangChain's MCP adapter and Composio SDK
  • The dotenv/config import loads environment variables from your .env file
  • This setup prepares the foundation for connecting LangChain with Humanloop functionality through MCP
6

Initialize Composio client

const composioApiKey = process.env.COMPOSIO_API_KEY;
const userId = process.env.COMPOSIO_USER_ID;

if (!composioApiKey) throw new Error('COMPOSIO_API_KEY is not set');
if (!userId) throw new Error('COMPOSIO_USER_ID is not set');

async function main() {
    const composio = new Composio({
        apiKey: composioApiKey as string,
        provider: new LangchainProvider()
    });
What's happening:
  • We're loading the COMPOSIO_API_KEY from environment variables and validating it exists
  • Creating a Composio instance that will manage our connection to Humanloop tools
  • Validating that COMPOSIO_USER_ID is also set before proceeding
7

Create a Tool Router session

const session = await composio.create(
    userId as string,
    {
        toolkits: ['humanloop']
    }
);

const url = session.mcp.url;
What's happening:
  • We're creating a Tool Router session that gives your agent access to Humanloop tools
  • The create method takes the user ID and specifies which toolkits should be available
  • The returned session.mcp.url is the MCP server URL that your agent will use
  • This approach allows the agent to dynamically load and use Humanloop tools as needed
8

Configure the agent with the MCP URL

const client = new MultiServerMCPClient({
    "humanloop-agent": {
        transport: "http",
        url: url,
        headers: {
            "x-api-key": process.env.COMPOSIO_API_KEY
        }
    }
});

const tools = await client.getTools();

const agent = createAgent({ model: "gpt-5", tools });
What's happening:
  • We're creating a MultiServerMCPClient that connects to our Humanloop MCP server via HTTP
  • The client is configured with a name and the URL from our Tool Router session
  • getTools() retrieves all available Humanloop tools that the agent can use
  • We're creating a LangChain agent using the GPT-5 model
9

Set up interactive chat interface

let conversationHistory: any[] = [];

console.log("Chat started! Type 'exit' or 'quit' to end the conversation.\n");
console.log("Ask any Humanloop related question or task to the agent.\n");

const rl = readline.createInterface({
    input: process.stdin,
    output: process.stdout,
    prompt: 'You: '
});

rl.prompt();

rl.on('line', async (userInput: string) => {
    const trimmedInput = userInput.trim();

    if (['exit', 'quit', 'bye'].includes(trimmedInput.toLowerCase())) {
        console.log("\nGoodbye!");
        rl.close();
        process.exit(0);
    }

    if (!trimmedInput) {
        rl.prompt();
        return;
    }

    conversationHistory.push({ role: "user", content: trimmedInput });
    console.log("\nAgent is thinking...\n");

    const response = await agent.invoke({ messages: conversationHistory });
    conversationHistory = response.messages;

    const finalResponse = response.messages[response.messages.length - 1]?.content;
    console.log(`Agent: ${finalResponse}\n`);
        
        rl.prompt();
    });

    rl.on('close', () => {
        console.log('\n👋 Session ended.');
        process.exit(0);
    });
What's happening:
  • We initialize an empty conversationHistory list to maintain context across interactions
  • A readline interface is used to continuously accept user input from the command line
  • When a user types a message, it's added to the conversation history and sent to the agent
  • The agent processes the request using the invoke() method with the full conversation history
  • Users can type 'exit', 'quit', or 'bye' to end the chat session gracefully
10

Run the application

main().catch((err) => {
    console.error('Fatal error:', err);
    process.exit(1);
});
What's happening:
  • We call the main() function to start the application

Complete Code

Here's the complete code to get you started with Humanloop and LangChain:

import { Composio } from '@composio/core';
import { LangchainProvider } from '@composio/langchain';
import { MultiServerMCPClient } from "@langchain/mcp-adapters";  
import { createAgent } from "langchain";
import * as readline from 'readline';
import 'dotenv/config';

const composioApiKey = process.env.COMPOSIO_API_KEY;
const userId = process.env.COMPOSIO_USER_ID;

if (!composioApiKey) throw new Error('COMPOSIO_API_KEY is not set');
if (!userId) throw new Error('COMPOSIO_USER_ID is not set');

async function main() {
    const composio = new Composio({
        apiKey: composioApiKey as string,
        provider: new LangchainProvider()
    });

    const session = await composio.create(
        userId as string,
        {
            toolkits: ['humanloop']
        }
    );

    const url = session.mcp.url;
    
    const client = new MultiServerMCPClient({
        "humanloop-agent": {
            transport: "http",
            url: url,
            headers: {
                "x-api-key": process.env.COMPOSIO_API_KEY
            }
        }
    });
    
    const tools = await client.getTools();
  
    const agent = createAgent({ model: "gpt-5", tools });
    
    let conversationHistory: any[] = [];
    
    console.log("Chat started! Type 'exit' or 'quit' to end the conversation.\n");
    console.log("Ask any Humanloop related question or task to the agent.\n");
    
    const rl = readline.createInterface({
        input: process.stdin,
        output: process.stdout,
        prompt: 'You: '
    });

    rl.prompt();

    rl.on('line', async (userInput: string) => {
        const trimmedInput = userInput.trim();
        
        if (['exit', 'quit', 'bye'].includes(trimmedInput.toLowerCase())) {
            console.log("\nGoodbye!");
            rl.close();
            process.exit(0);
        }
        
        if (!trimmedInput) {
            rl.prompt();
            return;
        }
        
        conversationHistory.push({ role: "user", content: trimmedInput });
        console.log("\nAgent is thinking...\n");
        
        const response = await agent.invoke({ messages: conversationHistory });
        conversationHistory = response.messages;
        
        const finalResponse = response.messages[response.messages.length - 1]?.content;
        console.log(`Agent: ${finalResponse}\n`);
        
        rl.prompt();
    });

    rl.on('close', () => {
        console.log('\nSession ended.');
        process.exit(0);
    });
}

main().catch((err) => {
    console.error('Fatal error:', err);
    process.exit(1);
});

Conclusion

You've successfully built a LangChain agent that can interact with Humanloop through Composio's Tool Router.

Key features of this implementation:

  • Dynamic tool loading through Composio's Tool Router
  • Conversation history maintenance for context-aware responses
  • Async Python provides clean, efficient execution of agent workflows
You can extend this further by adding error handling, implementing specific business logic, or integrating additional Composio toolkits to create multi-app workflows.
TOOLS

Supported Tools

Every Humanloop action and event your agent gets out of the box.

Create Project

Creates a new Humanloop project using a required name and optional description and directory ID.

Delete Project

This tool allows you to delete a specific project from your Humanloop organization.

List Experiments

Retrieves an array of experiments associated with a Humanloop project identified by a project ID starting with 'pr_'.

List Sessions

This tool retrieves a paginated list of sessions for a specific project in Humanloop.

FRAMEWORKS

How to build Humanloop MCP Agent with another framework

Claude Code logo

Claude Code

Use Humanloop MCP with Claude Code

Claude Cowork logo

Claude Cowork

Use Humanloop MCP with Claude Cowork

OpenAI Agents SDK logo

OpenAI Agents SDK

Use Humanloop MCP with OpenAI Agents SDK

Claude Agent SDK logo

Claude Agent SDK

Use Humanloop MCP with Claude Agent SDK

Codex logo

Codex

Use Humanloop MCP with Codex

OpenClaw logo

OpenClaw

Use Humanloop MCP with OpenClaw

Hermes logo

Hermes

Use Humanloop MCP with Hermes

Atomic Agent logo

Atomic Agent

Use Humanloop MCP with Atomic Agent

Google ADK logo

Google ADK

Use Humanloop MCP with Google ADK

Vercel AI SDK logo

Vercel AI SDK

Use Humanloop MCP with Vercel AI SDK

Mastra AI logo

Mastra AI

Use Humanloop MCP with Mastra AI

LlamaIndex logo

LlamaIndex

Use Humanloop MCP with LlamaIndex

CrewAI logo

CrewAI

Use Humanloop MCP with CrewAI

Pydantic AI logo

Pydantic AI

Use Humanloop MCP with Pydantic AI

AutoGen logo

AutoGen

Use Humanloop MCP with AutoGen

FAQ

Frequently asked questions

Start with Humanloop.It takes 30 seconds.

Managed auth, hosted MCP servers, and every Humanloop tool your agent needs.Free to start.

Start building