How to integrate Apify MCP with Autogen

This guide walks you through connecting Apify to AutoGen using the Composio tool router. By the end, you'll have a working Apify agent that can create a new dataset for scraped results, fetch items from a specific apify dataset, get details of your latest apify actor through natural language commands. This guide will help you understand how to give your AutoGen agent real control over a Apify account through Composio's Apify MCP server. Before we dive in, let's take a quick look at the key ideas and tools involved.

Get started for freeGet a demo
Apify logoApify
Api Key

Apify is a cloud platform for building, deploying, and managing web scraping and automation tools called Actors. It lets you automate data extraction and workflow tasks at scale—no infrastructure headaches.

112 Tools

Introduction

This guide walks you through connecting Apify to AutoGen using the Composio tool router. By the end, you'll have a working Apify agent that can create a new dataset for scraped results, fetch items from a specific apify dataset, get details of your latest apify actor through natural language commands.

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

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

Also integrate Apify with

OpenAI Agents SDKClaude Agent SDKClaude CodeClaude CoworkCodexOpenClawHermesCLIGoogle ADKLangChainVercel AI SDKMastra AILlamaIndexCrewAI

TL;DR

Here's what you'll learn:
  • Get and set up your OpenAI and Composio API keys
  • Install the required dependencies for Autogen and Composio
  • Initialize Composio and create a Tool Router session for Apify
  • Wire that MCP URL into Autogen using McpWorkbench and StreamableHttpServerParams
  • Configure an Autogen AssistantAgent that can call Apify tools
  • Run a live chat loop where you ask the agent to perform Apify operations

What is AutoGen?

Autogen is a framework for building multi-agent conversational AI systems from Microsoft. It enables you to create agents that can collaborate, use tools, and maintain complex workflows.

Key features include:

  • Multi-Agent Systems: Build collaborative agent workflows
  • MCP Workbench: Native support for Model Context Protocol tools
  • Streaming HTTP: Connect to external services through streamable HTTP
  • AssistantAgent: Pre-built agent class for tool-using assistants

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

The Apify MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Apify account. It provides structured and secure access to your web scraping and automation workflows, so your agent can create actors, manage datasets, fetch scraped data, schedule tasks, and maintain webhooks on your behalf.

  • Automated Actor Creation and Management: Easily instruct your agent to programmatically create, configure, or delete Apify actors for custom web automation or scraping jobs.
  • Dataset Handling and Data Retrieval: Let your agent spin up new datasets, organize scraped results, and pull items from datasets for downstream analysis or reporting.
  • Task Scheduling and Automation: Have your agent create and manage recurring actor tasks, making it simple to automate data extraction or browser automation at set intervals.
  • Webhook Integration and Event Handling: Direct your agent to set up or remove webhooks for actor tasks, enabling real-time notifications or downstream integrations when a task completes or fails.
  • Actor and Build Metadata Access: Empower your agent to fetch detailed metadata about actors, including build information and configuration details, for monitoring or troubleshooting purposes.

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 step08 STEPS
1

Prerequisites

You will need:

  • A Composio API key
  • An OpenAI API key (used by Autogen's OpenAIChatCompletionClient)
  • A Apify account you can connect to Composio
  • Some basic familiarity with Autogen and Python async
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

bash
pip install composio python-dotenv
pip install autogen-agentchat autogen-ext-openai autogen-ext-tools

Install Composio, Autogen extensions, and dotenv.

What's happening:

  • composio connects your agent to Apify via MCP
  • autogen-agentchat provides the AssistantAgent class
  • autogen-ext-openai provides the OpenAI model client
  • autogen-ext-tools provides MCP workbench support

4

Set up environment variables

bash
COMPOSIO_API_KEY=your-composio-api-key
OPENAI_API_KEY=your-openai-api-key
USER_ID=your-user-identifier@example.com

Create a .env file in your project folder.

What's happening:

  • COMPOSIO_API_KEY is required to talk to Composio
  • OPENAI_API_KEY is used by Autogen's OpenAI client
  • USER_ID is how Composio identifies which user's Apify connections to use
5

Import dependencies and create Tool Router session

python
import asyncio
import os
from dotenv import load_dotenv
from composio import Composio

from autogen_agentchat.agents import AssistantAgent
from autogen_ext.models.openai import OpenAIChatCompletionClient
from autogen_ext.tools.mcp import McpWorkbench, StreamableHttpServerParams

load_dotenv()

async def main():
    # Initialize Composio and create a Apify session
    composio = Composio(api_key=os.getenv("COMPOSIO_API_KEY"))
    session = composio.create(
        user_id=os.getenv("USER_ID"),
        toolkits=["apify"]
    )
    url = session.mcp.url
What's happening:
  • load_dotenv() reads your .env file
  • Composio(api_key=...) initializes the SDK
  • create(...) creates a Tool Router session that exposes Apify tools
  • session.mcp.url is the MCP endpoint that Autogen will connect to
6

Configure MCP parameters for Autogen

python
# Configure MCP server parameters for Streamable HTTP
server_params = StreamableHttpServerParams(
    url=url,
    timeout=30.0,
    sse_read_timeout=300.0,
    terminate_on_close=True,
    headers={"x-api-key": os.getenv("COMPOSIO_API_KEY")}
)

Autogen expects parameters describing how to talk to the MCP server. That is what StreamableHttpServerParams is for.

What's happening:

  • url points to the Tool Router MCP endpoint from Composio
  • timeout is the HTTP timeout for requests
  • sse_read_timeout controls how long to wait when streaming responses
  • terminate_on_close=True cleans up the MCP server process when the workbench is closed
7

Create the model client and agent

python
# Create model client
model_client = OpenAIChatCompletionClient(
    model="gpt-5",
    api_key=os.getenv("OPENAI_API_KEY")
)

# Use McpWorkbench as context manager
async with McpWorkbench(server_params) as workbench:
    # Create Apify assistant agent with MCP tools
    agent = AssistantAgent(
        name="apify_assistant",
        description="An AI assistant that helps with Apify operations.",
        model_client=model_client,
        workbench=workbench,
        model_client_stream=True,
        max_tool_iterations=10
    )

What's happening:

  • OpenAIChatCompletionClient wraps the OpenAI model for Autogen
  • McpWorkbench connects the agent to the MCP tools
  • AssistantAgent is configured with the Apify tools from the workbench
8

Run the interactive chat loop

python
print("Chat started! Type 'exit' or 'quit' to end the conversation.\n")
print("Ask any Apify related question or task to the agent.\n")

# Conversation loop
while True:
    user_input = input("You: ").strip()

    if user_input.lower() in ["exit", "quit", "bye"]:
        print("\nGoodbye!")
        break

    if not user_input:
        continue

    print("\nAgent is thinking...\n")

    # Run the agent with streaming
    try:
        response_text = ""
        async for message in agent.run_stream(task=user_input):
            if hasattr(message, "content") and message.content:
                response_text = message.content

        # Print the final response
        if response_text:
            print(f"Agent: {response_text}\n")
        else:
            print("Agent: I encountered an issue processing your request.\n")

    except Exception as e:
        print(f"Agent: Sorry, I encountered an error: {str(e)}\n")
What's happening:
  • The script prompts you in a loop with You:
  • Autogen passes your input to the model, which decides which Apify tools to call via MCP
  • agent.run_stream(...) yields streaming messages as the agent thinks and calls tools
  • Typing exit, quit, or bye ends the loop

Complete Code

Here's the complete code to get you started with Apify and AutoGen:

python
import asyncio
import os
from dotenv import load_dotenv
from composio import Composio

from autogen_agentchat.agents import AssistantAgent
from autogen_ext.models.openai import OpenAIChatCompletionClient
from autogen_ext.tools.mcp import McpWorkbench, StreamableHttpServerParams

load_dotenv()

async def main():
    # Initialize Composio and create a Apify session
    composio = Composio(api_key=os.getenv("COMPOSIO_API_KEY"))
    session = composio.create(
        user_id=os.getenv("USER_ID"),
        toolkits=["apify"]
    )
    url = session.mcp.url

    # Configure MCP server parameters for Streamable HTTP
    server_params = StreamableHttpServerParams(
        url=url,
        timeout=30.0,
        sse_read_timeout=300.0,
        terminate_on_close=True,
        headers={"x-api-key": os.getenv("COMPOSIO_API_KEY")}
    )

    # Create model client
    model_client = OpenAIChatCompletionClient(
        model="gpt-5",
        api_key=os.getenv("OPENAI_API_KEY")
    )

    # Use McpWorkbench as context manager
    async with McpWorkbench(server_params) as workbench:
        # Create Apify assistant agent with MCP tools
        agent = AssistantAgent(
            name="apify_assistant",
            description="An AI assistant that helps with Apify operations.",
            model_client=model_client,
            workbench=workbench,
            model_client_stream=True,
            max_tool_iterations=10
        )

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

        # Conversation loop
        while True:
            user_input = input("You: ").strip()

            if user_input.lower() in ['exit', 'quit', 'bye']:
                print("\nGoodbye!")
                break

            if not user_input:
                continue

            print("\nAgent is thinking...\n")

            # Run the agent with streaming
            try:
                response_text = ""
                async for message in agent.run_stream(task=user_input):
                    if hasattr(message, 'content') and message.content:
                        response_text = message.content

                # Print the final response
                if response_text:
                    print(f"Agent: {response_text}\n")
                else:
                    print("Agent: I encountered an issue processing your request.\n")

            except Exception as e:
                print(f"Agent: Sorry, I encountered an error: {str(e)}\n")

if __name__ == "__main__":
    asyncio.run(main())

Conclusion

You now have an Autogen assistant wired into Apify through Composio's Tool Router and MCP. From here you can:
  • Add more toolkits to the toolkits list, for example notion or hubspot
  • Refine the agent description to point it at specific workflows
  • Wrap this script behind a UI, Slack bot, or internal tool
Once the pattern is clear for Apify, you can reuse the same structure for other MCP-enabled apps with minimal code changes.
TOOLS

Supported Tools

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

Build Actor

Tool to build an Actor with specified configuration.

Abort Actor Build

Tool to abort an Actor build that is starting or running.

Delete Actor Build

Tool to delete an Actor build permanently.

Get Actor Build

Tool to get detailed information about a specific Actor build.

Get Actor Build Log

Tool to retrieve the log file for a specific Actor build.

Get user builds list

Tool to get a paginated list of all builds for a user.

Abort Actor Run

Tool to abort a running or starting Actor run.

Delete Actor Run

Tool to delete a finished Actor run.

Get Actor Run

Tool to get details about a specific Actor run.

Update Actor Run Status Message

Tool to update the status message of an Actor run.

Delete Actor Task

Tool to delete an Actor task permanently.

Get Actor Task

Tool to get complete details about an Actor task.

Update Actor Task

Tool to update Actor task settings using JSON payload.

Get last actor task run

Tool to get the most recent run of a specific Actor task.

Run Task Sync (GET)

Tool to run a specific task synchronously and return its output.

Run Task Sync & Get Dataset Items

Tool to run an actor task synchronously and retrieve its dataset items.

Run Task Sync with Input Override & Get Dataset Items

Tool to run an actor task synchronously with input overrides and retrieve its dataset items.

Run Task Sync (POST)

Tool to run an Actor task synchronously with input override and return its output.

Update Actor

Tool to update Actor settings using JSON payload.

Get last actor run

Tool to get the most recent run of a specific Actor.

Run Actor Sync without Input (GET)

Tool to run a specific Actor synchronously without input and return its output.

Run Actor Sync & Get Dataset Items

Tool to run Actor synchronously and get dataset items.

Get list of Actors

Tool to get the list of all Actors that the user created or used.

Delete Actor Version

Tool to delete a specific version of an Actor's source code.

Delete Actor Version Environment Variable

Tool to delete an environment variable from a specific Actor version.

Get Actor Version Environment Variable

Tool to get environment variable details for a specific Actor version.

Update Actor Version Environment Variable

Tool to update environment variable for a specific Actor version using JSON payload.

Get list of Actor version environment variables

Tool to get the list of environment variables for a specific Actor version.

Create Actor Version Environment Variable

Tool to create an environment variable for a specific Actor version.

Get Actor version

Tool to get details about a specific version of an Actor.

Update Actor Version

Tool to update an Actor version's configuration and source code.

Get list of Actor versions

Tool to get the list of versions of a specific Actor.

Create Actor Version

Tool to create a new version of an Actor.

Get list of Actor webhooks

Tool to get a list of webhooks for a specific Actor.

Create Actor

Tool to create a new Actor with specified configuration.

Create Dataset

Tool to create a new dataset.

Create Actor Task

Tool to create a new Actor task with specified settings.

Create Task Webhook

Tool to create a webhook for an Actor task.

Delete Dataset

Tool to delete a dataset permanently.

Get Dataset

Tool to retrieve dataset metadata by dataset ID.

Update Dataset

Tool to update a dataset's name via JSON payload.

Get list of datasets

Tool to get list of datasets for a user.

Get Dataset Statistics

Tool to get dataset field statistics by dataset ID.

Delete Actor

Tool to delete an Actor permanently.

Delete Webhook

Tool to delete a webhook by its ID.

Get Actor Details

Tool to get details of a specific Actor.

Get Actor Last Run Dataset Items

Tool to get dataset items from the last run of an Actor.

Get all webhooks

Tool to get a list of all webhooks created by the user.

Get dataset items

Tool to retrieve items from a dataset.

Get Default Build

Tool to get the default build for an Actor.

Get Key-Value Record

Tool to retrieve a record from a key-value store.

Get list of builds

Tool to get a list of builds for a specific Actor.

Get list of runs

Tool to get a list of runs for a specific Actor.

Get list of task runs

Tool to get a list of runs for a specific Actor task.

Get list of tasks

Tool to fetch a paginated list of tasks belonging to the authenticated user.

Get list of task webhooks

Tool to get a list of webhooks for a specific Actor task.

Get log

Tool to retrieve logs for a specific Actor run or build.

Get OpenAPI Definition

Tool to get the OpenAPI definition for a specific Actor build.

Get Run Dataset Items

Tool to get dataset items from a specific Actor run.

Get Task Input

Tool to retrieve the input configuration of a specific task.

Get Task Last Run Dataset Items

Tool to get dataset items from the last run of an Actor task.

Delete Key-Value Store

Tool to delete a key-value store permanently.

Get Key-Value Store

Tool to retrieve key-value store metadata by store ID.

Get Key-Value Store Keys

Tool to retrieve a list of keys from a key-value store.

Delete Key-Value Store Record

Tool to delete a record from a key-value store.

Check Key-Value Store Record Exists

Tool to check if a record exists in a key-value store.

Get list of key-value stores

Tool to get the list of key-value stores owned by the user.

Create Key-Value Store

Tool to create a new key-value store or retrieve an existing one by name.

List User Actor Runs

Tool to get a paginated list of all Actor runs for the authenticated user.

Delete Request Queue

Tool to delete a request queue permanently.

Get Request Queue

Tool to retrieve request queue metadata by queue ID.

Get Request Queue Head

Tool to retrieve first requests from the queue for inspection.

Get Head and Lock Queue Requests

Tool to get and lock head requests from the queue.

Update Request Queue

Tool to update request queue name using JSON payload.

Delete Request from Queue

Tool to delete a specific request from a request queue.

Get Request from Queue

Tool to retrieve a specific request from a request queue by its ID.

Delete Request Lock

Tool to delete a request lock from a request queue.

Prolong Request Lock

Tool to prolong request lock in a request queue.

Update Request in Queue

Tool to update a request in a request queue.

Batch Delete Requests from Queue

Tool to batch-delete up to 25 requests from a queue.

Batch Add Requests to Queue

Tool to batch-add up to 25 requests to a request queue.

List Request Queue Requests

Tool to list requests in a request queue with pagination support.

Add Request to Queue

Tool to add a request to the queue.

Unlock Queue Requests

Tool to unlock requests in a request queue that are currently locked by the client.

Get list of request queues

Tool to get list of request queues for a user.

Create Request Queue

Tool to create a new request queue or retrieve an existing one by name.

Run Actor Asynchronously

Tool to run a specific Actor asynchronously.

Run Actor Sync

Tool to run a specific Actor synchronously with input and return its output record.

Run Actor Sync & Get Dataset Items

Tool to run an Actor synchronously and retrieve its dataset items.

Run Task Asynchronously

Tool to run a specific Actor task asynchronously.

Delete Schedule

Tool to delete a schedule by its ID.

Get Schedule

Tool to get schedule details by ID.

Get Schedule Log

Tool to get schedule log by ID.

Update Schedule

Tool to update an existing schedule with new settings.

Get list of schedules

Tool to get list of schedules created by the user.

Create Schedule

Tool to create a new schedule with specified settings.

Store Data in Dataset

Tool to store data items in a dataset.

Store Data in Key-Value Store

Tool to create or update a record in a key-value store.

Get list of Actors in Store

Tool to get list of public Actors from Apify Store.

Update Key-Value Store

Tool to update a key-value store's properties.

Update Task Input

Tool to update the input configuration of a specific Actor task.

Get Public User Data

Tool to get public user data.

Get Current User Account Data

Tool to get private user account information.

Get Account Limits

Tool to get a complete summary of account limits and usage.

Update Account Limits

Tool to update account limits manageable on the Limits page.

Get Monthly Usage

Tool to get monthly usage summary with daily breakdown.

Get list of webhook dispatches

Tool to get list of webhook dispatches for the user.

Get Webhook Dispatch

Tool to get webhook dispatch object with all details.

Get webhook

Tool to get webhook object with all details.

Update Webhook

Tool to update webhook using JSON payload.

Test Webhook

Tool to test a webhook by creating a test dispatch with a dummy payload.

Get webhook dispatches

Tool to get list of webhook dispatches for a specific webhook.

FRAMEWORKS

How to build Apify MCP Agent with another framework

OpenAI Agents SDK logo

OpenAI Agents SDK

Use Apify MCP with OpenAI Agents SDK

Claude Agent SDK logo

Claude Agent SDK

Use Apify MCP with Claude Agent SDK

Claude Code logo

Claude Code

Use Apify MCP with Claude Code

Claude Cowork logo

Claude Cowork

Use Apify MCP with Claude Cowork

Codex logo

Codex

Use Apify MCP with Codex

OpenClaw logo

OpenClaw

Use Apify MCP with OpenClaw

Hermes logo

Hermes

Use Apify MCP with Hermes

CLI logo

CLI

Use Apify MCP with CLI

Google ADK logo

Google ADK

Use Apify MCP with Google ADK

LangChain logo

LangChain

Use Apify MCP with LangChain

Vercel AI SDK logo

Vercel AI SDK

Use Apify MCP with Vercel AI SDK

Mastra AI logo

Mastra AI

Use Apify MCP with Mastra AI

LlamaIndex logo

LlamaIndex

Use Apify MCP with LlamaIndex

CrewAI logo

CrewAI

Use Apify MCP with CrewAI

FAQ

Frequently asked questions

With a standalone Apify MCP server, the agents and LLMs can only access a fixed set of Apify tools tied to that server. However, with the Composio Tool Router, agents can dynamically load tools from Apify and many other apps based on the task at hand, all through a single MCP endpoint.

Yes, you can. Autogen fully supports MCP integration. You get structured tool calling, message history handling, and model orchestration while Tool Router takes care of discovering and serving the right Apify tools.

Yes, absolutely. You can configure which Apify scopes and actions are allowed when connecting your account to Composio. You can also bring your own OAuth credentials or API configuration so you keep full control over what the agent can do.

All sensitive data such as tokens, keys, and configuration is fully encrypted at rest and in transit. Composio is SOC 2 Type 2 compliant and follows strict security practices so your Apify data and credentials are handled as safely as possible.

Start with Apify.It takes 30 seconds.

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

Start building
Customer stories

Teams ship real agents on Composio

GET STARTED FOR FREEBOOK A DEMO
agent.aiZoomLettaGleanHubSpotAWSWabi

Karan skipped his own birthday party to fix our critical issue. It was 10pm and he diverted his Waymo to help us instead. This really sets the bar — it shows the commitment you need when users rely on your software.

Harsha Gaddipati
Harsha Gaddipati
Co-founder, Slashy

A lot of students tell us that the moment their connected tools start talking to each other inside Opennote feels almost magical. The agent just knows them, and it's immensely helped keep new users on the platform.

Abhi Arya
Abhi Arya
Co-founder, Opennote

We chose Composio over Pipedream because it delivered depth where it mattered — niche tools and tricky edge cases other platforms simply ignored. That gave us confidence to scale without compromising.

Nirman Dave
Nirman Dave
CEO, Zams

As a solo builder, shipping fast is life or death. The only way I can outcompete incumbents is by outmanoeuvring them. Getting bogged down managing agent auth would have been a death sentence.

Ryan Yu
Ryan Yu
Founder, Extra Thursday

Before Composio, adding tool integrations was slow and resource-intensive. Each one could take weeks of engineering time, and maintaining them meant constantly keeping up with API changes.

Tomisin Jenrola
Tomisin Jenrola
Founder & CEO, SwarmZero

With hands-on help from their founder, we integrated Gmail and Google Drive in just 30 minutes. This level of personal support and commitment is exactly what startups should strive for.

Jerome Leclanche
Jerome Leclanche
Co-Founder, Ingram Technologies