How to integrate Retellai MCP with Autogen

Trusted by
AWS
Glean
Zoom
Airtable

30 min · no commitment · see it on your stack

Retellai logo
AutoGen logo
divider

Introduction

This guide walks you through connecting Retellai to AutoGen using the Composio tool router. By the end, you'll have a working Retellai agent that can list all phone numbers linked to your account, retrieve call details for a specific agent this week, buy a new phone number with area code 415 through natural language commands.

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

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

Also integrate Retellai with

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 Retellai
  • Wire that MCP URL into Autogen using McpWorkbench and StreamableHttpServerParams
  • Configure an Autogen AssistantAgent that can call Retellai tools
  • Run a live chat loop where you ask the agent to perform Retellai 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 Retellai MCP server, and what's possible with it?

The Retellai MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Retellai account. It provides structured and secure access to your call records, phone numbers, and conversation transcripts, so your agent can perform actions like retrieving call details, managing phone numbers, initiating outbound calls, and analyzing voice data on your behalf.

  • Retrieve and analyze call records: Your agent can fetch detailed call logs, filter by agent or time, and surface insights from past conversations.
  • Initiate outbound and web-based calls: Easily direct your agent to start new phone or web calls between specific numbers or agents, supporting various business workflows.
  • Manage phone numbers and assignments: Buy, update, or delete phone numbers, and bind them to agents for streamlined inbound and outbound call handling.
  • Access and review call transcripts and details: Let your agent drill down into specific calls, pulling transcripts and metadata for compliance, training, or analytics.
  • Explore and configure voice settings: Fetch detailed information about available voice options, including provider, accent, gender, and preview audio for customization of call experiences.

Supported Tools & Triggers

Tools
Add community voiceAdd a community voice from ElevenLabs to your Retell voice library.
Add sources to knowledge baseTool to add sources (documents, URLs, text) to an existing knowledge base in Retell AI.
Buy a new phone number bind agentsThis endpoint allows purchasing a new phone number with a specified area code and binding it to designated agents for inbound and outbound calls.
Create Voice AI AgentCreate a new voice AI agent with specified configuration.
Create a new outbound phone callInitiate an outbound call by POST to '/v2/create-phone-call'.
Create a new web callThe /v2/create-web-call endpoint creates a web call with a unique agent ID, returning call details like type, token, call ID, and status in JSON format, with a 201 response.
Create Batch TestTool to create a batch test job that runs multiple test cases against an agent.
Create a new chat sessionTool to create a new chat session with a chat agent.
Create a new chat agentCreate a new chat agent with specified configuration.
Create chat completionTool to create a chat completion for an existing chat session, generating the agent's response to a user message.
Create conversation flowCreate a new Conversation Flow that can be attached to an agent for response generation.
Create conversation flow componentCreates a new shared conversation flow component at POST '/create-conversation-flow-component'.
Create a new knowledge baseTool to create a new knowledge base in Retell AI with texts, files, and URLs.
Create Retell LLM Response EngineCreate a new Retell LLM Response Engine that can be attached to an agent.
Create Test Case DefinitionTool to create a test case definition for agent QA testing in Retell AI.
Delete agentDeletes an existing agent by its unique identifier.
Delete callDelete a specific call and its associated data by call ID.
Delete chat agentDelete an existing chat agent by its unique identifier.
Delete conversation flowDelete a conversation flow and all its versions.
Delete conversation flow componentDelete a shared conversation flow component.
Delete knowledge baseDelete an existing knowledge base by its unique identifier.
Delete knowledge base sourceDelete an existing source from a knowledge base.
Delete phone numberTool to delete an existing phone number from Retell AI.
Delete Retell LLMDelete an existing Retell LLM Response Engine by its unique identifier.
Delete test case definitionDelete a test case definition by its unique identifier.
End chatTool to end an active chat session.
Retrieve details of a specific agentRetrieve details of a specific agent by its unique identifier.
Get agent versionsTool to retrieve all versions of a specific agent.
Get batch testRetrieve details and results of a specific batch test job.
Get chat detailsTool to retrieve details of a specific chat session by chat ID.
Retrieve details of a specific chat agentRetrieve details of a specific chat agent by its unique identifier.
Get all versions of a chat agentRetrieve all versions of a specific chat agent by its unique identifier.
Get concurrencyRetrieves the current concurrency and concurrency limits for the organization.
Get Conversation FlowRetrieve details of a specific Conversation Flow by its ID.
Get conversation flow componentRetrieves a shared conversation flow component by its unique identifier.
Get knowledge baseRetrieve details of a specific knowledge base by its unique identifier.
Retrieve details of a specific Retell LLMRetrieve details of a specific Retell LLM Response Engine by its unique identifier.
Import phone numberTool to import a phone number from custom telephony and bind agents to it.
List agentsRetrieves a list of all agents associated with the account.
List all chatsTool to retrieve a list of all chats associated with the account.
List all phone numbersRetrieves a list of all phone numbers associated with the account.
List batch testsTool to list batch test jobs for a response engine.
List chat agentsTool to retrieve a list of all chat agents associated with the account.
List conversation flow componentsRetrieves a list of all shared conversation flow components.
List conversation flowsTool to list all conversation flows that can be attached to an agent.
List knowledge basesTool to retrieve all knowledge bases associated with the account.
List Retell LLMsTool to list all Retell LLM Response Engines that can be attached to an agent.
List test case definitionsTool to list test case definitions for a response engine (Retell LLM or Conversation Flow).
List test runsTool to list all test case jobs (test runs) for a batch test job.
List voicesList all voices available to the user.
Publish agentPublishes the latest version of the agent and creates a new draft agent with a newer version.
Publish chat agentPublishes the latest version of the chat agent and creates a new draft chat agent with a newer version.
Register phone callRegister a phone call for custom telephony integration with Retell AI.
Retrieve call detailsTool to retrieve call details with filtering options.
Retrieve call details by idRetrieve call details by ID for web/phone calls, including type, agent ID, status, timestamps, and web access token; covering responses from success to server errors.
Retrieve phone number detailsTool to retrieve details of a specific phone number from Retell AI.
Retrieve details of a specific voiceTool to retrieve details of a specific voice by its voice_id.
Search community voiceSearch for community voices from voice providers.
Update agentUpdate an existing agent's latest draft version.
Update callUpdate an active call's parameters such as metadata, dynamic variables, or data storage settings.
Update chat agentUpdate an existing chat agent configuration.
Update chat metadataTool to update metadata and sensitive data storage settings for an existing chat.
Update conversation flowUpdate an existing conversation flow configuration.
Update conversation flow componentUpdate an existing shared conversation flow component by its ID.
Update phone number configurationUpdate agent bound to a purchased phone number.
Update Retell LLM Response EngineUpdate an existing Retell LLM Response Engine by its unique identifier.
Update test case definitionUpdate a test case definition for agent testing.

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

Prerequisites

You will need:

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

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.

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 Retellai via MCP
  • autogen-agentchat provides the AssistantAgent class
  • autogen-ext-openai provides the OpenAI model client
  • autogen-ext-tools provides MCP workbench support

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 Retellai connections to use

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 Retellai session
    composio = Composio(api_key=os.getenv("COMPOSIO_API_KEY"))
    session = composio.create(
        user_id=os.getenv("USER_ID"),
        toolkits=["retellai"]
    )
    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 Retellai tools
  • session.mcp.url is the MCP endpoint that Autogen will connect to

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

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 Retellai assistant agent with MCP tools
    agent = AssistantAgent(
        name="retellai_assistant",
        description="An AI assistant that helps with Retellai 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 Retellai tools from the workbench

Run the interactive chat loop

python
print("Chat started! Type 'exit' or 'quit' to end the conversation.\n")
print("Ask any Retellai 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 Retellai 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 Retellai 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 Retellai session
    composio = Composio(api_key=os.getenv("COMPOSIO_API_KEY"))
    session = composio.create(
        user_id=os.getenv("USER_ID"),
        toolkits=["retellai"]
    )
    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 Retellai assistant agent with MCP tools
        agent = AssistantAgent(
            name="retellai_assistant",
            description="An AI assistant that helps with Retellai 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 Retellai 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 Retellai 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 Retellai, you can reuse the same structure for other MCP-enabled apps with minimal code changes.

How to build Retellai MCP Agent with another framework

FAQ

What are the differences in Tool Router MCP and Retellai MCP?

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

Can I use Tool Router MCP with Autogen?

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 Retellai tools.

Can I manage the permissions and scopes for Retellai while using Tool Router?

Yes, absolutely. You can configure which Retellai 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.

How safe is my data with Composio Tool Router?

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 Retellai data and credentials are handled as safely as possible.

Used by agents from

Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai
Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai
Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai

Never worry about agent reliability

We handle tool reliability, observability, and security so you never have to second-guess an agent action.