# How to integrate Icypeas MCP with LlamaIndex

```json
{
  "title": "How to integrate Icypeas MCP with LlamaIndex",
  "toolkit": "Icypeas",
  "toolkit_slug": "icypeas",
  "framework": "LlamaIndex",
  "framework_slug": "llama-index",
  "url": "https://composio.dev/toolkits/icypeas/framework/llama-index",
  "markdown_url": "https://composio.dev/toolkits/icypeas/framework/llama-index.md",
  "updated_at": "2026-05-12T10:15:36.053Z"
}
```

## Introduction

This guide walks you through connecting Icypeas to LlamaIndex using the Composio tool router. By the end, you'll have a working Icypeas agent that can find verified email for john doe at acme.com, bulk search emails for 100 new leads, list all role-based emails at example.org through natural language commands.
This guide will help you understand how to give your LlamaIndex agent real control over a Icypeas account through Composio's Icypeas MCP server.
Before we dive in, let's take a quick look at the key ideas and tools involved.

## Also integrate Icypeas with

- [OpenAI Agents SDK](https://composio.dev/toolkits/icypeas/framework/open-ai-agents-sdk)
- [Claude Agent SDK](https://composio.dev/toolkits/icypeas/framework/claude-agents-sdk)
- [Claude Code](https://composio.dev/toolkits/icypeas/framework/claude-code)
- [Claude Cowork](https://composio.dev/toolkits/icypeas/framework/claude-cowork)
- [Codex](https://composio.dev/toolkits/icypeas/framework/codex)
- [OpenClaw](https://composio.dev/toolkits/icypeas/framework/openclaw)
- [Hermes](https://composio.dev/toolkits/icypeas/framework/hermes-agent)
- [CLI](https://composio.dev/toolkits/icypeas/framework/cli)
- [Google ADK](https://composio.dev/toolkits/icypeas/framework/google-adk)
- [LangChain](https://composio.dev/toolkits/icypeas/framework/langchain)
- [Vercel AI SDK](https://composio.dev/toolkits/icypeas/framework/ai-sdk)
- [Mastra AI](https://composio.dev/toolkits/icypeas/framework/mastra-ai)
- [CrewAI](https://composio.dev/toolkits/icypeas/framework/crew-ai)

## TL;DR

Here's what you'll learn:
- Set your OpenAI and Composio API keys
- Install LlamaIndex and Composio packages
- Create a Composio Tool Router session for Icypeas
- Connect LlamaIndex to the Icypeas MCP server
- Build a Icypeas-powered agent using LlamaIndex
- Interact with Icypeas through natural language

## What is LlamaIndex?

LlamaIndex is a data framework for building LLM applications. It provides tools for connecting LLMs to external data sources and services through agents and tools.
Key features include:
- ReAct Agent: Reasoning and acting pattern for tool-using agents
- MCP Tools: Native support for Model Context Protocol
- Context Management: Maintain conversation context across interactions
- Async Support: Built for async/await patterns

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

The Icypeas MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Icypeas account. It provides structured and secure access to professional email discovery and verification, so your agent can perform actions like finding emails, verifying addresses, searching company data, and scanning domains on your behalf.
- Accurate email discovery and verification: Instantly find and verify professional email addresses using first name, last name, and company domain to supercharge your outreach or lead generation.
- Bulk prospecting and search management: Launch bulk email or profile URL searches for thousands of contacts at once, then track progress and fetch results without manual oversight.
- Comprehensive company and people lookup: Search for companies or filter people by name, title, company, and more to enrich your CRM or build targeted prospect lists efficiently.
- Domain scanning for role-based emails: Scan entire company domains to discover all available role-based email addresses, simplifying large-scale contact discovery.
- Subscription and usage insights: Check your Icypeas subscription details and remaining credits, helping you stay on top of your usage and plan outreach campaigns smarter.

## Supported Tools

| Tool slug | Name | Description |
|---|---|---|
| `ICYPEAS_BULK_EMAIL_SEARCH` | Bulk Email Search | Initiate a bulk email search job to find professional email addresses for multiple people at once. Use this tool when you need to find emails for more than one person. Provide names (first and/or last) along with company domains or names. The API queues the job and returns a file identifier that can be used with 'Fetch Bulk Search Info' and 'Retrieve Search Results' to check progress and get results. Rate limit: 1 request per second. Maximum 5000 rows per bulk request. |
| `ICYPEAS_BULK_FIND_PROFILE_URLS` | Find Profile URLs Bulk | Tool to perform bulk search for profile URLs based on firstname, lastname, and company/domain or job title. Use when you need to find LinkedIn or other social profile URLs for multiple prospects at once (up to 50 per request). This endpoint returns immediate results with profile URLs. Each result includes the found URL (or empty string if not found), a search ID, and a status indicator. |
| `ICYPEAS_BULK_REVERSE_EMAIL_LOOKUP` | Bulk Reverse Email Lookup | Tool to find LinkedIn profile URLs for multiple professional email addresses in a single request. Use when you need to reverse lookup 2-50 email addresses to find their associated LinkedIn profiles. Each lookup costs 10 credits per found profile. Returns results immediately (not async like other bulk operations). |
| `ICYPEAS_BULK_SCRAPE` | Scrape Bulk | Tool to scrape multiple LinkedIn profiles or companies in bulk (up to 50 per request). Use when you need to scrape multiple LinkedIn URLs at once. Returns a job ID that can be used with check_progress to fetch results. |
| `ICYPEAS_CHECK_PROGRESS` | Check Search Progress | Check the processing progress of a search by its ID. Use this tool after initiating a single or bulk search to monitor its status. For mode='single': Checks progress of individual search items via /bulk-single-searchs/read endpoint. For mode='bulk': Checks progress of bulk search files via /search-files/read endpoint. Poll this endpoint periodically until 'finished' is True or item status is 'DEBITED'. Note: ICYPEAS recommends using webhooks instead of polling for production use. |
| `ICYPEAS_COUNT_COMPANIES` | Count Companies | Tool to count companies in Icypeas database matching specified filters without returning data or being charged. Use when you need to know how many companies match specific criteria like industry, location, or headcount. |
| `ICYPEAS_COUNT_PEOPLE` | Count People | Tool to count people matching specified filters without retrieving data or consuming credits. Use when you need to assess the size of a people search result set before running the actual query. |
| `ICYPEAS_DOMAIN_SCAN` | Domain Scan | Tool to scan a domain for role-based email addresses. Use when you need to discover all role-based emails for a specific domain. |
| `ICYPEAS_FETCH_BULK_SEARCH_INFO` | Fetch Bulk Search Info | Retrieve bulk search files with their status and progress. Lists all bulk searches (email, profile, company searches) with pagination support. Use to monitor bulk operation progress, check completion status, or retrieve file IDs for further operations. |
| `ICYPEAS_FETCH_SUBSCRIPTION_INFO` | Fetch Subscription Information | Retrieves subscription details and remaining credits for an Icypeas account. Use this to check credit balances, subscription status, and user ID before performing searches. The email parameter must match the account owner's email associated with the API key. |
| `ICYPEAS_FIND_COMPANIES` | Find Companies | Tool to search companies in Icypeas database. Use when you need to find companies matching filters like industry, location, or headcount. |
| `ICYPEAS_FIND_COMPANY_URL` | Find Company URL | Tool to find a single company profile URL using a company name or domain. Use when you need to initiate a profile-URL search for a given company identifier. |
| `ICYPEAS_FIND_LINKEDIN_PROFILE_BY_EMAIL` | Reverse Email Lookup | Find the LinkedIn profile URL behind a single professional email address. Use when you need to identify the person associated with an email address. Costs 10 credits per found profile. |
| `ICYPEAS_FIND_PEOPLE` | Find People | Search for people/leads in the Icypeas database. Supports filtering by name, job title, company, location, skills, languages, school, and keywords. Each filter supports include/exclude lists. Returns matching leads with pagination support for large result sets. Requires credits to execute. |
| `ICYPEAS_FIND_PROFILE_URL` | Find Profile URL | Finds a person's LinkedIn profile URL using their name and company or job title. Use this tool when you need to find someone's LinkedIn profile URL. Provide the person's first name, last name, and either their company/domain OR job title. Tips for best results: - For company/domain, use the website domain (e.g., 'icypeas.com') or company name - Keep inputs simple - avoid over-specifying details - Costs 1 credit per successful search Example: Find LinkedIn URL for Pierre Landoin at icypeas.com |
| `ICYPEAS_FIND_SINGLE_EMAIL` | Find Single Email | Initiates an asynchronous email search to find a prospect's email address using their name and company. Returns a search item ID that can be used with ICYPEAS_RETRIEVE_SEARCH_RESULTS to get the actual email. Use for individual lookups; for bulk operations, use ICYPEAS_BULK_EMAIL_SEARCH instead. |
| `ICYPEAS_PARSE_BULK_SEARCH_STATISTICS` | Statistics Bulk Search | Tool to parse bulk search statistics webhook. Use when processing the completion notification of a bulk search. |
| `ICYPEAS_RETRIEVE_SEARCH_RESULTS` | Retrieve Search Results | Tool to retrieve the results of a search by ID or to paginate through bulk search results. Use after initiating a search to fetch individual or multiple search results. |
| `ICYPEAS_SCRAPE_COMPANY` | Scrape Company | Tool to initiate scraping of a LinkedIn company page. Use when you have a LinkedIn company URL and need to retrieve company profile data. Returns a job ID that can be used with check_progress to fetch the full results. |
| `ICYPEAS_SCRAPE_PROFILE` | Scrape Profile | Tool to initiate scraping of a LinkedIn profile. Use when you have a public LinkedIn profile URL and need a job ID and status to later fetch full details. |
| `ICYPEAS_SETUP_NOTIFICATIONS` | Setup Notifications | Provides instructions for setting up Icypeas push notifications/webhooks. Use this tool when you need to configure real-time notifications for bulk searches or single search updates. Note: Webhooks must be configured via the Icypeas dashboard; there is no API endpoint for programmatic setup. |
| `ICYPEAS_VERIFY_EMAIL` | Verify Email Address | Tool to verify if a specific email address exists and is valid. Use when you need to check email deliverability. Returns a verification item ID that can be used with ICYPEAS_RETRIEVE_SEARCH_RESULTS to get the verification result. |

## Supported Triggers

None listed.

## Creating MCP Server - Stand-alone vs Composio SDK

The Icypeas MCP server is an implementation of the Model Context Protocol that connects your AI agent to Icypeas. It provides structured and secure access so your agent can perform Icypeas operations on your behalf through a secure, permission-based interface.
With Composio's managed implementation, you don't have to create your own developer app. For production, if you're building an end product, we recommend using your own credentials. The managed server helps you prototype fast and go from 0-1 faster.

## Step-by-step Guide

### 1. Prerequisites

Before you begin, make sure you have:
- Python 3.8/Node 16 or higher installed
- A Composio account with the API key
- An OpenAI API key
- A Icypeas account and project
- Basic familiarity with async Python/Typescript

### 1. Getting API Keys for OpenAI, Composio, and Icypeas

No description provided.

### 2. Installing dependencies

No description provided.
```python
pip install composio-llamaindex llama-index llama-index-llms-openai llama-index-tools-mcp python-dotenv
```

```typescript
npm install @composio/llamaindex @llamaindex/openai @llamaindex/tools @llamaindex/workflow dotenv
```

### 3. Set environment variables

Create a .env file in your project root:
These credentials will be used to:
- Authenticate with OpenAI's GPT-5 model
- Connect to Composio's Tool Router
- Identify your Composio user session for Icypeas access
```bash
OPENAI_API_KEY=your-openai-api-key
COMPOSIO_API_KEY=your-composio-api-key
COMPOSIO_USER_ID=your-user-id
```

### 4. Import modules

No description provided.
```python
import asyncio
import os
import dotenv

from composio import Composio
from composio_llamaindex import LlamaIndexProvider
from llama_index.core.agent.workflow import ReActAgent
from llama_index.core.workflow import Context
from llama_index.llms.openai import OpenAI
from llama_index.tools.mcp import BasicMCPClient, McpToolSpec

dotenv.load_dotenv()
```

```typescript
import "dotenv/config";
import readline from "node:readline/promises";
import { stdin as input, stdout as output } from "node:process";

import { Composio } from "@composio/core";

import { mcp } from "@llamaindex/tools";
import { agent as createAgent } from "@llamaindex/workflow";
import { openai } from "@llamaindex/openai";

dotenv.config();
```

### 5. Load environment variables and initialize Composio

No description provided.
```python
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
COMPOSIO_API_KEY = os.getenv("COMPOSIO_API_KEY")
COMPOSIO_USER_ID = os.getenv("COMPOSIO_USER_ID")

if not OPENAI_API_KEY:
    raise ValueError("OPENAI_API_KEY is not set in the environment")
if not COMPOSIO_API_KEY:
    raise ValueError("COMPOSIO_API_KEY is not set in the environment")
if not COMPOSIO_USER_ID:
    raise ValueError("COMPOSIO_USER_ID is not set in the environment")
```

```typescript
const OPENAI_API_KEY = process.env.OPENAI_API_KEY;
const COMPOSIO_API_KEY = process.env.COMPOSIO_API_KEY;
const COMPOSIO_USER_ID = process.env.COMPOSIO_USER_ID;

if (!OPENAI_API_KEY) throw new Error("OPENAI_API_KEY is not set");
if (!COMPOSIO_API_KEY) throw new Error("COMPOSIO_API_KEY is not set");
if (!COMPOSIO_USER_ID) throw new Error("COMPOSIO_USER_ID is not set");
```

### 6. Create a Tool Router session and build the agent function

What's happening here:
- We create a Composio client using your API key and configure it with the LlamaIndex provider
- We then create a tool router MCP session for your user, specifying the toolkits we want to use (in this case, icypeas)
- The session returns an MCP HTTP endpoint URL that acts as a gateway to all your configured tools
- LlamaIndex will connect to this endpoint to dynamically discover and use the available Icypeas tools.
- The MCP tools are mapped to LlamaIndex-compatible tools and plug them into the Agent.
```python
async def build_agent() -> ReActAgent:
    composio_client = Composio(
        api_key=COMPOSIO_API_KEY,
        provider=LlamaIndexProvider(),
    )

    session = composio_client.create(
        user_id=COMPOSIO_USER_ID,
        toolkits=["icypeas"],
    )

    mcp_url = session.mcp.url
    print(f"Composio MCP URL: {mcp_url}")

    mcp_client = BasicMCPClient(mcp_url, headers={"x-api-key": COMPOSIO_API_KEY})
    mcp_tool_spec = McpToolSpec(client=mcp_client)
    tools = await mcp_tool_spec.to_tool_list_async()

    llm = OpenAI(model="gpt-5")

    description = "An agent that uses Composio Tool Router MCP tools to perform Icypeas actions."
    system_prompt = """
    You are a helpful assistant connected to Composio Tool Router.
    Use the available tools to answer user queries and perform Icypeas actions.
    """
    return ReActAgent(tools=tools, llm=llm, description=description, system_prompt=system_prompt, verbose=True)
```

```typescript
async function buildAgent() {

  console.log(`Initializing Composio client...${COMPOSIO_USER_ID!}...`);
  console.log(`COMPOSIO_USER_ID: ${COMPOSIO_USER_ID!}...`);

  const composio = new Composio({
    apiKey: COMPOSIO_API_KEY,
    provider: new LlamaindexProvider(),
  });

  const session = await composio.create(
    COMPOSIO_USER_ID!,
    {
      toolkits: ["icypeas"],
    },
  );

  const mcpUrl = session.mcp.url;
  console.log(`Composio Tool Router MCP URL: ${mcpUrl}`);

  const server = mcp({
    url: mcpUrl,
    clientName: "composio_tool_router_with_llamaindex",
    requestInit: {
      headers: {
        "x-api-key": COMPOSIO_API_KEY!,
      },
    },
    // verbose: true,
  });

  const tools = await server.tools();

  const llm = openai({ apiKey: OPENAI_API_KEY, model: "gpt-5" });

  const agent = createAgent({
    name: "composio_tool_router_with_llamaindex",
        description : "An agent that uses Composio Tool Router MCP tools to perform actions.",
    systemPrompt:
      "You are a helpful assistant connected to Composio Tool Router."+
"Use the available tools to answer user queries and perform Icypeas actions." ,
    llm,
    tools,
  });

  return agent;
}
```

### 7. Create an interactive chat loop

No description provided.
```python
async def chat_loop(agent: ReActAgent) -> None:
    ctx = Context(agent)
    print("Type 'quit', 'exit', or Ctrl+C to stop.")

    while True:
        try:
            user_input = input("\nYou: ").strip()
        except (KeyboardInterrupt, EOFError):
            print("\nBye!")
            break

        if not user_input or user_input.lower() in {"quit", "exit"}:
            print("Bye!")
            break

        try:
            print("Agent: ", end="", flush=True)
            handler = agent.run(user_input, ctx=ctx)

            async for event in handler.stream_events():
                # Stream token-by-token from LLM responses
                if hasattr(event, "delta") and event.delta:
                    print(event.delta, end="", flush=True)
                # Show tool calls as they happen
                elif hasattr(event, "tool_name"):
                    print(f"\n[Using tool: {event.tool_name}]", flush=True)

            # Get final response
            response = await handler
            print()  # Newline after streaming
        except KeyboardInterrupt:
            print("\n[Interrupted]")
            continue
        except Exception as e:
            print(f"\nError: {e}")
```

```typescript
async function chatLoop(agent: ReturnType<typeof createAgent>) {
  const rl = readline.createInterface({ input, output });

  console.log("Type 'quit' or 'exit' to stop.");

  while (true) {
    let userInput: string;

    try {
      userInput = (await rl.question("\nYou: ")).trim();
    } catch {
      console.log("\nAgent: Bye!");
      break;
    }

    if (!userInput) {
      continue;
    }

    const lower = userInput.toLowerCase();
    if (lower === "quit" || lower === "exit") {
      console.log("Agent: Bye!");
      break;
    }

    try {
      process.stdout.write("Agent: ");

      const stream = agent.runStream(userInput);
      let finalResult: any = null;

      for await (const event of stream) {
        // The event.data contains the streamed content
        const data: any = event.data;

        // Check for streaming delta content
        if (data?.delta) {
          process.stdout.write(data.delta);
        }

        // Store final result for fallback
        if (data?.result || data?.message) {
          finalResult = data;
        }
      }

      // If no streaming happened, show the final result
      if (finalResult) {
        const answer =
          finalResult.result ??
          finalResult.message?.content ??
          finalResult.message ??
          "";
        if (answer && typeof answer === "string" && !answer.includes("[object")) {
          process.stdout.write(answer);
        }
      }

      console.log(); // New line after streaming completes
    } catch (err: any) {
      console.error("\nAgent error:", err?.message ?? err);
    }
  }

  rl.close();
}
```

### 8. Define the main entry point

What's happening here:
- We're orchestrating the entire application flow
- The agent gets built with proper error handling
- Then we kick off the interactive chat loop so you can start talking to Icypeas
```python
async def main() -> None:
    agent = await build_agent()
    await chat_loop(agent)

if __name__ == "__main__":
    # Handle Ctrl+C gracefully
    signal.signal(signal.SIGINT, lambda s, f: (print("\nBye!"), exit(0)))
    try:
        asyncio.run(main())
    except KeyboardInterrupt:
        print("\nBye!")
```

```typescript
async function main() {
  try {
    const agent = await buildAgent();
    await chatLoop(agent);
  } catch (err) {
    console.error("Failed to start agent:", err);
    process.exit(1);
  }
}

main();
```

### 9. Run the agent

When prompted, authenticate and authorise your agent with Icypeas, then start asking questions.
```bash
python llamaindex_agent.py
```

```typescript
npx ts-node llamaindex-agent.ts
```

## Complete Code

```python
import asyncio
import os
import signal
import dotenv

from composio import Composio
from composio_llamaindex import LlamaIndexProvider
from llama_index.core.agent.workflow import ReActAgent
from llama_index.core.workflow import Context
from llama_index.llms.openai import OpenAI
from llama_index.tools.mcp import BasicMCPClient, McpToolSpec

dotenv.load_dotenv()

OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
COMPOSIO_API_KEY = os.getenv("COMPOSIO_API_KEY")
COMPOSIO_USER_ID = os.getenv("COMPOSIO_USER_ID")

if not OPENAI_API_KEY:
    raise ValueError("OPENAI_API_KEY is not set")
if not COMPOSIO_API_KEY:
    raise ValueError("COMPOSIO_API_KEY is not set")
if not COMPOSIO_USER_ID:
    raise ValueError("COMPOSIO_USER_ID is not set")

async def build_agent() -> ReActAgent:
    composio_client = Composio(
        api_key=COMPOSIO_API_KEY,
        provider=LlamaIndexProvider(),
    )

    session = composio_client.create(
        user_id=COMPOSIO_USER_ID,
        toolkits=["icypeas"],
    )

    mcp_url = session.mcp.url
    print(f"Composio MCP URL: {mcp_url}")

    mcp_client = BasicMCPClient(mcp_url, headers={"x-api-key": COMPOSIO_API_KEY})
    mcp_tool_spec = McpToolSpec(client=mcp_client)
    tools = await mcp_tool_spec.to_tool_list_async()

    llm = OpenAI(model="gpt-5")
    description = "An agent that uses Composio Tool Router MCP tools to perform Icypeas actions."
    system_prompt = """
    You are a helpful assistant connected to Composio Tool Router.
    Use the available tools to answer user queries and perform Icypeas actions.
    """
    return ReActAgent(
        tools=tools,
        llm=llm,
        description=description,
        system_prompt=system_prompt,
        verbose=True,
    );

async def chat_loop(agent: ReActAgent) -> None:
    ctx = Context(agent)
    print("Type 'quit', 'exit', or Ctrl+C to stop.")

    while True:
        try:
            user_input = input("\nYou: ").strip()
        except (KeyboardInterrupt, EOFError):
            print("\nBye!")
            break

        if not user_input or user_input.lower() in {"quit", "exit"}:
            print("Bye!")
            break

        try:
            print("Agent: ", end="", flush=True)
            handler = agent.run(user_input, ctx=ctx)

            async for event in handler.stream_events():
                # Stream token-by-token from LLM responses
                if hasattr(event, "delta") and event.delta:
                    print(event.delta, end="", flush=True)
                # Show tool calls as they happen
                elif hasattr(event, "tool_name"):
                    print(f"\n[Using tool: {event.tool_name}]", flush=True)

            # Get final response
            response = await handler
            print()  # Newline after streaming
        except KeyboardInterrupt:
            print("\n[Interrupted]")
            continue
        except Exception as e:
            print(f"\nError: {e}")

async def main() -> None:
    agent = await build_agent()
    await chat_loop(agent)

if __name__ == "__main__":
    # Handle Ctrl+C gracefully
    signal.signal(signal.SIGINT, lambda s, f: (print("\nBye!"), exit(0)))
    try:
        asyncio.run(main())
    except KeyboardInterrupt:
        print("\nBye!")
```

```typescript
import "dotenv/config";
import readline from "node:readline/promises";
import { stdin as input, stdout as output } from "node:process";

import { Composio } from "@composio/core";
import { LlamaindexProvider } from "@composio/llamaindex";

import { mcp } from "@llamaindex/tools";
import { agent as createAgent } from "@llamaindex/workflow";
import { openai } from "@llamaindex/openai";

dotenv.config();

const OPENAI_API_KEY = process.env.OPENAI_API_KEY;
const COMPOSIO_API_KEY = process.env.COMPOSIO_API_KEY;
const COMPOSIO_USER_ID = process.env.COMPOSIO_USER_ID;

if (!OPENAI_API_KEY) {
    throw new Error("OPENAI_API_KEY is not set in the environment");
  }
if (!COMPOSIO_API_KEY) {
    throw new Error("COMPOSIO_API_KEY is not set in the environment");
  }
if (!COMPOSIO_USER_ID) {
    throw new Error("COMPOSIO_USER_ID is not set in the environment");
  }

async function buildAgent() {

  console.log(`Initializing Composio client...${COMPOSIO_USER_ID!}...`);
  console.log(`COMPOSIO_USER_ID: ${COMPOSIO_USER_ID!}...`);

  const composio = new Composio({
    apiKey: COMPOSIO_API_KEY,
    provider: new LlamaindexProvider(),
  });

  const session = await composio.create(
    COMPOSIO_USER_ID!,
    {
      toolkits: ["icypeas"],
    },
  );

  const mcpUrl = session.mcp.url;
  console.log(`Composio Tool Router MCP URL: ${mcpUrl}`);

  const server = mcp({
    url: mcpUrl,
    clientName: "composio_tool_router_with_llamaindex",
    requestInit: {
      headers: {
        "x-api-key": COMPOSIO_API_KEY!,
      },
    },
    // verbose: true,
  });

  const tools = await server.tools();

  const llm = openai({ apiKey: OPENAI_API_KEY, model: "gpt-5" });

  const agent = createAgent({
    name: "composio_tool_router_with_llamaindex",
    description:
      "An agent that uses Composio Tool Router MCP tools to perform actions.",
    systemPrompt:
      "You are a helpful assistant connected to Composio Tool Router."+
"Use the available tools to answer user queries and perform Icypeas actions." ,
    llm,
    tools,
  });

  return agent;
}

async function chatLoop(agent: ReturnType<typeof createAgent>) {
  const rl = readline.createInterface({ input, output });

  console.log("Type 'quit' or 'exit' to stop.");

  while (true) {
    let userInput: string;

    try {
      userInput = (await rl.question("\nYou: ")).trim();
    } catch {
      console.log("\nAgent: Bye!");
      break;
    }

    if (!userInput) {
      continue;
    }

    const lower = userInput.toLowerCase();
    if (lower === "quit" || lower === "exit") {
      console.log("Agent: Bye!");
      break;
    }

    try {
      process.stdout.write("Agent: ");

      const stream = agent.runStream(userInput);
      let finalResult: any = null;

      for await (const event of stream) {
        // The event.data contains the streamed content
        const data: any = event.data;

        // Check for streaming delta content
        if (data?.delta) {
          process.stdout.write(data.delta);
        }

        // Store final result for fallback
        if (data?.result || data?.message) {
          finalResult = data;
        }
      }

      // If no streaming happened, show the final result
      if (finalResult) {
        const answer =
          finalResult.result ??
          finalResult.message?.content ??
          finalResult.message ??
          "";
        if (answer && typeof answer === "string" && !answer.includes("[object")) {
          process.stdout.write(answer);
        }
      }

      console.log(); // New line after streaming completes
    } catch (err: any) {
      console.error("\nAgent error:", err?.message ?? err);
    }
  }

  rl.close();
}

async function main() {
  try {
    const agent = await buildAgent();
    await chatLoop(agent);
  } catch (err: any) {
    console.error("Failed to start agent:", err?.message ?? err);
    process.exit(1);
  }
}

main();
```

## Conclusion

You've successfully connected Icypeas to LlamaIndex through Composio's Tool Router MCP layer.
Key takeaways:
- Tool Router dynamically exposes Icypeas tools through an MCP endpoint
- LlamaIndex's ReActAgent handles reasoning and orchestration; Composio handles integrations
- The agent becomes more capable without increasing prompt size
- Async Python provides clean, efficient execution of agent workflows
You can easily extend this to other toolkits like Gmail, Notion, Stripe, GitHub, and more by adding them to the toolkits parameter.

## How to build Icypeas MCP Agent with another framework

- [OpenAI Agents SDK](https://composio.dev/toolkits/icypeas/framework/open-ai-agents-sdk)
- [Claude Agent SDK](https://composio.dev/toolkits/icypeas/framework/claude-agents-sdk)
- [Claude Code](https://composio.dev/toolkits/icypeas/framework/claude-code)
- [Claude Cowork](https://composio.dev/toolkits/icypeas/framework/claude-cowork)
- [Codex](https://composio.dev/toolkits/icypeas/framework/codex)
- [OpenClaw](https://composio.dev/toolkits/icypeas/framework/openclaw)
- [Hermes](https://composio.dev/toolkits/icypeas/framework/hermes-agent)
- [CLI](https://composio.dev/toolkits/icypeas/framework/cli)
- [Google ADK](https://composio.dev/toolkits/icypeas/framework/google-adk)
- [LangChain](https://composio.dev/toolkits/icypeas/framework/langchain)
- [Vercel AI SDK](https://composio.dev/toolkits/icypeas/framework/ai-sdk)
- [Mastra AI](https://composio.dev/toolkits/icypeas/framework/mastra-ai)
- [CrewAI](https://composio.dev/toolkits/icypeas/framework/crew-ai)

## Related Toolkits

- [Aeroleads](https://composio.dev/toolkits/aeroleads) - Aeroleads is a B2B lead generation platform for finding business emails and phone numbers. Grow your sales pipeline faster with powerful prospecting tools.
- [Autobound](https://composio.dev/toolkits/autobound) - Autobound is an AI-powered sales engagement platform that crafts hyper-personalized outreach and insights. It helps sales teams boost response rates and close more deals through tailored content and recommendations.
- [Better proposals](https://composio.dev/toolkits/better_proposals) - Better Proposals is a web-based tool for crafting and sending professional proposals. It helps teams impress clients and close deals faster with slick, easy-to-use templates.
- [Bidsketch](https://composio.dev/toolkits/bidsketch) - Bidsketch is a proposal software that helps businesses create professional proposals quickly and efficiently. It streamlines the proposal process, saving time while boosting client win rates.
- [Bolna](https://composio.dev/toolkits/bolna) - Bolna is an AI platform for building conversational voice agents. It helps businesses automate support and streamline interactions through natural, voice-powered conversations.
- [Botsonic](https://composio.dev/toolkits/botsonic) - Botsonic is a no-code AI chatbot builder for easily creating and deploying chatbots to your website. It empowers businesses to offer conversational experiences without writing code.
- [Botstar](https://composio.dev/toolkits/botstar) - BotStar is a comprehensive chatbot platform for designing, developing, and training chatbots visually on Messenger and websites. It helps businesses automate conversations and customer interactions without coding.
- [Callerapi](https://composio.dev/toolkits/callerapi) - CallerAPI is a white-label caller identification platform for branded caller ID and fraud prevention. It helps businesses boost customer trust while stopping spam, fraud, and robocalls.
- [Callingly](https://composio.dev/toolkits/callingly) - Callingly is a lead response management platform that automates immediate call and text follow-ups with new leads. It helps sales teams boost response speed and close more deals by connecting seamlessly with CRMs and lead sources.
- [Callpage](https://composio.dev/toolkits/callpage) - Callpage is a lead capture platform that lets businesses instantly connect with website visitors via callback. It boosts lead generation and increases your sales conversion rates.
- [Clearout](https://composio.dev/toolkits/clearout) - Clearout is an AI-powered service for verifying, finding, and enriching email addresses. It boosts deliverability and helps you discover high-quality leads effortlessly.
- [Clientary](https://composio.dev/toolkits/clientary) - Clientary is a platform for managing clients, invoices, projects, proposals, and more. It streamlines client work and saves you serious admin time.
- [Convolo ai](https://composio.dev/toolkits/convolo_ai) - Convolo ai is an AI-powered communications platform for sales teams. It accelerates lead response and improves conversion rates by automating calls and integrating workflows.
- [Delighted](https://composio.dev/toolkits/delighted) - Delighted is a customer feedback platform based on the Net Promoter System®. It helps you quickly gather, track, and act on customer sentiment.
- [Docsbot ai](https://composio.dev/toolkits/docsbot_ai) - Docsbot ai is a platform that lets you build custom AI chatbots trained on your documentation. It automates customer support and content generation, saving time and improving response quality.
- [Emelia](https://composio.dev/toolkits/emelia) - Emelia is an all-in-one B2B prospecting platform for cold-email, LinkedIn outreach, and prospect research. It streamlines outbound campaigns so you can find, engage, and warm up leads faster.
- [Findymail](https://composio.dev/toolkits/findymail) - Findymail is a B2B data provider offering verified email and phone contacts for sales prospecting. Enhance outreach with automated exports, email verification, and CRM enrichment.
- [Freshdesk](https://composio.dev/toolkits/freshdesk) - Freshdesk is customer support software with ticketing and automation tools. It helps teams streamline helpdesk operations for faster, better customer support.
- [Fullenrich](https://composio.dev/toolkits/fullenrich) - FullEnrich is a B2B contact enrichment platform that aggregates emails and phone numbers from 15+ data vendors. Instantly find and verify lead contact data to boost your outreach.
- [Gatherup](https://composio.dev/toolkits/gatherup) - GatherUp is a customer feedback and online review management platform. It helps businesses boost their reputation by streamlining how they collect and manage customer feedback.

## Frequently Asked Questions

### What are the differences in Tool Router MCP and Icypeas MCP?

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

### Can I use Tool Router MCP with LlamaIndex?

Yes, you can. LlamaIndex 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 Icypeas tools.

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

Yes, absolutely. You can configure which Icypeas 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 Icypeas data and credentials are handled as safely as possible.

---
[See all toolkits](https://composio.dev/toolkits) · [Composio docs](https://docs.composio.dev/llms.txt)