# How to integrate ReferralRock MCP with LlamaIndex

```json
{
  "title": "How to integrate ReferralRock MCP with LlamaIndex",
  "toolkit": "ReferralRock",
  "toolkit_slug": "referralrock",
  "framework": "LlamaIndex",
  "framework_slug": "llama-index",
  "url": "https://composio.dev/toolkits/referralrock/framework/llama-index",
  "markdown_url": "https://composio.dev/toolkits/referralrock/framework/llama-index.md",
  "updated_at": "2026-03-29T06:47:32.628Z"
}
```

## Introduction

This guide walks you through connecting ReferralRock to LlamaIndex using the Composio tool router. By the end, you'll have a working ReferralRock agent that can list all active referral campaigns, add a new member to referral program, show referral stats for this week through natural language commands.
This guide will help you understand how to give your LlamaIndex agent real control over a ReferralRock account through Composio's ReferralRock MCP server.
Before we dive in, let's take a quick look at the key ideas and tools involved.

## Also integrate ReferralRock with

- [OpenAI Agents SDK](https://composio.dev/toolkits/referralrock/framework/open-ai-agents-sdk)
- [Claude Agent SDK](https://composio.dev/toolkits/referralrock/framework/claude-agents-sdk)
- [Claude Code](https://composio.dev/toolkits/referralrock/framework/claude-code)
- [Claude Cowork](https://composio.dev/toolkits/referralrock/framework/claude-cowork)
- [Codex](https://composio.dev/toolkits/referralrock/framework/codex)
- [OpenClaw](https://composio.dev/toolkits/referralrock/framework/openclaw)
- [Hermes](https://composio.dev/toolkits/referralrock/framework/hermes-agent)
- [CLI](https://composio.dev/toolkits/referralrock/framework/cli)
- [Google ADK](https://composio.dev/toolkits/referralrock/framework/google-adk)
- [LangChain](https://composio.dev/toolkits/referralrock/framework/langchain)
- [Vercel AI SDK](https://composio.dev/toolkits/referralrock/framework/ai-sdk)
- [Mastra AI](https://composio.dev/toolkits/referralrock/framework/mastra-ai)
- [CrewAI](https://composio.dev/toolkits/referralrock/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 ReferralRock
- Connect LlamaIndex to the ReferralRock MCP server
- Build a ReferralRock-powered agent using LlamaIndex
- Interact with ReferralRock 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 ReferralRock MCP server, and what's possible with it?

The ReferralRock MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your ReferralRock account. It provides structured and secure access so your agent can perform ReferralRock operations on your behalf.

## Supported Tools

| Tool slug | Name | Description |
|---|---|---|
| `REFERRALROCK_DELETE_EMAIL_REMOVEUNSUBSCRIBE` | Remove Email from Unsubscribe List | Tool to remove an email address from the unsubscribe list. Use when you need to resubscribe an email address that was previously unsubscribed. |
| `REFERRALROCK_DELETE_HOOKS` | Delete Webhook | Tool to delete a webhook subscription. Use when you need to cancel and stop data transmission to a registered webhook URL. |
| `REFERRALROCK_DELETE_MEMBERS_REMOVE` | Delete Members from Program | Tool to delete members from a referral program. Use when you need to remove one or more members. Each member requires either valid primaryInfo (memberId or referralCode) OR both secondaryInfo (externalIdentifier or email) and tertiaryInfo (program details) to be uniquely identified. |
| `REFERRALROCK_DELETE_REFERRAL_REMOVE` | Delete Referral | Tool to delete one or more referrals from a referral program. Use when you need to remove referrals by their unique ID or other identifying information. Each referral is deleted only when it can be uniquely identified. |
| `REFERRALROCK_DELETE_REWARDS` | Delete Rewards | Tool to delete rewards from members or referrals in a referral program. Use when you need to remove one or more rewards from the system. |
| `REFERRALROCK_GET_EMAIL_UNSUBSCRIBED_STATUS` | Get Email Unsubscribed Status | Tool to check if an email address is unsubscribed from communications. Use when you need to verify unsubscribe status before sending emails. |
| `REFERRALROCK_GET_WEBHOOK_HISTORY_FOR_EVENT` | Get Webhook History for Event | Tool to retrieve webhook history for a specific event. Use when you need to check the delivery status and details of webhooks for a particular event type. |
| `REFERRALROCK_LIST_PROGRAM_MEMBERS` | List Program Members | Tool to list program members with filtering and pagination. Use when you need to retrieve member information, optionally filtering by program, search query, status, or date range. Supports sorting and pagination for large datasets. |
| `REFERRALROCK_GET_MEMBER_STATISTICS` | Get Member Statistics | Tool to retrieve comprehensive statistics for a single member from ReferralRock. Use when you need to analyze a member's sharing, referral, and reward activities over a specified time period. |
| `REFERRALROCK_GET_PAYOUT` | Get Payout | Tool to retrieve payout list by ID. Use when you need to get specific payout information using a payout identifier. |
| `REFERRALROCK_GET_PAYOUT_DETAILS_BY_ID` | Get Payout Details by ID | Tool to get payout details by ID. Use when you need to retrieve specific payout information using a payout identifier. |
| `REFERRALROCK_GET_PENDING_PAYOUTS` | Get Pending Payouts | Tool to get pending payouts for a member or recipient. Use when you need to retrieve pending payout information, optionally filtering by recipient and including ineligible rewards. |
| `REFERRALROCK_GET_PAYOUT_TRANSACTIONS` | Get Payout Transactions | Tool to get payout transaction history for referral program participants. Use when you need to retrieve payout transaction information, optionally filtering by recipient ID or specific transaction ID. |
| `REFERRALROCK_GET_SINGLE_PROGRAM` | Get Single Program | Tool to retrieve a single referral program by name. Use when you need details about a specific program. |
| `REFERRALROCK_LIST_REFERRAL_PROGRAMS` | List Referral Programs | List all referral programs in ReferralRock. Use this to retrieve details about available referral programs with pagination support. |
| `REFERRALROCK_GET_REFERRALS` | Get Referrals | Tool to list referrals with filtering and pagination options. Use when you need to retrieve referrals from ReferralRock, optionally filtered by program, member, date range, status, or search query. Supports sorting and pagination. |
| `REFERRALROCK_GET_SINGLE_REFERRAL` | Get Single Referral | Tool to retrieve a single referral by email, ID, or external ID. Use when you need to fetch detailed information about a specific referral from ReferralRock. |
| `REFERRALROCK_GET_REWARD_RULES` | Get Reward Rules | Tool to get reward rules for a program. Use when you need to retrieve member rewards rules configuration including triggers, payouts, and delivery settings for a specific program. |
| `REFERRALROCK_GET_REWARDS` | Get Rewards | Tool to list rewards with optional filtering and pagination. Use when you need to retrieve rewards filtered by program, member, status, date ranges, or specific search queries. Supports sorting and pagination via offset/count parameters. |
| `REFERRALROCK_DELETE_EMAIL_REMOVEUNSUBSCRIBE` | Remove Email from Unsubscribe List | Tool to remove an email address from the unsubscribe list. Use when you need to resubscribe a previously unsubscribed email address to allow them to receive emails again. |
| `REFERRALROCK_UNSUBSCRIBE_EMAIL_ADDRESS` | Unsubscribe Email Address | Tool to unsubscribe an email address from the ReferralRock mailing list. Use when a user wants to opt out of email communications. |
| `REFERRALROCK_CREATE_WEBHOOK_SUBSCRIPTION` | Create Webhook Subscription | Tool to create or update a webhook subscription for event notifications. Use when you need to register a callback URL to receive POST requests with JSON data when specified events occur. |
| `REFERRALROCK_CREATE_INVITE_FEED_ENTRY` | Create Invite Feed Entry | Tool to create an invite feed entry for automatic invitations. Use when you need to add a contact to the referral program's invite queue. |
| `REFERRALROCK_CREATE_INVITE_FEED_ENTRIES` | Create Invite Feed Entries | Tool to create multiple invite feed entries for automatic invites in ReferralRock. Use when you need to send a set of contacts to be used with automatic member invitations. |
| `REFERRALROCK_GENERATE_MEMBER_ACCESS_TOKEN` | Generate Member Access Token | Tool to generate pre-authenticated member access URLs for sharing and portal experiences. Use when you need to create temporary access tokens for members to access their portal or share content. Note: This endpoint is deprecated; consider using memberaccessurls endpoint instead. |
| `REFERRALROCK_GENERATE_MEMBER_ACCESS_URLS` | Generate Member Access URLs | Tool to generate member access URLs for custom share experience or secure portal embedding. Use when you need to create unique access and share links for members with optional expiration time. |
| `REFERRALROCK_ADD_NEW_MEMBER_TO_PROGRAM` | Add new member to program | Tool to create a new member and add them to a referral program. Use when you need to add a new member with their program ID, name, and email. Returns HTTP 201 (Created) on success with member confirmation details. |
| `REFERRALROCK_REMOVE_MEMBERS_FROM_PROGRAM` | Remove Members from Program | Tool to remove members from a referral program. Use when you need to delete members from the system. A member is removed when uniquely identified by either primaryInfo (memberId or referralCode) OR both secondaryInfo and tertiaryInfo. |
| `REFERRALROCK_CREATE_REFERRAL_ACTION` | Create Referral Action | Tool to create a referral action for processing recurring rewards. Use when you need to record a referral action with an amount value for calculating percentage-based recurring rewards. The action can be linked to referrals via email, ID, or external identifier. |
| `REFERRALROCK_REMOVE_REFERRAL` | Remove Referral | Tool to remove referrals from a referral program. Use when you need to delete referrals using their referral ID or alternative identifiers. |
| `REFERRALROCK_CREATE_NEW_REFERRAL` | Create New Referral | Tool to create a new referral in ReferralRock based on a member's referral code. Use when you need to add a referral to a program, providing details like contact information, transaction amount, and custom fields. |
| `REFERRALROCK_UPDATE_REFERRAL_STATUS` | Update Referral Status | Tool to update referral status in ReferralRock. Use when you need to change a referral's status to pending, qualified, approved, or denied. Note: This endpoint is deprecated; use POST /api/referral/update instead. |
| `REFERRALROCK_UPDATE_REFERRAL_DETAILS` | Update Referral Details | Tool to update referral details in a referral program. Use when you need to modify existing referral information such as contact details, status, notes, or custom fields. Requires either a referralId OR both secondaryInfo and tertiaryInfo to identify referrals. Supports bulk updates by providing multiple update items. |
| `REFERRALROCK_ISSUE_REWARD_TO_MEMBER` | Issue Reward to Member | Tool to issue a reward to a member in the Referral Rock system. Use when you need to issue a specific reward to a recipient, with the option to override eligibility date rules. |
| `REFERRALROCK_REMOVE_REWARDS` | Remove Rewards | Tool to remove rewards from members or referrals of a referral program. Use when you need to delete one or more rewards by their IDs. The endpoint returns a result for each reward ID indicating whether the removal was successful. |

## Supported Triggers

None listed.

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

The ReferralRock MCP server is an implementation of the Model Context Protocol that connects your AI agent to ReferralRock. It provides structured and secure access so your agent can perform ReferralRock 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 ReferralRock account and project
- Basic familiarity with async Python/Typescript

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

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 ReferralRock 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, referralrock)
- 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 ReferralRock 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=["referralrock"],
    )

    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 ReferralRock actions."
    system_prompt = """
    You are a helpful assistant connected to Composio Tool Router.
    Use the available tools to answer user queries and perform ReferralRock 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: ["referralrock"],
    },
  );

  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 ReferralRock 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 ReferralRock
```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 ReferralRock, 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=["referralrock"],
    )

    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 ReferralRock actions."
    system_prompt = """
    You are a helpful assistant connected to Composio Tool Router.
    Use the available tools to answer user queries and perform ReferralRock 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: ["referralrock"],
    },
  );

  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 ReferralRock 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 ReferralRock to LlamaIndex through Composio's Tool Router MCP layer.
Key takeaways:
- Tool Router dynamically exposes ReferralRock 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 ReferralRock MCP Agent with another framework

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

## Related Toolkits

- [Google Sheets](https://composio.dev/toolkits/googlesheets) - Google Sheets is a cloud-based spreadsheet tool for real-time collaboration and data analysis. It lets teams work together from anywhere, updating information instantly.
- [Notion](https://composio.dev/toolkits/notion) - Notion is a collaborative workspace for notes, docs, wikis, and tasks. It streamlines team knowledge, project tracking, and workflow customization in one place.
- [Airtable](https://composio.dev/toolkits/airtable) - Airtable combines the flexibility of spreadsheets with the power of a database for easy project and data management. Teams use Airtable to organize, track, and collaborate with custom views and automations.
- [Asana](https://composio.dev/toolkits/asana) - Asana is a collaborative work management platform for teams to organize and track projects. It streamlines teamwork, boosts productivity, and keeps everyone aligned on goals.
- [Google Tasks](https://composio.dev/toolkits/googletasks) - Google Tasks is a to-do list and task management tool integrated into Gmail and Google Calendar. It helps you organize, track, and complete tasks across your Google ecosystem.
- [Linear](https://composio.dev/toolkits/linear) - Linear is a modern issue tracking and project planning tool for fast-moving teams. It helps streamline workflows, organize projects, and boost productivity.
- [Jira](https://composio.dev/toolkits/jira) - Jira is Atlassian’s platform for bug tracking, issue tracking, and agile project management. It helps teams organize work, prioritize tasks, and deliver projects efficiently.
- [Clickup](https://composio.dev/toolkits/clickup) - ClickUp is an all-in-one productivity platform for managing tasks, docs, goals, and team collaboration. It streamlines project workflows so teams can work smarter and stay organized in one place.
- [Monday](https://composio.dev/toolkits/monday) - Monday.com is a customizable work management platform for project planning and collaboration. It helps teams organize tasks, automate workflows, and track progress in real time.
- [Reddit](https://composio.dev/toolkits/reddit) - Reddit is a social news platform with thriving user-driven communities (subreddits). It's the go-to place for discussion, content sharing, and viral marketing.
- [Facebook](https://composio.dev/toolkits/facebook) - Facebook is a social media and advertising platform for businesses and creators. It helps you connect, share, and manage content across your public Facebook Pages.
- [Linkedin](https://composio.dev/toolkits/linkedin) - LinkedIn is a professional networking platform for connecting, sharing content, and engaging with business opportunities. It's the go-to place for building your professional brand and unlocking new career connections.
- [Active campaign](https://composio.dev/toolkits/active_campaign) - ActiveCampaign is a marketing automation and CRM platform for managing email campaigns, sales pipelines, and customer segmentation. It helps businesses engage customers and drive growth through smart automation and targeted outreach.
- [ActiveTrail](https://composio.dev/toolkits/active_trail) - ActiveTrail is a user-friendly email marketing and automation platform. It helps you reach subscribers and automate campaigns with ease.
- [Addressfinder](https://composio.dev/toolkits/addressfinder) - Addressfinder is a data quality platform for verifying addresses, emails, and phone numbers. It helps you ensure accurate customer and contact data every time.
- [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.
- [Agiled](https://composio.dev/toolkits/agiled) - Agiled is an all-in-one business management platform for CRM, projects, and finance. It helps you streamline workflows, consolidate client data, and manage business processes in one place.
- [Ahrefs](https://composio.dev/toolkits/ahrefs) - Ahrefs is an SEO and marketing platform for site audits, keyword research, and competitor insights. It helps you improve search rankings and drive organic traffic.
- [Amcards](https://composio.dev/toolkits/amcards) - AMCards lets you create and mail personalized greeting cards online. Build stronger customer relationships with easy, automated card campaigns.
- [Ascora](https://composio.dev/toolkits/ascora) - Ascora is a cloud-based field service management platform for service businesses. It streamlines scheduling, invoicing, and customer operations in one place.

## Frequently Asked Questions

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

With a standalone ReferralRock MCP server, the agents and LLMs can only access a fixed set of ReferralRock tools tied to that server. However, with the Composio Tool Router, agents can dynamically load tools from ReferralRock 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 ReferralRock tools.

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

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

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