# How to integrate Classmarker MCP with LlamaIndex

```json
{
  "title": "How to integrate Classmarker MCP with LlamaIndex",
  "toolkit": "Classmarker",
  "toolkit_slug": "classmarker",
  "framework": "LlamaIndex",
  "framework_slug": "llama-index",
  "url": "https://composio.dev/toolkits/classmarker/framework/llama-index",
  "markdown_url": "https://composio.dev/toolkits/classmarker/framework/llama-index.md",
  "updated_at": "2026-05-12T10:06:19.859Z"
}
```

## Introduction

This guide walks you through connecting Classmarker to LlamaIndex using the Composio tool router. By the end, you'll have a working Classmarker agent that can add student to biology exam access list, create a new question for math quiz, delete user account for withdrawn student through natural language commands.
This guide will help you understand how to give your LlamaIndex agent real control over a Classmarker account through Composio's Classmarker MCP server.
Before we dive in, let's take a quick look at the key ideas and tools involved.

## Also integrate Classmarker with

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

The Classmarker MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Classmarker account. It provides structured and secure access to your quiz management tools, so your agent can create tests, manage users and groups, add questions, and control access codes—without manual intervention.
- Automated user and group management: Let your agent create new users, add them to groups, or delete users and groups for streamlined participant organization.
- Dynamic question and category creation: Instruct your agent to add new questions or categories to your exams, helping you build tests faster and keep content organized.
- Access code and permissions control: Enable your agent to generate, assign, or delete access codes for specific exams, giving or revoking test access instantly as needed.
- Test link and API key management: Allow your agent to manage test links or revoke API keys to maintain secure and up-to-date exam distribution.
- Efficient data cleanup: Ask your agent to remove users, groups, test links, or access codes, keeping your Classmarker account tidy and up to date with minimal effort.

## Supported Tools

| Tool slug | Name | Description |
|---|---|---|
| `CLASSMARKER_CREATE_ACCESS_LIST_ITEM` | Create Access List Item | Tool to add one or more access codes to an access list. Use after obtaining a valid access_list_id to grant exam access. |
| `CLASSMARKER_CREATE_CATEGORY` | Create a new question category | Tool to create a new question category. Use after confirming the parent category ID. |
| `CLASSMARKER_CREATE_GROUP` | Create Group | Tool to create a new group. Use when you need to organize users into a specific group after planning your user structure. |
| `CLASSMARKER_CREATE_QUESTION` | Create Question | Tool to create a new question with specified text, type, and category in ClassMarker. Supports multiplechoice, multipleresponse, truefalse, and essay question types. |
| `CLASSMARKER_CREATE_USER` | Create ClassMarker User | Tool to create a new user in ClassMarker. Use after gathering user details to provision accounts. Provide user info and optional group assignments. Returns the complete API response. |
| `CLASSMARKER_DELETE_ACCESS_LIST_ITEM` | Delete Access List Item | Removes one or more access codes from a ClassMarker access list. Access lists are used to control exam access by requiring users to provide unique identifiers (e.g., email addresses, employee IDs, student numbers). Each code can be used a limited number of times per exam, and codes are recorded with exam results. To use this action: 1. First call GET_ALL_GROUPS_LINKS_EXAMS to find available access list IDs 2. Provide the access_list_id and the list of codes to remove 3. Up to 100 codes can be removed per request Note: This action is idempotent - deleting codes that don't exist will not cause an error. |
| `CLASSMARKER_DELETE_API_KEY` | Delete API Key | Tool to delete an API key by its ID. Use when you need to revoke an API key immediately. |
| `CLASSMARKER_DELETE_GROUP` | Delete Group | Tool to delete a group by its ID. Use when you need to remove a group from ClassMarker. |
| `CLASSMARKER_DELETE_TEST_LINK` | Delete Test Link | Deletes a specific link assignment from a test/exam in ClassMarker. Links are unique URLs that provide access to tests. This action is idempotent - deleting a non-existent link returns success. Use this when you need to remove a link from a test to revoke access via that specific URL. |
| `CLASSMARKER_DELETE_USER` | Delete User | Tool to delete a specific user by ID. Use when you need to remove a user from ClassMarker after verifying the user's identity. |
| `CLASSMARKER_DELETE_WEBHOOK` | Delete Webhook | Deletes a webhook listener from your ClassMarker account. This action removes a webhook configuration that was previously set up to receive real-time exam results. The ClassMarker API is idempotent - attempting to delete a non-existent webhook will return success without error. Webhook IDs are typically obtained from the ClassMarker web interface (My Account > Webhooks / API Keys) or through other webhook management tools if available. Note: This is a destructive operation and cannot be undone through the API. |
| `CLASSMARKER_GET_GROUP_DETAILS` | Get Group Details | Retrieve detailed information about a specific group by ID. Returns group name and assigned tests. Note: This fetches all accessible groups and filters for the specified group_id, as ClassMarker API doesn't provide a direct single-group endpoint. |
| `CLASSMARKER_GET_INITIAL_FINISHED_AFTER_TIMESTAMP` | Get Initial Finished After Timestamp | Compute the initial finishedAfterTimestamp for paginating ClassMarker test results. Use this action when you need to retrieve recent test results for the first time and don't have a previous 'next_finished_after_timestamp' value. This returns a Unix timestamp set to 2 weeks ago, which is the ClassMarker API's default starting point for result pagination. **When to use this:** - Starting a new pagination sequence for test results - You don't have a stored 'next_finished_after_timestamp' from a previous request **When NOT to use this:** - You have a 'next_finished_after_timestamp' from a previous response - use that value instead - For subsequent pagination requests - use the timestamp from the last successful response **Important:** ClassMarker API requires the finishedAfterTimestamp to be less than 3 months old. This action ensures compliance by setting it to 2 weeks ago. |
| `CLASSMARKER_GET_QUESTION` | Get Question | Tool to retrieve a specific question by its ID. Use when you need detailed information of a question after confirming the question_id. |
| `CLASSMARKER_GET_RECENT_RESULTS_GROUP_EXAM` | Get Recent Results For Group Exam | Tool to fetch recent test results for a specific group and exam. Use after determining a UNIX timestamp to retrieve only results finished after that time for the group and exam. |
| `CLASSMARKER_GET_RECENT_RESULTS_LINK_EXAM` | Get Recent Results Link Exam | Fetch recent exam results for a specific link and test combination. Use CLASSMARKER_GET_ALL_GROUPS_LINKS_EXAMS first to obtain valid link_id and test_id values. This endpoint returns paginated results based on finishedAfterTimestamp for efficient incremental data retrieval. Supports up to 200 results per request. |
| `CLASSMARKER_GET_TEST_DETAILS` | Get Test Details | Retrieves detailed information for a specific test (exam) including its name and all assignment contexts. Returns a list of groups and/or links where the test is assigned, with their IDs, names, and URL identifiers. Use when you need to find where a specific test is deployed or to get the complete deployment context for a test. Returns 'no_results' status if test ID doesn't exist. |
| `CLASSMARKER_GET_USER_DETAILS` | Get User Details | Tool to retrieve detailed information about a specific user. Use after you have the user's ID and need the full user profile. |
| `CLASSMARKER_LIST_CATEGORIES` | List Question Categories | Retrieves all question categories organized in a hierarchical structure. Returns parent categories with their nested sub-categories. Use this to browse available categories before creating questions or filtering content. |
| `CLASSMARKER_LIST_CERTIFICATES` | List Certificates | Tool to list all certificates. Use when you need to retrieve all certificates available in the account. |
| `CLASSMARKER_LIST_QUESTIONS` | List Questions | Retrieve a paginated list of questions from your question bank. Returns up to 200 questions per page. Use the optional 'page' parameter to navigate through multiple pages. Each question includes type, content, points, options, correct answers, and metadata. |
| `CLASSMARKER_LIST_RECENT_RESULTS_FOR_GROUPS` | List Recent Results For Groups | Retrieve recent exam results for all groups in your ClassMarker account. Returns results from the last 2 weeks by default, or from a specified timestamp (must be within 3 months). Use this to fetch user performance data including scores, pass/fail status, and certificates. |
| `CLASSMARKER_LIST_RECENT_RESULTS_FOR_LINKS` | List Recent Results for Links | Retrieves recent test results from ALL links accessible to your API key. Returns results completed after a specified timestamp (defaults to 2 weeks ago). Supports pagination for large result sets (max 200 per request). Use the next_finished_after_timestamp from the response for subsequent requests to fetch newer results without duplicates. Ideal for syncing or monitoring exam completions across all your test links. |
| `CLASSMARKER_LIST_USERS` | List Users | Tool to list all users. Use when you need to retrieve every user in your account after confirming authentication. |
| `CLASSMARKER_LIST_WEBHOOKS` | List Webhooks | Tool to retrieve all configured webhooks. Use when you need to programmatically list your webhook configurations. |
| `CLASSMARKER_PUT_CATEGORY` | Update Sub-Category | Tool to update an existing sub-category. Use when you need to rename or move a sub-category after confirming its IDs. |
| `CLASSMARKER_PUT_PARENT_CATEGORY` | Update an existing parent category | Tool to update an existing parent category. Use after confirming the parent category ID. |
| `CLASSMARKER_PUT_QUESTION` | Update Question | Updates an existing question in the ClassMarker question bank. IMPORTANT CONSTRAINTS: - Cannot change the question_type of an existing question (must match original) - random_answers is only valid for multiplechoice and multipleresponse types - grade_style is only valid for multipleresponse type - options and correct_options are required for multiplechoice, multipleresponse, and truefalse - essay questions cannot have options, correct_options, or grade_style Set verify_only=True to validate changes without applying them. |

## Supported Triggers

None listed.

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

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

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

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 Classmarker 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, classmarker)
- 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 Classmarker 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=["classmarker"],
    )

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

  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 Classmarker 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 Classmarker
```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 Classmarker, 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=["classmarker"],
    )

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

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

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

## Related Toolkits

- [Canvas](https://composio.dev/toolkits/canvas) - Canvas is a learning management system for online courses, assignments, grading, and collaboration. It's trusted by educators and students to streamline virtual classrooms and enhance digital learning.
- [Accredible certificates](https://composio.dev/toolkits/accredible_certificates) - Accredible Certificates is a platform for creating and managing digital certificates, badges, and blockchain credentials. It streamlines issuing, tracking, and verifying professional achievements for organizations of any size.
- [Api bible](https://composio.dev/toolkits/api_bible) - API.Bible is a developer platform for Scripture content and passage search. Easily integrate Bible verses and translations into your apps or chatbots.
- [Blackboard](https://composio.dev/toolkits/blackboard) - Blackboard is a digital learning platform for higher education and schools, offering tools to manage courses, track engagement, and deliver interactive content. It helps institutions improve student outcomes through actionable analytics and in-app guidance.
- [Certifier](https://composio.dev/toolkits/certifier) - Certifier is a platform for creating, managing, and issuing digital certificates and credentials. Organizations use it to automate and secure the entire credentialing process.
- [Coassemble](https://composio.dev/toolkits/coassemble) - Coassemble is a flexible platform for building, managing, and delivering online training courses. It helps teams streamline onboarding, upskilling, and ongoing learning for employees or partners.
- [D2lbrightspace](https://composio.dev/toolkits/d2lbrightspace) - D2L Brightspace is a learning management system for delivering and managing online courses and assessments. It helps educators streamline digital teaching, assignments, and communication with students.
- [Dictionary api](https://composio.dev/toolkits/dictionary_api) - Dictionary api is the Merriam-Webster API providing rich dictionary and thesaurus data for developers. Instantly access definitions, synonyms, etymologies, and audio pronunciations in your apps.
- [Google Classroom](https://composio.dev/toolkits/google_classroom) - Google Classroom is a free web service for educators and students to manage assignments and communication. It streamlines classroom collaboration and grading, making teaching simpler and more connected.
- [Lessonspace](https://composio.dev/toolkits/lessonspace) - Lessonspace is an online collaborative classroom platform offering video, whiteboards, and real-time interaction for educators and students. It streamlines remote teaching with integrated tools for engagement and communication.
- [Linguapop](https://composio.dev/toolkits/linguapop) - Linguapop is a web platform for administering language placement tests in English, German, Spanish, Italian, and French. It helps schools and organizations efficiently manage multilingual assessments and analyze results.
- [Memberspot](https://composio.dev/toolkits/memberspot) - Memberspot is an online course and video-hosting platform for business learning. It helps teams manage, deliver, and track knowledge efficiently.
- [Membervault](https://composio.dev/toolkits/membervault) - Membervault is a platform for hosting courses, memberships, and digital products in one place. It helps you build stronger relationships with your audience by centralizing digital offers and customer engagement.
- [Gmail](https://composio.dev/toolkits/gmail) - Gmail is Google's email service with powerful spam protection, search, and G Suite integration. It keeps your inbox organized and makes communication fast and reliable.
- [Google Calendar](https://composio.dev/toolkits/googlecalendar) - Google Calendar is a time management service for scheduling meetings, events, and reminders. It streamlines personal and team organization with integrated notifications and sharing options.
- [Google Drive](https://composio.dev/toolkits/googledrive) - Google Drive is a cloud storage platform for uploading, sharing, and collaborating on files. It's perfect for keeping your documents accessible and organized across devices.
- [Outlook](https://composio.dev/toolkits/outlook) - Outlook is Microsoft's email and calendaring platform for unified communications and scheduling. It helps users stay organized with powerful email, contacts, and calendar management.
- [Twitter](https://composio.dev/toolkits/twitter) - Twitter is a social media platform for sharing real-time updates, conversations, and news. Stay connected, informed, and engaged with communities worldwide.
- [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.
- [Supabase](https://composio.dev/toolkits/supabase) - Supabase is an open-source backend platform offering scalable Postgres databases, authentication, storage, and real-time APIs. It lets developers build modern apps without managing infrastructure.

## Frequently Asked Questions

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

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

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

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

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