How to integrate Wrike MCP with Vercel AI SDK

BOOK A CALLGET STARTED FOR FREE
Framework Integration Gradient
Wrike Logo
Vercel AI SDK Logo
divider

Introduction

This guide walks you through connecting Wrike to Vercel AI SDK using the Composio tool router. By the end, you'll have a working Wrike agent that can create a new task in marketing folder, add multiple users to project group, invite a teammate to the workspace, delete completed tasks from design folder through natural language commands.

This guide will help you understand how to give your Vercel AI SDK agent real control over a Wrike account through Composio's Wrike MCP server.

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

TL;DR

Here's what you'll learn:
  • How to set up and configure a Vercel AI SDK agent with Wrike integration
  • Using Composio's Tool Router to dynamically load and access Wrike tools
  • Creating an MCP client connection using HTTP transport
  • Building an interactive CLI chat interface with conversation history management
  • Handling tool calls and results within the Vercel AI SDK framework

What is Vercel AI SDK?

The Vercel AI SDK is a TypeScript library for building AI-powered applications. It provides tools for creating agents that can use external services and maintain conversation state.

Key features include:

  • streamText: Core function for streaming responses with real-time tool support
  • MCP Client: Built-in support for Model Context Protocol
  • Step Counting: Control multi-step tool execution
  • OpenAI Provider: Native integration with OpenAI models

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

The Wrike MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Wrike account. It provides structured and secure access to your project spaces, so your agent can perform actions like creating tasks, managing folders, handling group memberships, sending workspace invitations, and automating project workflows on your behalf.

  • Automated task creation and management: Let your agent quickly create new tasks in specific folders, assign details, and keep your projects moving forward without manual input.
  • Dynamic folder and project organization: Have your agent generate new folders or subfolders to structure work, or clean up by deleting old folders and their contents when projects wrap up.
  • Efficient user and group management: Easily add, remove, or modify group memberships and create new user groups to keep team permissions organized and up-to-date.
  • Seamless workspace invitations: Direct your agent to invite teammates or collaborators to your Wrike workspace via email, including customizing invitation details for better onboarding.
  • Custom field and data cleanup: Empower your agent to delete custom fields, tasks, or groups when they're no longer needed, helping you maintain a clean and efficient workspace.

Supported Tools & Triggers

Tools
Bulk modify group membersAdds or removes members for multiple wrike groups in a single request; all specified user ids must correspond to existing wrike users.
Create a folderCreates a new wrike subfolder within the specified `folderid`, optionally as a project if `customitemtypeid` is given; the folder is auto-shared with its creator.
Create a groupCreates a new user group in wrike with a specified title, optionally setting members, parent group, avatar, and custom metadata.
Create invitationInvites a user to a wrike workspace by email, optionally with name, specifying either `usertypeid` or a combination of `role`/`external`; custom email subject/message available for paid accounts.
Create task in folderCreates a new task in a specified wrike folder; if setting priority with `prioritybefore` or `priorityafter`, the referenced task must be in the same folder or project.
Delete custom field by idPermanently deletes a custom field by its id; this action is irreversible and requires a valid, existing custom field id.
Delete folderPermanently deletes the folder specified by `folderid` and all its contents (e.
Delete group by idPermanently deletes a group by its `groupid`; this action is irreversible and does not affect user accounts that were members of the group.
Delete invitationPermanently deletes an existing invitation, specified by its unique `invitationid`; this action cannot be undone.
Delete taskPermanently deletes a wrike task and all its associated data by its id; this action is irreversible and the task must exist.
Fetch all tasksFetches tasks from a wrike account, allowing filtering by status, due date, and subfolder inclusion, with customizable response fields and pagination.
Get account informationRetrieves detailed wrike account information, where the response content is influenced by selected fields, account subscription, and user permissions.
Get all custom fieldsRetrieves all custom field definitions (including id, name, type, and settings) from the wrike account; this returns the definitions themselves, not their specific values on wrike items, and is useful for obtaining custom field ids.
Get contactsRetrieves a list of wrike contacts (e.
Get foldersRetrieves folders and/or projects from wrike, with filters; when using `nextpagetoken`, all other filter parameters must match the initial request.
Get specific contact informationRetrieves detailed information for a specific wrike contact using their unique `contactid`, optionally including `metadata` and `customfields` if specified in the `fields` parameter.
Get specific userRetrieves detailed information about a specific user in wrike using their unique user id.
Get task by idRetrieves read-only detailed information for a specific wrike task by its unique id, optionally allowing specification of fields to include in the response.
Launch folder blueprint asyncAsynchronously launches a new project or folder structure in wrike from a specified folder blueprint, typically returning a task id to track progress.
Launch Task Blueprint AsyncAsynchronously launches a wrike task blueprint to create tasks/projects, requiring either `super task id` (parent task) or `parent id` (parent folder/project) for placement.
List Folder BlueprintsRetrieves all account-level folder blueprints, which are templates for standardizing folder/project creation with predefined structures, custom fields, and workflows.
List space folder blueprintsLists all folder blueprints (templates for new folders/projects) within a specified wrike space, requiring a valid and accessible space id.
List space task blueprintsLists task blueprints (templates for creating tasks with consistent structures) available in a specific, accessible wrike space.
List subfolders by folder idLists subfolders (metadata only, not their contents) for an existing wrike folder specified by `folderid`, supporting recursive descent, filtering, and pagination.
List Task BlueprintsRetrieves a list of defined task blueprints (predefined task templates) from the wrike account, supporting pagination.
Update account metadataUpdates or adds custom key-value metadata to the wrike account, useful for integrations, storing app-specific data, or mapping external system identifiers.
Modify folder attributesModifies an existing wrike folder: updates title, description, parents (not root/recycle bin), sharing, metadata, custom fields/columns; restores, converts to project, or manages access roles.
Modify groupUpdates an existing wrike user group's attributes like title, members, parent, avatar, or metadata, using its `groupid` and specifying only the fields to change.
Modify taskModifies an existing wrike task by its id, allowing updates to attributes such as title, status, dates, assignees, and custom fields; `prioritybefore` and `priorityafter` are mutually exclusive, and parent folder ids for `addparents`/`removeparents` cannot be the recycle bin.
Retrieve custom field by idRetrieves a wrike custom field's detailed information (e.
Query invitationsRetrieves all active invitations in wrike, useful for viewing and auditing pending invitations or managing user onboarding.
Get group by idRetrieves detailed information for a specific wrike group using its `groupid`, optionally including 'metadata'.
Query workflowsFetches a list of all workflows with their detailed information from the wrike account; this is a read-only action and does not support pagination or filtering through its parameters.
Retrieve list of groupsRetrieves a list of user groups from the wrike account, supporting metadata filtering, pagination, and inclusion of specific fields; this is a read-only operation.
Update custom field by idUpdates properties of an existing wrike custom field by its id, such as its title, type, scope, or sharing settings.
Update invitationUpdates a pending wrike invitation (`invitationid`) to resend it or change user's role/type (use `usertypeid` over deprecated `role`/`external`).
Update metadata on a specific contactUpdates metadata, job role, or custom fields for an existing wrike contact specified by `contactid`; if `jobroleid` is provided, it must be a valid id.
Update a specific userUpdates specified profile attributes (e.

What is the Composio tool router, and how does it fit here?

What is Tool Router?

Composio's Tool Router helps agents find the right tools for a task at runtime. You can plug in multiple toolkits (like Gmail, HubSpot, and GitHub), and the agent will identify the relevant app and action to complete multi-step workflows. This can reduce token usage and improve the reliability of tool calls. Read more here: Getting started with Tool Router

The tool router generates a secure MCP URL that your agents can access to perform actions.

How the Tool Router works

The Tool Router follows a three-phase workflow:

  1. Discovery: Searches for tools matching your task and returns relevant toolkits with their details.
  2. Authentication: Checks for active connections. If missing, creates an auth config and returns a connection URL via Auth Link.
  3. Execution: Executes the action using the authenticated connection.

Step-by-step Guide

Prerequisites

Before you begin, make sure you have:
  • Node.js and npm installed
  • A Composio account with API key
  • An OpenAI API key

Getting API Keys for OpenAI and Composio

OpenAI API Key
  • Go to the OpenAI dashboard and create an API key. You'll need credits to use the models, or you can connect to another model provider.
  • Keep the API key safe.
Composio API Key
  • Log in to the Composio dashboard.
  • Navigate to your API settings and generate a new API key.
  • Store this key securely as you'll need it for authentication.

Install required dependencies

bash
npm install @ai-sdk/openai @ai-sdk/mcp @composio/core ai dotenv

First, install the necessary packages for your project.

What you're installing:

  • @ai-sdk/openai: Vercel AI SDK's OpenAI provider
  • @ai-sdk/mcp: MCP client for Vercel AI SDK
  • @composio/core: Composio SDK for tool integration
  • ai: Core Vercel AI SDK
  • dotenv: Environment variable management

Set up environment variables

bash
OPENAI_API_KEY=your_openai_api_key_here
COMPOSIO_API_KEY=your_composio_api_key_here
COMPOSIO_USER_ID=your_user_id_here

Create a .env file in your project root.

What's needed:

  • OPENAI_API_KEY: Your OpenAI API key for GPT model access
  • COMPOSIO_API_KEY: Your Composio API key for tool access
  • COMPOSIO_USER_ID: A unique identifier for the user session

Import required modules and validate environment

typescript
import "dotenv/config";
import { openai } from "@ai-sdk/openai";
import { Composio } from "@composio/core";
import * as readline from "readline";
import { streamText, type ModelMessage, stepCountIs } from "ai";
import { experimental_createMCPClient as createMCPClient } from "@ai-sdk/mcp";

const composioAPIKey = process.env.COMPOSIO_API_KEY;
const composioUserID = process.env.COMPOSIO_USER_ID;

if (!process.env.OPENAI_API_KEY) throw new Error("OPENAI_API_KEY is not set");
if (!composioAPIKey) throw new Error("COMPOSIO_API_KEY is not set");
if (!composioUserID) throw new Error("COMPOSIO_USER_ID is not set");

const composio = new Composio({
  apiKey: composioAPIKey,
});
What's happening:
  • We're importing all necessary libraries including Vercel AI SDK's OpenAI provider and Composio
  • The dotenv/config import automatically loads environment variables
  • The MCP client import enables connection to Composio's tool server

Create Tool Router session and initialize MCP client

typescript
async function main() {
  // Create a tool router session for the user
  const { session } = await composio.create(composioUserID!, {
    toolkits: ["wrike"],
  });

  const mcpUrl = session.mcp.url;
What's happening:
  • We're creating a Tool Router session that gives your agent access to Wrike tools
  • The create method takes the user ID and specifies which toolkits should be available
  • The returned mcp object contains the URL and authentication headers needed to connect to the MCP server
  • This session provides access to all Wrike-related tools through the MCP protocol

Connect to MCP server and retrieve tools

typescript
const mcpClient = await createMCPClient({
  transport: {
    type: "http",
    url: mcpUrl,
    headers: session.mcp.headers, // Authentication headers for the Composio MCP server
  },
});

const tools = await mcpClient.tools();
What's happening:
  • We're creating an MCP client that connects to our Composio Tool Router session via HTTP
  • The mcp.url provides the endpoint, and mcp.headers contains authentication credentials
  • The type: "http" is important - Composio requires HTTP transport
  • tools() retrieves all available Wrike tools that the agent can use

Initialize conversation and CLI interface

typescript
let messages: ModelMessage[] = [];

console.log("Chat started! Type 'exit' or 'quit' to end the conversation.\n");
console.log(
  "Ask any questions related to wrike, like summarize my last 5 emails, send an email, etc... :)))\n",
);

const rl = readline.createInterface({
  input: process.stdin,
  output: process.stdout,
  prompt: "> ",
});

rl.prompt();
What's happening:
  • We initialize an empty messages array to maintain conversation history
  • A readline interface is created to accept user input from the command line
  • Instructions are displayed to guide the user on how to interact with the agent

Handle user input and stream responses with real-time tool feedback

typescript
rl.on("line", async (userInput: string) => {
  const trimmedInput = userInput.trim();

  if (["exit", "quit", "bye"].includes(trimmedInput.toLowerCase())) {
    console.log("\nGoodbye!");
    rl.close();
    process.exit(0);
  }

  if (!trimmedInput) {
    rl.prompt();
    return;
  }

  messages.push({ role: "user", content: trimmedInput });
  console.log("\nAgent is thinking...\n");

  try {
    const stream = streamText({
      model: openai("gpt-5"),
      messages,
      tools,
      toolChoice: "auto",
      stopWhen: stepCountIs(10),
      onStepFinish: (step) => {
        for (const toolCall of step.toolCalls) {
          console.log(`[Using tool: ${toolCall.toolName}]`);
          }
          if (step.toolCalls.length > 0) {
            console.log(""); // Add space after tool calls
          }
        },
      });

      for await (const chunk of stream.textStream) {
        process.stdout.write(chunk);
      }

      console.log("\n\n---\n");

      // Get final result for message history
      const response = await stream.response;
      if (response?.messages?.length) {
        messages.push(...response.messages);
      }
    } catch (error) {
      console.error("\nAn error occurred while talking to the agent:");
      console.error(error);
      console.log(
        "\nYou can try again or restart the app if it keeps happening.\n",
      );
    } finally {
      rl.prompt();
    }
  });

  rl.on("close", async () => {
    await mcpClient.close();
    console.log("\nđź‘‹ Session ended.");
    process.exit(0);
  });
}

main().catch((err) => {
  console.error("Fatal error:", err);
  process.exit(1);
});
What's happening:
  • We use streamText instead of generateText to stream responses in real-time
  • toolChoice: "auto" allows the model to decide when to use Wrike tools
  • stopWhen: stepCountIs(10) allows up to 10 steps for complex multi-tool operations
  • onStepFinish callback displays which tools are being used in real-time
  • We iterate through the text stream to create a typewriter effect as the agent responds
  • The complete response is added to conversation history to maintain context
  • Errors are caught and displayed with helpful retry suggestions

Complete Code

Here's the complete code to get you started with Wrike and Vercel AI SDK:

typescript
import "dotenv/config";
import { openai } from "@ai-sdk/openai";
import { Composio } from "@composio/core";
import * as readline from "readline";
import { streamText, type ModelMessage, stepCountIs } from "ai";
import { experimental_createMCPClient as createMCPClient } from "@ai-sdk/mcp";

const composioAPIKey = process.env.COMPOSIO_API_KEY;
const composioUserID = process.env.COMPOSIO_USER_ID;

if (!process.env.OPENAI_API_KEY) throw new Error("OPENAI_API_KEY is not set");
if (!composioAPIKey) throw new Error("COMPOSIO_API_KEY is not set");
if (!composioUserID) throw new Error("COMPOSIO_USER_ID is not set");

const composio = new Composio({
  apiKey: composioAPIKey,
});

async function main() {
  // Create a tool router session for the user
  const { session } = await composio.create(composioUserID!, {
    toolkits: ["wrike"],
  });

  const mcpUrl = session.mcp.url;

  const mcpClient = await createMCPClient({
    transport: {
      type: "http",
      url: mcpUrl,
      headers: session.mcp.headers, // Authentication headers for the Composio MCP server
    },
  });

  const tools = await mcpClient.tools();

  let messages: ModelMessage[] = [];

  console.log("Chat started! Type 'exit' or 'quit' to end the conversation.\n");
  console.log(
    "Ask any questions related to wrike, like summarize my last 5 emails, send an email, etc... :)))\n",
  );

  const rl = readline.createInterface({
    input: process.stdin,
    output: process.stdout,
    prompt: "> ",
  });

  rl.prompt();

  rl.on("line", async (userInput: string) => {
    const trimmedInput = userInput.trim();

    if (["exit", "quit", "bye"].includes(trimmedInput.toLowerCase())) {
      console.log("\nGoodbye!");
      rl.close();
      process.exit(0);
    }

    if (!trimmedInput) {
      rl.prompt();
      return;
    }

    messages.push({ role: "user", content: trimmedInput });
    console.log("\nAgent is thinking...\n");

    try {
      const stream = streamText({
        model: openai("gpt-5"),
        messages,
        tools,
        toolChoice: "auto",
        stopWhen: stepCountIs(10),
        onStepFinish: (step) => {
          for (const toolCall of step.toolCalls) {
            console.log(`[Using tool: ${toolCall.toolName}]`);
          }
          if (step.toolCalls.length > 0) {
            console.log(""); // Add space after tool calls
          }
        },
      });

      for await (const chunk of stream.textStream) {
        process.stdout.write(chunk);
      }

      console.log("\n\n---\n");

      // Get final result for message history
      const response = await stream.response;
      if (response?.messages?.length) {
        messages.push(...response.messages);
      }
    } catch (error) {
      console.error("\nAn error occurred while talking to the agent:");
      console.error(error);
      console.log(
        "\nYou can try again or restart the app if it keeps happening.\n",
      );
    } finally {
      rl.prompt();
    }
  });

  rl.on("close", async () => {
    await mcpClient.close();
    console.log("\nđź‘‹ Session ended.");
    process.exit(0);
  });
}

main().catch((err) => {
  console.error("Fatal error:", err);
  process.exit(1);
});

Conclusion

You've successfully built a Wrike agent using the Vercel AI SDK with streaming capabilities! This implementation provides a powerful foundation for building AI applications with natural language interfaces and real-time feedback.

Key features of this implementation:

  • Real-time streaming responses for a better user experience with typewriter effect
  • Live tool execution feedback showing which tools are being used as the agent works
  • Dynamic tool loading through Composio's Tool Router with secure authentication
  • Multi-step tool execution with configurable step limits (up to 10 steps)
  • Comprehensive error handling for robust agent execution
  • Conversation history maintenance for context-aware responses

You can extend this further by adding custom error handling, implementing specific business logic, or integrating additional Composio toolkits to create multi-app workflows.

How to build Wrike MCP Agent with another framework

OpenAI Agents SDK

OpenAI Agents SDK

Use Wrike MCP with OpenAI Agents SDK

Claude Agent SDK

Claude Agent SDK

Use Wrike MCP with Claude Agent SDK

Google ADK

Google ADK

Use Wrike MCP with Google ADK

LangChain

LangChain

Use Wrike MCP with LangChain

Mastra AI

Mastra AI

Use Wrike MCP with Mastra AI

LlamaIndex

LlamaIndex

Use Wrike MCP with LlamaIndex

CrewAI

CrewAI

Use Wrike MCP with CrewAI

FAQ

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

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

Can I use Tool Router MCP with Vercel AI SDK?

Yes, you can. Vercel AI SDK 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 Wrike tools.

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

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

Used by agents from

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

Never worry about agent reliability

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

Book a call↗Get started for free