wrike-mcp-server

wrike-mcp-server

4

The Wrike MCP server is a lightweight bridge connecting Wrike's project management platform with language learning models, enabling AI-powered interactions with Wrike data through API endpoints. It supports tasks and comment management on Wrike via simple integrations.

wrike-mcp-server

A lightweight implementation of the Wrike MCP server to help you integrate Wrike with your favorite LLM tools.

Overview

The Wrike MCP (Model Control Protocol) server acts as a bridge between Wrike's project management platform and Language Learning Models (LLMs). It enables AI-powered tools to interact with Wrike data through a set of well-defined API endpoints.

Features

  • Fetch task details using permalink URLs
  • Retrieve tasks assigned to the authenticated user
  • Get tasks from specific folders
  • Write html comment to Wrike task specifed by given permalink url with @mentions support
  • Create a new Wrike task in the folder specified by the given permalink url.

Requirements

  • Python 3.x
  • Wrike API token
  • Required Python packages (see requirements.txt)

Setup

  1. Clone the repository

  2. Create and activate a virtual environment:

    # Create a virtual environment
    python -m venv .venv
    
    # Activate the virtual environment
    # On macOS and Linux:
    source venv/bin/activate
    
  3. Install dependencies:

    pip install -r requirements.txt
    
  4. Configure your client application (example: Cursor IDE) to work with the MCP server:
    a. Go to Settings → Cursor Settings → MCP and click Add new global MCP server

    b. Add the following configuration with the correct paths:

    {
      "mcpServers": {
        "wrike-mcp-server": {
          "command": "/Users/username/wrike-mcp-server/.venv/bin/python /Users/username/wrike-mcp-server/server.py",
          "env": {
            "WRIKE_API_TOKEN": "<YOUR WRIKE API TOKEN>"
          }
        }
      }
    }
    

    c. Enable the MCP server

Usage

The server provides several MCP tools to interact with Wrike. Here are some example prompts:

  1. Fetch a specific task:

    Please analyze the following task and provide a short summary: https://www.wrike.com/open.htm?id=12345678
    
  2. Get tasks assigned to you:

    Please gather all my Wrike tasks and build a prioritized list with links to each task. Tasks involving my boss John Doe should be treated as more important. Also, include a link to each task under the title.
    
  3. Get tasks from a specific folder:

    Please summarize all Wrike tasks from the following folder: https://www.wrike.com/open.htm?id=12345678  
    Please exclude any information about prices and budgets in your output.
    
  4. Write html comment to Wrike task:

    Please write a comment to this Wrike task with Lorem Ipsum text and mention Jhon Doe @https://www.wrike.com/open.htm?id=123456789 
    
  5. Create a new Wrike task in the folder

    Please create Wrike task in the specified folder @https://www.wrike.com/open.htm?id=123456789  with well formated lorem ipsum text 
    

License

This project is licensed under the MIT License – see the file for details.