Twitter MCP Server

A Model Context Protocol (MCP) server that provides Twitter functionality using the twikit library. This server allows AI assistants to interact with Twitter through a standardized protocol with cookie-based authentication - the LLM model provides ct0 and auth_token cookies directly in tool calls.

Features

• Cookie Authentication: LLM model provides ct0 and auth_token cookies directly in tool calls
• Session Caching: Automatically caches authenticated sessions for efficiency
• Timeline Access: Get tweets from your timeline
• User Information: Retrieve user profiles and statistics
• Tweet Search: Search for tweets with specific queries
• Tweet Management: Post, like, and retweet tweets
• User Tweets: Get tweets from specific users
• Direct Messaging: Send DMs, get DM history, react to messages, and delete messages
• Authentication Testing: Test cookies before use
• Tweet Operations: Post tweets, like/unlike tweets, retweet/delete retweets, bookmark tweets
• Tweet Retrieval: Get tweets by ID, search tweets, get user timelines, get tweet replies
• User Operations: Follow/unfollow users, get user information, search users
• Trending Topics: Get trending topics across different categories (trending, news, sports, entertainment, for-you)

Disclaimer

This project utilizes an unofficial API to interact with X (formerly Twitter) through the twikit library. The methods employed for authentication and data retrieval are not officially endorsed by X/Twitter and may be subject to change or discontinuation without notice.

This tool is intended for educational and experimental purposes only. Users should be aware of the potential risks associated with using unofficial APIs, including but not limited to account restrictions or suspension. The developers of this project are not responsible for any misuse or consequences arising from the use of this tool.

Installation

1. Clone this repository:

git clone <repository-url>
cd twitter-mcp

2. Install dependencies:

pip install -r requirements.txt

3. Run the server:

python server.py

Authentication

The server expects the LLM model to provide Twitter cookies directly in each tool call via the ct0 and auth_token parameters. No pre-configuration is required!

Getting Twitter Cookies

The LLM model will need to provide both Twitter cookies. Here's how to obtain them:

1. Open your browser and go to Twitter/X
2. Log in to your account
3. Open Developer Tools (F12)
4. Go to Application/Storage → Cookies → twitter.com (or x.com)
5. Find and copy these cookie values:
   • ct0 - CSRF token cookie
   • auth_token - Authentication token cookie

Both cookies are required for all operations.

Usage

Available Tools

1. Authenticate

Test authentication with cookies:

{
  "tool": "authenticate",
  "arguments": {
    "ct0": "your_ct0_cookie_here",
    "auth_token": "your_auth_token_cookie_here"
  }
}

2. Tweet

Post a new tweet:

{
  "tool": "tweet",
  "arguments": {
    "text": "Hello from MCP! 🚀",
    "ct0": "your_ct0_cookie_here",
    "auth_token": "your_auth_token_cookie_here"
  }
}

3. Get User Info

Get information about a Twitter user:

{
  "tool": "get_user_info",
  "arguments": {
    "username": "elonmusk",
    "ct0": "your_ct0_cookie_here",
    "auth_token": "your_auth_token_cookie_here"
  }
}

4. Search Tweets

Search for tweets:

{
  "tool": "search_tweets",
  "arguments": {
    "query": "artificial intelligence",
    "count": 10,
    "ct0": "your_ct0_cookie_here",
    "auth_token": "your_auth_token_cookie_here"
  }
}

5. Get Timeline

Get tweets from your timeline:

{
  "tool": "get_timeline",
  "arguments": {
    "count": 20,
    "ct0": "your_ct0_cookie_here",
    "auth_token": "your_auth_token_cookie_here"
  }
}

6. Like Tweet

Like a tweet by ID: ``

... View full README on GitHub