About Twitter Client MCP
A Model Context Protocol server that gives clients access to Twitter operations through the `agent-twitter-client` library. It covers profile and tweet actions, search, and follower relationships, with credential-handling protections.
Key features
- 01Retrieve profiles and tweets
- 02Post, like, and retweet
- 03Search tweets and profiles
- 04List followers and following
- 05Follow users
- 06Isolate and sanitize credentials
Use cases
- 01Search for tweets or Twitter profiles
- 02Post a tweet or interact with a tweet
- 03Review a user's followers and following
MCP support
Exposes Twitter profile, tweet, search, and relationship tools to MCP clients. Twitter credentials are loaded by the server and protected from logging and LLM access.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
npx -y github:mzkrasner/twitter-client-mcpTWITTER_USERNAMETWITTER_PASSWORDSecretTWITTER_EMAILLocal servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMETwitter Client MCP
A Model Context Protocol (MCP) server that provides secure access to Twitter functionality through the agent-twitter-client library.
Features
The Twitter Client MCP provides the following capabilities:
-
Profile Operations
- Get profile information by username
- Get the authenticated user's profile
-
Tweet Operations
- Get a specific tweet by ID
- Get tweets from a user
- Send new tweets
- Like tweets
- Retweet tweets
-
Search Operations
- Search for tweets
- Search for profiles
-
Relationship Operations
- Get followers of a user
- Get users that a user is following
- Follow a user
Security ⚠️
IMPORTANT: TWITTER CREDENTIAL PROTECTION
This MCP server requires Twitter credentials to operate. To protect these sensitive credentials:
- NEVER share your credentials
- NEVER run commands that display your credentials
- NEVER allow the LLM to execute shell commands directly without your approval
Multiple Layers of Protection
This server implements several layers of security to keep your credentials safe:
1. Credential Isolation
- Your credentials are only loaded during initialization
- After loading, credentials are immediately removed from environment variables
- Credentials are never logged or transmitted to the LLM
2. Memory Protection
- Secure memory allocation using sodium-native
- Memory locking to prevent swapping to disk
- Zero-out memory buffers after use
3. Access Prevention
- Secure environment variable handling
- Strict validation of required environment variables
- Console output sanitization to prevent leaking secrets
Setup
- Clone the repository
- Install dependencies:
npm install - Build the project:
npm run build
Required Environment Variables
The Twitter Client MCP requires the following environment variables:
TWITTER_USERNAME=your_username
TWITTER_PASSWORD=your_password
[email protected]
Optionally, you can add Twitter API v2 credentials for advanced functionality:
TWITTER_API_KEY=your_api_key
TWITTER_API_SECRET_KEY=your_api_secret
TWITTER_ACCESS_TOKEN=your_access_token
TWITTER_ACCESS_TOKEN_SECRET=your_access_secret
Tools
The server exposes the following MCP tools:
| Tool Name | Description | Parameters |
|---|---|---|
profileByUsername | Get a Twitter profile by username | username: String |
myProfile | Get the authenticated user's profile | check: Boolean |
getTweet | Get a specific tweet by ID | tweetId: String |
getUserTweets | Get tweets from a Twitter user | username: String, count?: Number |
sendTweet | Post a new tweet | text: String, inReplyToId?: String |
likeTweet | Like a tweet | tweetId: String |
retweet | Retweet a tweet | tweetId: String |
searchTweets | Search for tweets | query: String, count?: Number, searchMode?: String ('top', 'latest', 'photos', 'videos') |
searchProfiles | Search for Twitter profiles | query: String, count?: Number |
getFollowers | Get a list of users following a Twitter user | username: String, count?: Number |
getFollowing | Get a list of users that a Twitter user is following | username: String, count?: Number |
followUser | Follow a Twitter user | username: String |
Usage
Running Locally
# Start the server with environment variables
TWITTER_USERNAME=your_username TWITTER_PASSWORD=your_password [email protected] npm start
Adding to Cursor
To add this MCP server to Cursor:
- In Cursor, go to Settings > MCP Servers
- Click "Add Server"
- Configure the server with the following settings:
- Name:
Twitter Client MCP(or any name you prefer) - Type:
command - Command:
node - Arguments:
/path/to/twitter-client-mcp/dist/index.js(replace with your actual path) - Environment Variables:
TWITTER_USERNAME: Your Twitter usernameTWITTER_PASSWORD: Your Twitter passwordTWITTER_EMAIL: Your Twitter email- Any other variables you want to set
- Name:
- Click "Save"
Using NPX (Recommended)
You can also use npx to run the MCP server directly from GitHub:
TWITTER_USERNAME=your_username TWITTER_PASSWORD=your_password [email protected] npx -y github:mzkrasner/twitter-client-mcp
Using Environment Variables in Cursor Configuration
For more security and ease of use, configure Cursor via the .cursor/mcp.json file in your home directory:
{
"mcpServers": {
"twitter-client-mcp": {
"command": "npx",
"args": [
"-y",
"github:mzkrasner/twitter-client-mcp"
],
"env": {
"TWITTER_USERNAME": "your_username",
"TWITTER_PASSWORD": "your_password",
"TWITTER_EMAIL": "[email protected]"
}
}
}
}
Development
Important Note for Development
When developing the MCP server, use console.error() instead of console.log() for all debugging and logging statements. The MCP protocol communicates with the client via stdout, so any console.log() statements will interfere with this communication.
License
MIT
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.