Connect your AI Agents to X in minutes
Available tools
like_post
Like a post/tweet by its ID. Requires like.write scope.
unlike_post
Unlike a post/tweet by its ID. Removes your like from the post.
retweet
Retweet a post by its ID. Shares the post with your followers.
unretweet
Undo a retweet by its original tweet ID. Removes the retweet from your profile.
add_bookmark
Bookmark a post/tweet for later. Private bookmarks are only visible to you.
remove_bookmark
Remove a post from your bookmarks by its tweet ID.
get_bookmarks
Get your bookmarked posts. Returns up to 100 posts per page with pagination.
get_liking_users
Get users who liked a tweet. Returns up to 100 users with pagination.
get_liked_tweets
Get tweets a user has liked. Use get_me first to get your own user ID.
get_retweeters
Get users who retweeted a post. Returns up to 100 users.
get_quote_tweets
Get quote tweets of a post. Returns tweets that quote the specified tweet.
create_post
Create a new post/tweet. Max 280 chars. Supports replies (reply_to_tweet_id), quotes (quote_tweet_id), and polls (2-4 options).
delete_post
Delete a post/tweet by its ID. You can only delete your own posts.
get_post
Get a single post/tweet by its ID. Returns text, author, metrics, and referenced tweets.
search_recent_posts
Search posts from the last 7 days. Supports operators: from:user, to:user, -is:retweet, has:media, lang:en. Max 100 results per page.
follow_user
Follow a user by their user ID. Use get_user_by_username first if you only have the @handle.
unfollow_user
Unfollow a user by their user ID. Also cancels pending follow requests.
mute_user
Mute a user by their user ID. Their posts won't appear in your timeline but they can still follow you.
unmute_user
Unmute a user by their user ID. Their posts will appear in your timeline again.
get_muted_users
Get your list of muted users. Muted users' posts don't appear in your timeline.
get_user_tweets
Get a user's recent tweets by user ID. Can exclude retweets/replies. Use get_me first to get your own user ID.
get_user_mentions
Get tweets that mention a user by their user ID. Use get_me first to get mentions of yourself.
get_me
Get the authenticated user's profile including ID, username, name, bio, and metrics. Use this first to get your user ID for other operations.
get_user_by_id
Get a user's profile by their unique ID. Returns username, name, bio, follower counts, and verification status.
get_user_by_username
Get a user's profile by @username (without the @ symbol). Returns ID, name, bio, metrics, and verification status.
get_followers
Get a list of users who follow the specified user. Returns up to 1000 followers per request with pagination.
get_following
Get a list of users that the specified user follows. Returns up to 1000 accounts per request with pagination.
validate_credential
Validate X credentials. Verifies credentials during setup.
How to set up Merge Agent Handler
In an mcp.json file, add the configuration below, and restart Cursor.
Learn more in the official documentation ↗
1{
2 "mcpServers": {
3 "agent-handler": {
4 "url": "https://ah-api-develop.merge.dev/api/v1/tool-packs/{TOOL_PACK_ID}/registered-users/{REGISTERED_USER_ID}/mcp",
5 "headers": {
6 "Authorization": "Bearer yMt*****"
7 }
8 }
9 }
10}
11Open your Claude Desktop configuration file and add the server configuration below. You'll also need to restart the application for the changes to take effect.
Make sure Claude is using the Node v20+.
Learn more in the official documentation ↗
1{
2 "mcpServers": {
3 "agent-handler": {
4 "command": "npx",
5 "args": [
6 "-y",
7 "mcp-remote@latest",
8 "https://ah-api-develop.merge.dev/api/v1/tool-packs/{TOOL_PACK_ID}/registered-users/{REGISTERED_USER_ID}/mcp",
9 "--header",
10 "Authorization: Bearer ${AUTH_TOKEN}"
11 ],
12 "env": {
13 "AUTH_TOKEN": "yMt*****"
14 }
15 }
16 }
17}Open your Windsurf MCP configuration file and add the server configuration below.
Click on the refresh button in the top right of the Manage MCP server page or in the top right of the chat box in the box icon.
Learn more in the official documentation ↗
1{
2 "mcpServers": {
3 "agent-handler": {
4 "command": "npx",
5 "args": [
6 "-y",
7 "mcp-remote@latest",
8 "https://ah-api.merge.dev/api/v1/tool-packs/<tool-pack-id>/registered-users/<registered-user-id>/mcp",
9 "--header",
10 "Authorization: Bearer ${AUTH_TOKEN}"
11 ],
12 "env": {
13 "AUTH_TOKEN": "<ah-production-access-key>"
14 }
15 }
16 }
17 }In Command Palette (Cmd+Shift+P on macOS, Ctrl+Shift+P on Windows), run "MCP: Open User Configuration".
You can then add the configuration below and press "start" right under servers. Enter the auth token when prompted.
Learn more in the official documentation ↗
1{
2 "inputs": [
3 {
4 "type": "promptString",
5 "id": "agent-handler-auth",
6 "description": "Agent Handler AUTH_TOKEN", // "yMt*****" when prompt
7 "password": true
8 }
9 ],
10 "servers": {
11 "agent-handler": {
12 "type": "stdio",
13 "command": "npx",
14 "args": [
15 "-y",
16 "mcp-remote@latest",
17 "https://ah-api-develop.merge.dev/api/v1/tool-packs/{TOOL_PACK_ID}/registered-users/{REGISTERED_USER_ID}/mcp",
18 "--header",
19 "Authorization: Bearer ${input:agent-handler-auth}"
20 ]
21 }
22 }
23}FAQs on using Merge's X MCP server
FAQs on using Merge's X MCP server
What is an X (formerly Twitter) MCP server?
It’s an MCP server that lets your agents access data and functionality from X. This includes everything from fetching reposts to writing posts on behalf of a specific account.
X doesn’t offer an official MCP server, so you’ll need to use one that’s hosted and managed by a 3rd-party.
How can I use the X MCP server?
Here are a few common agentic use cases:
- Social listening and brand monitoring: Track mentions of your company, executives, products, and competitors. Then summarize what’s trending and driving sentiment
- Customer support triage (public): Detect posts that look like support requests, pull the full thread context, and draft a response or route it to the right team
- Incident detection and escalation: Spot early outage reports and recurring error strings, then compile examples and escalate to on-call with links and a short summary
- Product feedback mining: Extract feature requests and pain points at scale, cluster them by theme, and quantify frequency to inform roadmap discussions
- Competitive intelligence: Monitor competitor announcements and customer reactions, then produce a weekly digest of notable launches, positioning changes, and “why users care.”
What makes Merge Agent Handler’s X MCP server better than alternative X MCP servers?
Here are a few reasons:
- Enterprise-grade security platform: Agent Handler includes a Security Gateway with Data Loss Protection (DLP) that scans tool inputs and outputs for sensitive data and can block, redact, or mask fields based on configured rules. This enables proactive governance rather than relying solely on model behavior
- Real-time observability with searchable logs: Every tool call and underlying API request is logged, creating a complete audit trail. Logs are reviewable and searchable for debugging, auditing, and optimization
- Built-in authentication and credential management: Agent Handler dynamically validates authentication, prompts users when credentials are missing, and supports both shared and individual authentication per connector
- Evaluation Suite before production deployment: You can define evaluation scenarios to validate tool outputs against expected results before going live, helping improve hit rates and reliability
Can I set custom security rules for X tool calls in Merge Agent Handler?
Yes. Agent Handler includes default security rules and allows you to build and test custom Data Loss Protection (DLP) rules. These rules can block, redact, or mask sensitive data in tool inputs and outputs.
Here are some examples:
- Sensitive-data leak protection (DLP): If your agent tries to include an API key, token, or customer email in a post or reply, Agent Handler can block the tool call and alert your team
- Account safety and identity control: Your can require approval before posting from an exec handle or replying to high-visibility threads
- Blast-radius limits for automation: You can rate-limit replies, cap posts per hour, and block repeated near-duplicate content
- Topic and compliance guardrails: You can block tool calls that include restricted claims (legal, regulated, pricing guarantees) and route them for review
- High-risk action escalation: You can block deleting posts or mass-blocking accounts unless a human approves
How can I start using Merge Agent Handler’s X MCP server?
Follow these steps:
1. Create (or open) an Agent Handler account and go to the dashboard where you can manage connectors and tool packs.
2. Create or choose a Tool Pack for the agent experience you are building (this is the bundle of tools your agent will be allowed to call).
3. Add the X MCP server / connector to that Tool Pack and select the specific tools you want enabled (start with read-only tools, then add write tools as needed).
4. Set up authentication. Decide whether X access should be shared auth (admin-managed credential reused across users) or individual auth (each end user connects their own X account). Then have the relevant user(s) complete the connection flow so tool calls can run under the right identity.
5. Configure security rules and monitoring. Turn on the preconfigured security rules and add any custom rules you need for X actions (for example, preventing sensitive data from being posted).
6. Get the MCP endpoint for your Tool Pack and registered user. You’ll use the Agent Handler MCP URL pattern that includes your {TOOL_PACK_ID} and {REGISTERED_USER_ID}.
7. Connect an MCP client to Agent Handler. Configure your MCP client (for example, Claude Desktop, Cursor, VS Code, or Windsurf) with the MCP URL plus the required auth header/token, then refresh tools in the client.
8. Test in a safe mode first. Run a few “read” prompts (e.g., fetch recent posts) and verify the logs show the right arguments and responses. Then try a small “write” action (e.g., draft a post to a private test account), and confirm your rules behave as expected.
Ready to try it out?
Whether you're an engineer experimenting with agents or a product manager looking to add tools, you can get started for free now
