SocialData is a remote MCP server that gives an AI agent 14 tools, among them check_is_following, get_list_tweets, get_tweet. It answers at https://api.socialdata.tools/mcp over streamable HTTP, and it listed its tools to an anonymous client, so no sign-in was needed for that. Our probes reached it 6 of 6 times since 12 September 2026; the last check, on 24 October 2026, succeeded.
Twitter (X) API alternative for AI agents: tweet search, profiles, followers. $0.0002 per result.
Description as the publisher wrote it.
| Tool | What it does |
|---|---|
check_is_following | Check whether one Twitter/X user follows another. Far cheaper than fetching a follower list — use this whenever the question is about a specific pair of accounts. Both IDs must be numeric. Inputs: source_user_id, target_user_id |
get_list_tweets | Fetch one page of posts from a Twitter/X list, newest first. A page is large — around 100 posts — plus a next_cursor. Billing is per result, so one call here costs several times a single-user timeline call; do not page through a busy list u Inputs: list_id, cursor |
get_tweet | Fetch a single Twitter/X post by its numeric ID, including engagement counts. Bills one post request against the caller's balance. Inputs: tweet_id |
get_tweet_comments | Fetch one page of replies to a Twitter/X post. Returns about 20 replies plus a next_cursor. The first item of a page can repeat the last item of the previous page, so de-duplicate by post id when paging. Each call bills against the caller's Inputs: tweet_id, cursor |
get_tweet_thread | Fetch one page of a Twitter/X thread — the chain of follow-up posts the author wrote under their own post ("1/", "2/", ...). Give it the ID of any post in the chain and it returns that chain in order. It does NOT return what other people re Inputs: tweet_id, cursor |
get_user_by_id | Look up a Twitter/X user profile by numeric user ID. Bills one user request against the caller's balance. Inputs: user_id |
get_user_by_username | Look up a Twitter/X user profile by username (handle). Returns the profile including follower and following counts. Bills one user request against the caller's balance. Inputs: username |
get_user_followers | Fetch one page of a user's followers. This is one of the most expensive tools: a page is set upstream and can run to ~100 user records, every one of which is billed. Never page through a large account's followers without the user explicitly Inputs: user_id, cursor |
get_user_following | Fetch one page of the accounts a user follows. Expensive: a page is set upstream and can run to ~100 user records, every one of which is billed. Never page through a large account's following list without the user explicitly asking. Require Inputs: user_id, cursor |
get_user_mentions | Fetch one page of posts mentioning a given user, newest first. Returns about 20 posts plus a next_cursor. The first item of a page can repeat the last item of the previous page, so de-duplicate by post id when paging. Each call bills agains Inputs: username, cursor |
get_user_tweets | Fetch one page of a user's own posts (excluding replies), newest first. Returns about 20 posts plus a next_cursor. If the account has a pinned post it is prepended to every page, so de-duplicate by post id when paging. Each call bills again Inputs: user_id, cursor |
get_user_tweets_and_replies | Fetch one page of a user's posts including their replies, newest first. Returns about 20 items plus a next_cursor. If the account has a pinned post it is prepended to every page, so de-duplicate by post id when paging. Each call bills again Inputs: user_id, cursor |
search_tweets | Search Twitter/X posts. Returns one page of about 20 results plus a next_cursor, and bills one search request against the caller's balance. The query supports Twitter's advanced search operators, and using them is almost always better than Inputs: query, type, cursor |
search_users | Search Twitter/X user profiles by name or handle. Returns one page of matching profiles and bills against the caller's balance. Use this when you know roughly who someone is but not their exact handle. Inputs: query |
Every block below points at https://api.socialdata.tools/mcp. Pick your client and copy.
claude mcp add --transport http socialdata https://api.socialdata.tools/mcp
{
"mcpServers": {
"socialdata": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://api.socialdata.tools/mcp"]
}
}
}{
"mcpServers": {
"socialdata": {
"url": "https://api.socialdata.tools/mcp"
}
}
}{
"servers": {
"socialdata": {
"type": "http",
"url": "https://api.socialdata.tools/mcp"
}
}
}{
"context_servers": {
"socialdata": {
"url": "https://api.socialdata.tools/mcp"
}
}
}{
"mcpServers": {
"socialdata": {
"serverUrl": "https://api.socialdata.tools/mcp"
}
}
}{
"mcpServers": {
"socialdata": {
"url": "https://api.socialdata.tools/mcp"
}
}
}name: SocialData
version: 0.0.1
schema: v1
mcpServers:
- name: SocialData
type: streamable-http
url: https://api.socialdata.tools/mcp{
"mcpServers": {
"socialdata": {
"httpUrl": "https://api.socialdata.tools/mcp"
}
}
}{
"mcpServers": {
"socialdata": {
"type": "http",
"url": "https://api.socialdata.tools/mcp"
}
}
}{
"mcpServers": {
"socialdata": {
"url": "https://api.socialdata.tools/mcp"
}
}
}goose://extension?url=https%3A%2F%2Fapi.socialdata.tools%2Fmcp&type=streamable_http&id=socialdata&name=SocialData
It answered our last check on 24 October 2026 in 170 ms. Over 6 checks since 12 September 2026 it answered 6. Status here comes from Neuronto's own probes, not from the publisher.
Yes. It is published by socialdata.tools and served from api.socialdata.tools, the vendor's own domain: https://api.socialdata.tools/mcp.
14, as returned by the server's own tools/list on 23 September 2026: check_is_following, get_list_tweets, get_tweet, get_tweet_comments, get_tweet_thread, get_user_by_id, get_user_by_username, get_user_followers, and 6 more.
It answered our last check on 24 October 2026 in 170 ms. Over 6 checks since 12 September 2026 it answered 6.
Put the server under mcpServers in ~/.cursor/mcp.json with the URL https://api.socialdata.tools/mcp. The exact JSON is above, with the same setup for Claude Code, Claude Desktop, VS Code, Windsurf, Zed, Gemini CLI and the rest.
It listed its tools to an anonymous client, so reading the tool list needs no key. Individual tools can still ask for credentials in their own inputs.
Publisher: tools.socialdata. Endpoint host: api.socialdata.tools. ARD identifier: urn:air:tools.socialdata:mcp:socialdata.
First indexed 10 September 2026. The tools are exactly what the server returned; nothing on this page
is inferred from its name. Ranking in Neuronto search is never for sale.