threads-mcp
MCP server for the Threads API, enabling profile management, content reading, publishing, replies, and discovery through 26 tools.
README
threads-mcp
A Model Context Protocol (MCP) server for the Threads API. Exposes 26 tools covering profile management, content reading, publishing, replies, and discovery.
Quick Start
1. Get a Threads Access Token
You can start using Threads API immediately for your own account via the Threads Tester — no App Review required for self-use.
- Go to developers.facebook.com/tools/explorer
- Select your Threads App from the app dropdown (or create one at developers.facebook.com/apps with the Threads product)
- In the Threads Tester, generate a short-lived token
- To get a long-lived token (60 days), call:
curl "https://graph.threads.net/access_token?grant_type=th_exchange_token&client_id=YOUR_APP_ID&client_secret=YOUR_APP_SECRET&access_token=SHORT_LIVED_TOKEN"
- To refresh a long-lived token before it expires:
curl "https://graph.threads.net/refresh_access_token?grant_type=th_refresh_token&access_token=LONG_LIVED_TOKEN"
Note: Long-lived tokens expire after 60 days. Refresh them at least once per 60 days. There is no refresh token — you must always refresh before expiry.
2. Set Environment Variables
export THREADS_ACCESS_TOKEN=your_access_token_here
# Optional: skip the GET /me call at startup by providing your user ID directly
export THREADS_USER_ID=your_threads_user_id
3. Run the Server
# Install globally
npm install -g threads-mcp
# Or run directly
npx threads-mcp
Or build from source:
git clone https://github.com/your-repo/threads-mcp
cd threads-mcp
pnpm install
pnpm build
THREADS_ACCESS_TOKEN=your_token node dist/index.js
Claude Desktop Configuration
Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):
{
"mcpServers": {
"threads": {
"command": "npx",
"args": ["-y", "threads-mcp"],
"env": {
"THREADS_ACCESS_TOKEN": "your_access_token_here"
}
}
}
}
Or if installed globally:
{
"mcpServers": {
"threads": {
"command": "threads-mcp",
"env": {
"THREADS_ACCESS_TOKEN": "your_access_token_here",
"THREADS_USER_ID": "your_user_id_optional"
}
}
}
}
Cursor Configuration
Add to .cursor/mcp.json in your project root, or ~/.cursor/mcp.json globally:
{
"mcpServers": {
"threads": {
"command": "npx",
"args": ["-y", "threads-mcp"],
"env": {
"THREADS_ACCESS_TOKEN": "your_access_token_here"
}
}
}
}
Tools Reference
Profile Tools
| Tool | Description | Required Permission |
|---|---|---|
get_profile |
Get your Threads profile (username, bio, follower count) | threads_basic |
get_publishing_limit |
Check remaining publishing quota (250 posts/24h) | threads_content_publish |
get_account_insights |
Account-level analytics (views, likes, followers) | threads_read_replies or threads_manage_insights |
Content Tools
| Tool | Description | Required Permission |
|---|---|---|
list_my_threads |
Paginated list of your posts | threads_basic |
list_my_replies |
Paginated list of your replies | threads_basic |
get_thread |
Get details of a specific post | threads_basic |
get_thread_insights |
Engagement metrics for a post | threads_manage_insights |
get_mentions |
Posts where you've been @mentioned | threads_basic |
Publishing Tools
| Tool | Description | Required Permission |
|---|---|---|
create_text_post |
Create/publish a text post | threads_content_publish |
create_image_post |
Create/publish a post with an image | threads_content_publish |
create_video_post |
Create a video post container | threads_content_publish |
create_carousel_post |
Create a multi-media carousel post | threads_content_publish |
publish_thread |
Publish a container created with publish:false |
threads_content_publish |
get_container_status |
Check if a video container finished processing | threads_content_publish |
delete_thread |
Delete one of your posts | threads_content_publish |
repost_thread |
Repost another user's thread | threads_content_publish |
quote_thread |
Quote-post another thread with your commentary | threads_content_publish |
Replies Tools
| Tool | Description | Required Permission |
|---|---|---|
get_replies |
Get direct replies to a post | threads_read_replies |
get_conversation |
Get full conversation thread | threads_read_replies |
reply_to_thread |
Post a reply to a thread | threads_content_publish |
manage_reply |
Hide or unhide a reply on your posts | threads_manage_replies |
Discovery Tools
| Tool | Description | Required Permission |
|---|---|---|
search_keyword |
Search posts by keyword | threads_keyword_search ⚠️ |
lookup_profile |
Look up a user profile by username | threads_profile_discovery ⚠️ |
list_profile_posts |
List posts from any public user | threads_basic |
search_locations |
Search for location tags | threads_basic |
get_location |
Get details about a location | threads_basic |
⚠️ App Review required for
threads_keyword_searchandthreads_profile_discovery. These permissions require Meta's App Review process before they can be used in production. For self-use development, you may be able to use them in Development mode with your own account only.
Publishing Workflow
Text Post (simplest)
Tool: create_text_post
text: "Hello, Threads!"
publish: true ← publishes immediately
Video Post (requires polling)
1. Tool: create_video_post
video_url: "https://..."
publish: false ← don't auto-publish; get container ID
2. Tool: get_container_status
container_id: "from step 1"
→ poll until status = FINISHED
3. Tool: publish_thread
container_id: "from step 1"
Carousel Post
1. Tool: create_carousel_post
items: [{type: IMAGE, url: "..."}, {type: IMAGE, url: "..."}]
publish: true ← auto-publish carousel
Rate Limits
- Publishing: 250 posts per 24 hours (use
get_publishing_limitto check) - API rate limits: Follow standard Meta API rate limits (~200 calls/hour per user for most endpoints)
- Video processing: Videos typically take 10–30 seconds; poll with
get_container_status
Security
- The
THREADS_ACCESS_TOKENis never logged or included in tool responses - All outgoing URLs with the token are built internally and never surfaced to the MCP layer
- Error messages from the API that may reference the token URL are automatically redacted
Development
pnpm install
pnpm test # run all tests
pnpm test:watch # watch mode
pnpm typecheck # TypeScript check
pnpm build # build dist/
Running locally
THREADS_ACCESS_TOKEN=your_token pnpm build && node dist/index.js
Verify tool count
echo '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}' \
| THREADS_ACCESS_TOKEN=test node dist/index.js 2>/dev/null \
| python3 -c "import sys,json; d=json.load(sys.stdin); print(len(d['result']['tools']), 'tools')"
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.