FreeFeed MCP Server
MCP server for the FreeFeed social network API. Enables reading timelines, posts, comments, and attachments, as well as writing posts, comments, and managing subscriptions.
README
FreeFeed MCP Server
MCP server for the FreeFeed API, a social network that replaces FriendFeed.
Features
Read
get_timeline- Get a user timeline or the home feedget_directs- Get direct (private) posts timelineget_post- Get a specific post with commentsget_post_attachments- Get attachment list with URLssearch_posts- Search posts with advanced operatorsget_user_profile- Get user profile infoget_user_subscriptions- Get subscriptions/subscribersget_my_groups- Get user's groupsget_group_timeline- Get group postsget_group_info- Get group info
Write
upload_attachment- Upload files (images, video, etc.)download_attachment- Download post attachments (returns image content when possible)get_attachment_image- Download attachment as image content with URL fallbackcreate_direct_post- Send a direct post to one or more recipientsleave_direct- Leave a direct post threadcreate_post- Create a new post (auto uploads files, supports posting to groups)update_post- Edit a postdelete_post- Delete a postlike_post/unlike_post- Likesadd_comment- Add a commentupdate_comment/delete_comment- Manage commentssubscribe_user/unsubscribe_user- Subscribe/unsubscribe
Privacy & Opt-out
Users can opt out of AI analysis by:
- Adding one of these tags to their profile description: #noai, #opt-out-ai, #no-bots, #ai-free
- Setting their account to private
- Contacting the server maintainer to be added to the manual opt-out list
The server automatically excludes:
- Private accounts (isPrivate: "1")
- Paused accounts (isGone: true)
- Users with opt-out tags in their profile
Opt-out filtering is disabled by default. Enable it with configuration:
FREEFEED_OPTOUT_ENABLED=trueFREEFEED_OPTOUT_USERS=oneuser,anotheruserFREEFEED_OPTOUT_TAGS=#noai,#opt-out-ai,#no-bots,#ai-freeFREEFEED_OPTOUT_RESPECT_PRIVATE=trueFREEFEED_OPTOUT_RESPECT_PAUSED=trueFREEFEED_OPTOUT_CONFIG=/path/to/opt_out.json
Example opt-out config file:
{
"enabled": true,
"manual_opt_out": ["someuser"],
"tags": ["#noai", "#opt-out-ai", "#no-bots", "#ai-free"],
"respect_private": true,
"respect_paused": true
}
Installation
- Install dependencies:
pip install -e .
- Create a
.envfile:
FREEFEED_BASE_URL=https://freefeed.net
FREEFEED_API_VERSION=4
FREEFEED_APP_TOKEN=your_app_token
# or
FREEFEED_USERNAME=your_username
FREEFEED_PASSWORD=your_password
Usage
Start server
python -m freefeed_mcp_server
REST API server
# Start HTTP API server
python -m freefeed_mcp_server.api
# Or with uvicorn
uvicorn freefeed_mcp_server.api:app --reload
The API will be available at http://localhost:8000
📚 API documentation: see API.md 🌐 Swagger UI: http://localhost:8000/docs
Claude Desktop integration
Add to Claude Desktop config (claude_desktop_config.json):
{
"mcpServers": {
"freefeed": {
"command": "python",
"args": ["-m", "freefeed_mcp_server"],
"env": {
"FREEFEED_BASE_URL": "https://freefeed.net",
"FREEFEED_API_VERSION": "4",
"FREEFEED_APP_TOKEN": "your_app_token"
}
}
}
}
Replace `/path/to/venv/bin/python` with your actual virtualenv path. On macOS it is often
`./.venv/bin/python` if you created a local venv in the project directory.
Claude Desktop integration (system Python)
Use the system Python executable:
{
"mcpServers": {
"freefeed": {
"command": "/usr/bin/python3",
"args": ["-m", "freefeed_mcp_server"],
"env": {
"FREEFEED_BASE_URL": "https://freefeed.net",
"FREEFEED_API_VERSION": "4",
"FREEFEED_APP_TOKEN": "your_app_token"
}
}
}
}
Claude Desktop integration (virtualenv)
Use the Python executable from your virtualenv:
{
"mcpServers": {
"freefeed": {
"command": "/path/to/venv/bin/python",
"args": ["-m", "freefeed_mcp_server"],
"env": {
"FREEFEED_BASE_URL": "https://freefeed.net",
"FREEFEED_API_VERSION": "4",
"FREEFEED_APP_TOKEN": "your_app_token"
}
}
}
}
### MCP tool examples
```json
{
"name": "get_directs",
"arguments": {
"limit": 20
}
}
{
"name": "create_direct_post",
"arguments": {
"body": "Hi from MCP!",
"recipients": ["alice", "bob"]
}
}
{
"name": "leave_direct",
"arguments": {
"post_id": "POST_ID"
}
}
{
"name": "get_attachment_image",
"arguments": {
"attachment_url": "https://freefeed.net/attachments/ATTACHMENT_ID",
"max_bytes": 2000000
}
}
{
"name": "download_attachment",
"arguments": {
"attachment_url": "https://freefeed.net/attachments/ATTACHMENT_ID",
"prefer_image": true,
"max_bytes": 2000000
}
}
## Request examples
### Get home feed
```json
{
"name": "get_timeline",
"arguments": {
"timeline_type": "home"
}
}
Search posts
{
"name": "search_posts",
"arguments": {
"query": "intitle:MCP from:username"
}
}
Create a post
{
"name": "create_post",
"arguments": {
"body": "Testing the FreeFeed MCP server!"
}
}
Create a post with an image
{
"name": "create_post",
"arguments": {
"body": "Check out this photo!",
"attachment_paths": ["/path/to/image.jpg"]
}
}
Create a post in a group
{
"name": "create_post",
"arguments": {
"body": "A post for my group!",
"group_names": ["mygroup"]
}
}
Get group posts
{
"name": "get_group_timeline",
"arguments": {
"group_name": "mygroup",
"limit": 20
}
}
Get post attachments and download
{
"name": "get_post_attachments",
"arguments": {
"post_id": "post-id-here"
}
}
Then download:
{
"name": "download_attachment",
"arguments": {
"attachment_url": "https://freefeed.net/attachments/att-id",
"save_path": "/path/to/save/image.jpg"
}
}
FreeFeed API
API docs: https://github.com/FreeFeed/freefeed-server/wiki/API
Supported search operators:
intitle:query- search in post bodyincomment:query- search in commentsfrom:username- search by authorAND,OR- logical operators
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。