Feather MCP Server
Enables AI agents to automate Facebook lead generation and CRM tasks such as searching leads, sending friend requests, managing contacts, broadcasting messages, and tracking growth metrics.
README
Feather MCP Server
Connect AI agents to Facebook lead generation and CRM automation.
An MCP (Model Context Protocol) server that exposes Feather Connector and Feather Messenger CRM tools to any AI agent. Search leads, manage contacts, send broadcasts, and track growth metrics through a universal plugin interface.
Quick Start
Zero-install via npx:
npx feather-mcp-server
Or install and build locally:
npm install
npm run build
node dist/index.js
Test with the MCP Inspector:
npx @modelcontextprotocol/inspector node dist/index.js
Demo Mode
Try all 10 tools without installing the Chrome extension!
The server runs in demo mode by default, returning realistic mock data (leads, contacts, stats, growth metrics) so you can explore every tool immediately. Set
FEATHER_DEMO=falsewhen you're ready to connect to the live Feather extensions.
| Variable | Default | Description |
|---|---|---|
FEATHER_DEMO |
true |
Use mock data (true) or connect to live extensions (false) |
Works With
- Claude Desktop — add to
claude_desktop_config.json - Cursor — add as an MCP server in settings
- VS Code — via MCP extension
- Windsurf — add as an MCP server
- OpenClaw — add to your OpenClaw config
- Any MCP client — standard stdio transport
Tools
Feather Connector (Facebook Automation)
| Tool | Description |
|---|---|
search_leads |
Search for potential leads by keywords, groups, or interests |
send_friend_request |
Send a friend request with an optional note |
get_request_status |
Check status of sent friend requests (pending/accepted/declined) |
get_connector_stats |
Get overall statistics: sent, accepted, declined, accept rate |
Feather Messenger CRM
| Tool | Description |
|---|---|
list_contacts |
List CRM contacts with optional label, tag, or search filters |
add_label |
Add a label to a contact |
add_tag |
Add a tag to a contact |
send_broadcast |
Send a broadcast message to all contacts with a given label |
get_crm_stats |
Get aggregate CRM stats: contact count, label distribution, recent activity |
Analytics
| Tool | Description |
|---|---|
get_growth_metrics |
Get growth metrics (connections, messages, engagement) over 7d/30d/90d |
Example Usage
Search for leads
{
"tool": "search_leads",
"arguments": {
"query": "marketing",
"limit": 5,
"source": "groups"
}
}
Send a broadcast
{
"tool": "send_broadcast",
"arguments": {
"label": "warm",
"message": "New webinar this Thursday! Link in bio.",
"dryRun": true
}
}
Check growth
{
"tool": "get_growth_metrics",
"arguments": {
"metric": "connections",
"period": "30d"
}
}
Integration
Claude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"feather": {
"command": "npx",
"args": ["-y", "feather-mcp-server"]
}
}
}
OpenClaw
Add to your OpenClaw config:
mcp:
servers:
feather:
command: npx
args:
- -y
- feather-mcp-server
Architecture
src/
index.ts Entry point (stdio transport)
server.ts MCP server setup + tool registration
types.ts TypeScript types
utils.ts Helpers (demo mode check, JSON results, error handling)
tools/
connector.ts Feather Connector tools (4 tools)
messenger.ts Feather Messenger CRM tools (5 tools)
analytics.ts Analytics tools (1 tool)
mock/
data.ts Realistic demo data (leads, contacts, stats, growth series)
Tech Stack
- TypeScript + Node.js
@modelcontextprotocol/sdk(official MCP SDK)zodfor input validation- stdio transport (MCP standard)
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。