pipedriver
MCP server for Pipedrive API v2, enabling management of deals, persons, organizations, products, activities, and pipelines through natural language in MCP clients like Cursor and Claude Desktop.
README
pipedriver
MCP server for the Pipedrive API v2. Expose Pipedrive (deals, persons, organizations, products, activities, pipelines, and more) as tools to any Model Context Protocol client—Cursor, Claude Desktop, and others.
Built with the official Pipedrive JavaScript SDK and the MCP TypeScript SDK.
Quick start (npx)
No install. Run the server with your Pipedrive API token:
PIPEDRIVE_API_KEY=your_api_token npx pipedriver
Then point your MCP client at this command. Example for Cursor (Settings → MCP → Edit config):
{
"mcpServers": {
"pipedriver": {
"command": "npx",
"args": ["-y", "pipedriver"],
"env": {
"PIPEDRIVE_API_KEY": "your_api_token"
}
}
}
}
Get an API token in Pipedrive → Settings → API.
Usage
How it runs
The server uses stdio transport: your MCP client starts it as a subprocess and talks over stdin/stdout. No HTTP server or port.
Option A: npx (recommended for trying it out)
npx -y pipedriver
Set PIPEDRIVE_API_KEY in the environment (or in your MCP config env, as in the example above).
Option B: Install and run
npm install -g pipedriver
# or: yarn global add pipedriver
export PIPEDRIVE_API_KEY=your_api_token
pipedriver
Option C: From a local clone (development)
git clone https://github.com/vaibhavpandeyvpz/pipedriver.git
cd pipedriver
yarn install
export PIPEDRIVE_API_KEY=your_api_token
yarn dev
In your MCP config, use the path to the repo:
{
"mcpServers": {
"pipedriver": {
"command": "node",
"args": ["/path/to/pipedriver/dist/index.js"],
"env": { "PIPEDRIVE_API_KEY": "your_api_token" }
}
}
}
Or without building (runs TypeScript with tsx):
{
"mcpServers": {
"pipedriver": {
"command": "npx",
"args": ["tsx", "/path/to/pipedriver/src/index.ts"],
"env": { "PIPEDRIVE_API_KEY": "your_api_token" }
}
}
}
Tool naming
Tools follow pipedrive_<domain>_<action>, e.g. pipedrive_deals_list, pipedrive_persons_add. Arguments and request bodies match the Pipedrive API v2 (filters, sort, pagination, and JSON bodies where applicable).
Tools overview
| Domain | Examples |
|---|---|
| Activities | add, delete, list, get, update |
| Activity fields | get, list (metadata) |
| Deals | add, delete, list, get, update, search; products (add/delete/update/many); followers; convert to lead; additional discounts; installments |
| Deal fields | get, list (metadata) |
| Item search | search, search by field |
| Leads | search, convert to deal, get conversion status |
| Organizations | add, delete, list, get, update, search; followers |
| Organization fields | get, list (metadata) |
| Persons | add, delete, list, get, update, search; picture; followers |
| Person fields | get, list (metadata) |
| Pipelines | add, delete, list, get, update |
| Products | add, delete, list, get, update, search; followers; duplicate; variations; image |
| Product fields | get, list (metadata) |
| Stages | add, delete, list, get, update |
| Users | get followers |
Each tool’s input schema in your MCP client describes parameters and optional fields.
Development
Prerequisites
- Node.js 18+
- Yarn (or npm)
Setup
git clone https://github.com/vaibhavpandeyvpz/pipedriver.git
cd pipedriver
yarn install
Create a .env or export your token:
export PIPEDRIVE_API_KEY=your_api_token
Commands
| Command | Description |
|---|---|
yarn dev |
Run server from source (no build) via tsx |
yarn build |
Compile TypeScript to dist/ |
yarn start |
Run compiled server (node dist/index.js) |
yarn test |
Run Jest tests (start server, connect, list tools) |
Project layout
src/– Server entry, MCP server setup, Pipedrive client wrappersrc/tools/– Tool handlers per domain (deals, persons, products, etc.)src/schemas/– Zod schemas for request bodiestests/– Integration tests (Jest)
Deployment
This server is not a long-lived HTTP service. It is started by your MCP client when needed and communicates over stdio.
- Production use = your MCP client (e.g. Cursor, Claude Desktop) runs
npx pipedriverorpipedriverwithPIPEDRIVE_API_KEYset. No extra deployment step. - Publishing: The package is published to npm via GitHub Actions on version tags (
v*). See .github/workflows/publish-npm.yml.
License
MIT © Vaibhav Pandey
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。