Kommo CRM MCP Server
Enables AI agents to manage Kommo CRM (formerly AmoCRM) entities including leads, contacts, companies, tasks, notes, pipelines, and products through natural language commands via the Model Context Protocol.
README
Kommo CRM MCP Server & SDK
Model Context Protocol (MCP) server and TypeScript/JavaScript SDK for Kommo CRM (formerly AmoCRM) API v4.
This package provides a dual-purpose solution:
- MCP Server: Connect Kommo CRM to AI agents like Claude Desktop, providing them with tools to manage Leads, Contacts, Companies, Tasks, and more.
- SDK: A type-safe, complete TypeScript SDK for interacting with the Kommo API v4 in your own applications.
🚀 Features
- Complete CRUD: Full management of Leads, Contacts, Companies, Deals, Tasks, and Notes.
- Advanced Features:
- Manage Pipelines and Statuses.
- Handle Custom Fields intelligently (search by name, metadata awareness).
- Catalogs & Products support (including linking products to leads).
- AI-Ready: Designed with semantic tools optimized for LLM usage.
- Type-Safe: Built with TypeScript and Zod for robust validation.
- Dual Build: Supports both ESM (
import) and CommonJS (require).
📦 Installation
npm install @syncraengine/kommo-mcp
🤖 Usage as MCP Server (Claude Desktop)
To use this with Claude Desktop or any MCP client, add the following configuration to your claude_desktop_config.json (Mac: ~/Library/Application Support/Claude/, Windows: %APPDATA%\Claude\).
Configuration
{
"mcpServers": {
"kommo-crm": {
"command": "npx",
"args": [
"-y",
"@syncraengine/kommo-mcp"
],
"env": {
"KOMMO_SUBDOMAIN": "your-subdomain",
"KOMMO_ACCESS_TOKEN": "your-long-lived-access-token"
}
}
}
}
Environment Variables
| Variable | Description | Required |
|---|---|---|
KOMMO_ACCESS_TOKEN |
OAuth 2.0 Access Token (Long-lived recommended for MCP) | ✅ Yes |
KOMMO_SUBDOMAIN |
Your Kommo subdomain (e.g., mycompany for mycompany.kommo.com) |
✅ Yes (or BASE_URL) |
KOMMO_BASE_URL |
Full URL (e.g., https://mycompany.kommo.com) |
Optional |
📚 Usage as Library (SDK)
You can use the exported functions directly in your TypeScript/Node.js projects.
import {
searchContactByPhone,
createLead,
getKommoConfig,
type KommoConfig
} from '@syncraengine/kommo-mcp';
// 1. Configure
const config: KommoConfig = {
baseUrl: 'https://your-subdomain.kommo.com',
accessToken: process.env.KOMMO_ACCESS_TOKEN!
};
// 2. Use functions
async function main() {
// Search for a contact
const searchResult = await searchContactByPhone({ phone: '+1234567890' }, config);
if (searchResult.success && searchResult.contacts.length > 0) {
const contact = searchResult.contacts[0];
console.log(`Found contact: ${contact.name}`);
// Create a lead for this contact
const leadResult = await createLead({
name: 'New Deal from Website',
price: 500,
_embedded: {
contacts: [{ id: contact.id }]
}
}, config);
console.log('Lead created:', leadResult);
}
}
main();
🛠️ Available Tools (MCP)
The server exposes ~40 tools to the AI, covering:
- Contacts:
kommo_search_contact_by_phone,kommo_create_contact,kommo_update_contact, etc. - Leads:
kommo_list_leads,kommo_create_lead,kommo_update_lead,kommo_update_lead_custom_fields, etc. - Companies:
kommo_search_company,kommo_create_company,kommo_update_company. - Tasks:
kommo_list_tasks,kommo_create_task,kommo_update_task. - Notes:
kommo_list_notes,kommo_create_note. - Pipelines:
kommo_list_pipelines,kommo_get_pipeline_statuses. - Products:
kommo_list_catalogs,kommo_link_product_to_lead.
👨💻 Development
If you want to contribute or modify the package:
-
Clone the repository:
git clone https://github.com/syncraengine/kommo-mcp.git cd kommo-mcp -
Install dependencies:
npm install -
Build:
npm run buildThis generates both ESM (
dist/esm) and CommonJS (dist/cjs) builds. -
Test locally (MCP): You can use the MCP Inspector to test the server.
npx @modelcontextprotocol/inspector npx tsx src/bin/mcp-server.ts
📄 License
This project is licensed under the MIT License - see the LICENSE file for details.
Built with ❤️ by Eduardo Mendez
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。