hubspot-mcp-server
Enables interaction with HubSpot CRM through MCP, providing tools to manage contacts, companies, deals, and search/associations via natural language.
README
hubspot-mcp-server
A Model Context Protocol (MCP) server for the HubSpot CRM API — works both locally via Claude Desktop (stdio) and remotely via claude.ai (HTTP + bearer token).
Tools
| Domain | Tools |
|---|---|
| Contacts | hubspot_get_contact, hubspot_list_contacts, hubspot_create_contact, hubspot_update_contact, hubspot_delete_contact |
| Companies | hubspot_get_company, hubspot_list_companies, hubspot_create_company, hubspot_update_company, hubspot_delete_company |
| Deals | hubspot_get_deal, hubspot_list_deals, hubspot_create_deal, hubspot_update_deal, hubspot_delete_deal |
| Search & Assoc. | hubspot_search, hubspot_associate_records, hubspot_list_associations |
Setup
1. Create a HubSpot Private App
- Go to HubSpot → Settings → Integrations → Private Apps
- Click Create a private app, then under Scopes add:
crm.objects.contacts.read+writecrm.objects.companies.read+writecrm.objects.deals.read+write
- Copy the generated Access Token
2. Install & Build
npm install
npm run build
Option A — Claude Desktop (local, stdio)
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"hubspot": {
"command": "node",
"args": ["/absolute/path/to/hubspot-mcp-server/dist/index.js"],
"env": {
"HUBSPOT_ACCESS_TOKEN": "pat-your-token-here"
}
}
}
}
Restart Claude Desktop — the tools will appear automatically.
Option B — claude.ai browser (remote, HTTP via Render)
Step 1 — Deploy to Render
-
Push this repo to GitHub
-
Go to render.com → New → Web Service
-
Connect your GitHub repo
-
Render detects
render.yamlautomatically. Set these Environment Variables in the dashboard:Key Value HUBSPOT_ACCESS_TOKENYour HubSpot private app token MCP_API_KEYA strong random secret you generate (e.g. openssl rand -hex 32)TRANSPORThttp -
Deploy. Your URL will be:
https://hubspot-mcp-server.onrender.com
Free tier note: Render's free tier spins down after 15 minutes of inactivity. The first request after spin-down takes ~30 seconds. Upgrade to Starter ($7/mo) to keep it always-on.
Step 2 — Connect in claude.ai
- Go to claude.ai → Settings → Integrations
- Click Add MCP Server
- Fill in:
- URL:
https://hubspot-mcp-server.onrender.com/mcp - Header name:
Authorization - Header value:
Bearer <your MCP_API_KEY>
- URL:
- Click Connect — the 15 HubSpot tools will appear in your next conversation
Environment Variables
| Variable | Required | Description |
|---|---|---|
HUBSPOT_ACCESS_TOKEN |
✅ | HubSpot private app token |
MCP_API_KEY |
✅ (HTTP mode) | Bearer token that Claude.ai sends to authenticate |
TRANSPORT |
No | stdio (default) or http |
PORT |
No | HTTP port, default 3000 (Render sets this automatically) |
Generate a secure MCP_API_KEY
openssl rand -hex 32
Copy the output into Render's environment variables and into the claude.ai header value field.
Local HTTP testing
HUBSPOT_ACCESS_TOKEN=pat-xxx MCP_API_KEY=mysecret TRANSPORT=http node dist/index.js
# Test auth
curl -X POST http://localhost:3000/mcp \
-H "Authorization: Bearer mysecret" \
-H "Content-Type: application/json" \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'
# Health check (no auth)
curl http://localhost:3000/health
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。