hubspot-mcp

hubspot-mcp

A read-only MCP server that exposes HubSpot CRM data (contacts, deals, companies, quotes) to AI agents, enabling natural language queries.

Category
访问服务器

README

hubspot-mcp

CI Python License: MIT

Let Claude and other AI agents query your HubSpot CRM in natural language.

hubspot-mcp is a Model Context Protocol server that exposes your HubSpot CRM — contacts, deals, companies, and quotes — to AI agents. Ask "What deals are in the proposal stage?" or "Find the contact for acme.com" and the agent answers from live HubSpot data.

Read-only and safe by default. The server only uses read scopes; it cannot modify your CRM.


Features

  • 🔎 Search contacts, deals, companies, and quotes
  • 📇 Fetch full details for any record by id (with associations)
  • 🧭 List deal pipelines and stages so the agent knows valid filters
  • 🔒 Read-only by design — no write scopes requested or used
  • 🐍 Pure Python, built on the official MCP SDK (FastMCP)

Tools

Tool What it does
search_contacts(query, limit) Search contacts by name/email/company
get_contact(contact_id) Full detail for one contact
search_deals(query, stage, limit) Search/filter deals by name and stage
get_deal(deal_id) Full deal detail + associated contacts/company
search_companies(query, limit) Search companies by name/domain
get_company(company_id) Full company detail
list_quotes(deal_id, limit) List quotes, optionally for one deal
get_quote(quote_id) Full quote detail + line items
get_pipeline_stages() List deal pipelines and their stages

Install

Install directly from GitHub:

pip install git+https://github.com/MBAD753/hubspot-mcp
# or, isolated:
pipx install git+https://github.com/MBAD753/hubspot-mcp

A PyPI release (pip install hubspot-mcp) is planned — see the roadmap below.

Setup

  1. In HubSpot, go to Settings → Integrations → Private Apps → Create a private app.
  2. Grant read-only scopes: crm.objects.contacts.read, crm.objects.deals.read, crm.objects.companies.read, crm.objects.quotes.read.
  3. Copy the access token and set it in your environment:
export HUBSPOT_ACCESS_TOKEN="pat-na1-xxxx..."

Tip: Use a free HubSpot developer/test account while evaluating — never point a tool at a production account until you trust it.

Use with Claude Code

Register the server (user scope makes it available in every project):

claude mcp add hubspot --scope user \
  --env HUBSPOT_ACCESS_TOKEN=pat-naX-xxxx... \
  -- hubspot-mcp

Or add it directly to your ~/.claude.json under mcpServers:

{
  "mcpServers": {
    "hubspot": {
      "type": "stdio",
      "command": "hubspot-mcp",
      "env": { "HUBSPOT_ACCESS_TOKEN": "pat-naX-xxxx..." }
    }
  }
}

Reload the window, then ask Claude: "Use my hubspot tools — what deals do I have and what stage is each in?"

Use with Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "hubspot": {
      "command": "hubspot-mcp",
      "env": {
        "HUBSPOT_ACCESS_TOKEN": "pat-naX-xxxx..."
      }
    }
  }
}

Restart Claude Desktop, and the HubSpot tools appear. Try: "Search my HubSpot for deals in the proposal stage and show their amounts."

Development

git clone https://github.com/MBAD753/hubspot-mcp
cd hubspot-mcp
python -m venv .venv
.venv\Scripts\Activate.ps1      # Windows (PowerShell)
# source .venv/bin/activate     # macOS / Linux
pip install -e ".[dev]"
pytest          # run tests (mocked HTTP — no real API calls)
ruff check .    # lint

Roadmap

  • v1 (now): read-only search + fetch across contacts, deals, companies, quotes.
  • v2: opt-in write operations (create contact, update deal stage, add note, create quote), gated behind an explicit environment flag and separate write scopes.

License

MIT © Marcus Bickel

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选