HubSpot MCP Server Integration

What does “HubSpot” MCP Server do?

The HubSpot MCP (Model Context Protocol) Server is designed to enable AI assistants to interact directly with HubSpot CRM data. By acting as a bridge between AI models and your HubSpot account, this server allows for seamless access to contacts, companies, and engagement metrics. It incorporates built-in vector storage (using FAISS) for semantic search and caching mechanisms that help overcome HubSpot API limitations, ensuring faster and more reliable responses. The focus is on high-value, frequently used CRM operations, with robust error handling and AI-friendly optimization. This makes complex multi-step CRM workflows more efficient and enhances the capabilities of AI development workflows by providing direct, context-rich access to business data.

List of Prompts

• (No explicit prompt templates found in the repository. The section “Example Prompts” in the README contains only user prompt examples, not reusable templates.)

List of Resources

(No explicit resources are described in the repository or documentation. No MCP resource primitives are listed.)

List of Tools

• hubspot_create_contact
  Create HubSpot contacts with duplicate prevention logic.

• hubspot_create_company
  Create HubSpot companies with duplicate prevention logic.

• hubspot_get_company_activity
  Retrieve activity for specific companies.

• hubspot_get_active_companies
  Retrieve most recently active companies.

• hubspot_get_active_contacts
  Retrieve most recently active contacts.

• hubspot_get_recent_conversations
  Retrieve recent conversation threads with messages.

• hubspot_search_data
  Semantic search across previously retrieved HubSpot data.

Use Cases of this MCP Server

• Automated Contact Creation
  Enable AI assistants to create new contacts in HubSpot directly from conversations, emails, or LinkedIn profile text, streamlining lead intake and reducing manual data entry.

• Company Data Management
  Simplify the process of creating and updating company profiles in HubSpot through AI-driven workflows, ensuring accurate and up-to-date CRM records.

• Activity and Engagement Monitoring
  Retrieve the latest activity on companies and contacts, allowing sales teams and AI agents to monitor engagement trends and follow up more effectively.

• Conversation Analysis
  Access and analyze recent conversation threads, enabling AI to summarize interactions or identify follow-up opportunities.

• Semantic CRM Search
  Use built-in vector storage to perform semantic searches across HubSpot data, making it easy for AI to find relevant information in past interactions and CRM records.

How to set it up

Windsurf

1. Ensure you have Docker installed.
2. Obtain your HubSpot access token with required scopes.
3. In the Windsurf configuration, locate the section for MCP servers.
4. Add the HubSpot MCP server using the following JSON snippet:

   {
     "mcpServers": {
       "hubspot": {
         "command": "docker",
         "args": [
           "run", "-i", "--rm",
           "-e", "HUBSPOT_ACCESS_TOKEN=your_token",
           "-v", "/path/to/storage:/storage",
           "buryhuang/mcp-hubspot:latest"
         ]
       }
     }
   }
   

5. Save your configuration and restart Windsurf.
6. Verify the setup by querying your HubSpot data via Windsurf.

Claude

1. Install Node.js if not present.
2. Obtain your HubSpot access token.
3. Edit the Claude configuration to include the MCP server.
4. Insert the HubSpot MCP server with:

   {
     "mcpServers": {
       "hubspot": {
         "command": "docker",
         "args": [
           "run", "-i", "--rm",
           "-e", "HUBSPOT_ACCESS_TOKEN=your_token",
           "-v", "/path/to/storage:/storage",
           "buryhuang/mcp-hubspot:latest"
         ]
       }
     }
   }
   

5. Save and restart Claude.
6. Confirm the MCP server is listed and responding.

Cursor

1. Install Docker and Node.js.
2. Obtain the HubSpot access token.
3. Open Cursor’s configuration file.
4. Add this under mcpServers:

   {
     "mcpServers": {
       "hubspot": {
         "command": "docker",
         "args": [
           "run", "-i", "--rm",
           "-e", "HUBSPOT_ACCESS_TOKEN=your_token",
           "-v", "/path/to/storage:/storage",
           "buryhuang/mcp-hubspot:latest"
         ]
       }
     }
   }
   

5. Save changes and restart Cursor.

Cline

1. Ensure Docker is installed.
2. Acquire the HubSpot access token.
3. In your Cline environment, open the relevant config file.
4. Add:

   {
     "mcpServers": {
       "hubspot": {
         "command": "docker",
         "args": [
           "run", "-i", "--rm",
           "-e", "HUBSPOT_ACCESS_TOKEN=your_token",
           "-v", "/path/to/storage:/storage",
           "buryhuang/mcp-hubspot:latest"
         ]
       }
     }
   }
   

5. Save and restart Cline.

Securing API Keys
It is best practice to secure your HubSpot access token using environment variables rather than hard-coding values. Example:

{
  "mcpServers": {
    "hubspot": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-e", "HUBSPOT_ACCESS_TOKEN=${HUBSPOT_ACCESS_TOKEN}",
        "-v", "/path/to/storage:/storage",
        "buryhuang/mcp-hubspot:latest"
      ],
      "env": {
        "HUBSPOT_ACCESS_TOKEN": "your_token"
      }
    }
  }
}

How to use this MCP inside flows

Using MCP in FlowHunt

To integrate MCP servers into your FlowHunt workflow, start by adding the MCP component to your flow and connecting it to your AI agent:

Click on the MCP component to open the configuration panel. In the system MCP configuration section, insert your MCP server details using this JSON format:

{
  "hubspot": {
    "transport": "streamable_http",
    "url": "https://yourmcpserver.example/pathtothemcp/url"
  }
}

Once configured, the AI agent is now able to use this MCP as a tool with access to all its functions and capabilities. Remember to change “hubspot” to whatever the actual name of your MCP server is (e.g., “github-mcp”, “weather-api”, etc.) and replace the URL with your own MCP server URL.

Overview

My overall rating for the HubSpot MCP server is:
While the server is robust in its tooling and documentation for setup, the lack of explicit prompt templates and MCP resource primitives limits its flexibility for advanced AI workflows. Sampling and roots support are not mentioned. It is fit for practical CRM use, but could benefit from fuller MCP feature coverage.

MCP Score

MCP Table Score: 7/10