HubSpot Extended MCP Server
A standalone MCP server that extends HubSpot functionality for post-call processing and pre-call preparation workflows.
README
HubSpot Extended MCP Server
A standalone MCP (Model Context Protocol) server that extends HubSpot functionality for post-call processing and pre-call preparation workflows.
Features
This server provides access to:
- Meeting Details: Retrieve complete meeting information including descriptions and properties
- Meeting Creation: Create new meetings with associations to deals and contacts
- Deal Meetings: Get all meetings associated with deals, with filtering and sorting
- Deal Notes: Get all notes associated with specific deals, sorted by timestamp
- Task Management: Create, retrieve, update, complete, and manage tasks with proper associations
- Task Lookup: Find tasks by deal or contact with fuzzy name matching support
- Overdue Tasks: Efficiently retrieve overdue tasks with owner filtering
- Meeting Search: Search meetings by keywords in descriptions
- Filtering & Sorting: Advanced filtering by outcome, exclude Calendly meetings, sort by date
Prerequisites
- Python 3.8 or higher
- A HubSpot account with access to create private apps
- Claude Desktop (or another MCP-compatible client)
Installation
1. Clone or Download the Repository
git clone <your-repo-url>
cd hubspotMcpExtended
Or download and extract the ZIP file to a location on your machine.
2. Install Python Dependencies
pip install -r requirements.txt
3. Set Up HubSpot Access Token
Create a HubSpot Private App:
- Go to your HubSpot account
- Navigate to Settings → Integrations → Private Apps
- Click "Create a private app"
- Give it a name (e.g., "MCP Server")
- In the "Scopes" tab, select the following scopes:
crm.objects.contacts.readcrm.objects.deals.readcrm.objects.meetings.readcrm.objects.meetings.writecrm.objects.notes.readcrm.objects.tasks.readcrm.objects.tasks.write
- Click "Create app" and copy the access token
Configure Environment:
Create a .env file in the project directory:
HUBSPOT_ACCESS_TOKEN=your_token_here
LOG_LEVEL=INFO
Replace your_token_here with the access token from your HubSpot private app.
4. Configure Claude Desktop
Add the server to your Claude Desktop configuration file:
On macOS:
open -a "TextEdit" ~/Library/Application\ Support/Claude/claude_desktop_config.json
On Windows:
%APPDATA%\Claude\claude_desktop_config.json
Add this configuration:
{
"mcpServers": {
"hubspot-extended": {
"command": "python",
"args": ["/FULL/PATH/TO/hubspotMcpExtended/main_fastmcp.py"],
"cwd": "/FULL/PATH/TO/hubspotMcpExtended"
}
}
}
Important: Replace /FULL/PATH/TO/hubspotMcpExtended with the actual full path to where you downloaded/cloned this project.
5. Restart Claude Desktop
- Quit Claude Desktop completely (Cmd+Q on macOS, or close all windows on Windows)
- Reopen Claude Desktop
- The HubSpot Extended server should now be available
Verify Installation
In Claude Desktop, try asking:
"Get the details for HubSpot meeting ID 12345"
If configured correctly, Claude will use the get_meeting_details tool from this server.
Available Tools
1. get_meeting_details
Retrieve complete meeting information.
Parameters:
meeting_id(required): HubSpot meeting IDproperties(optional): Array of specific properties to retrieve
2. create_meeting
Create a new meeting with associations to contacts and deals.
Parameters:
title(required): Meeting title/namestart_time(required): Meeting start time in ISO format (e.g., "2025-10-30T14:00:00Z")end_time(optional): Meeting end time in ISO formatdescription(optional): Meeting description/bodyowner_id(optional): HubSpot owner ID for the meeting creatoroutcome(optional): Meeting outcome (SCHEDULED, COMPLETED, RESCHEDULED, NO_SHOW, CANCELED)location(optional): Meeting locationcontact_ids(optional): List of contact IDs to associate with the meetingdeal_ids(optional): List of deal IDs to associate with the meetingmeeting_type(optional): Type of meeting based on account meeting typesinternal_notes(optional): Internal team notes about the meeting
Examples:
- Create simple meeting:
title="Client Call", start_time="2025-10-30T14:00:00Z" - Create with associations:
title="Sales Meeting", start_time="2025-10-30T14:00:00Z", deal_ids=["12345"], contact_ids=["67890"]
3. get_deal_notes
Get all notes associated with a specific deal.
Parameters:
deal_id(required): HubSpot deal IDlimit(optional): Number of notes to retrieve (default: 100)
4. create_task
Create a new task with associations to contacts and deals.
Parameters:
title(required): Task titleassigned_to_user_id(required): HubSpot user ID to assign todescription(optional): Task descriptiondue_date(optional): Due date in ISO formatpriority(optional): HIGH, MEDIUM, or LOWcontact_id(optional): Contact ID to associate withdeal_id(optional): Deal ID to associate withtask_type(optional): TODO, CALL, EMAIL, etc.
5. get_tasks
Retrieve tasks with optional filtering.
Parameters:
owner_id(optional): Filter by task ownercontact_id(optional): Filter by associated contactdeal_id(optional): Filter by associated dealstatus(optional): Filter by task statuslimit(optional): Number of tasks to retrieve (default: 100)
6. get_task_details
Get detailed information for a specific task.
Parameters:
task_id(required): HubSpot task IDproperties(optional): Array of specific properties to retrieve
7. complete_task
Mark a task as completed.
Parameters:
task_id(required): HubSpot task IDcompletion_notes(optional): Notes about the completionupdate_properties(optional): Additional properties to update
8. update_task
Update an existing task with new property values.
Parameters:
task_id(required): HubSpot task IDtitle(optional): New task titledescription(optional): New task description/notesstatus(optional): Task status (NOT_STARTED, IN_PROGRESS, COMPLETED, WAITING, DEFERRED)priority(optional): Task priority (HIGH, MEDIUM, LOW)assigned_to_user_id(optional): HubSpot user ID to reassign task todue_date(optional): New due date in ISO formattask_type(optional): Type of task (TODO, CALL, EMAIL, etc.)
Examples:
- Update title:
task_id="123", title="Follow up with client" - Change priority:
task_id="123", priority="HIGH", status="IN_PROGRESS" - Reassign task:
task_id="123", assigned_to_user_id="456"
9. get_deal_meetings
Retrieve all meetings associated with a deal, with filtering and sorting.
Parameters:
deal_id(required): HubSpot deal IDlimit(optional): Number of meetings to return (default: 100)outcome_filter(optional): Filter by outcome (e.g., "COMPLETED", "SCHEDULED")exclude_calendly(optional): Exclude automated Calendly meetingssort_direction(optional): "DESCENDING" (newest first, default) or "ASCENDING"
10. get_overdue_tasks
Efficiently retrieve overdue tasks.
Parameters:
owner_id(optional): Filter by task ownerlimit(optional): Number of tasks to return (default: 100)
11. search_meetings
Search meetings by keywords in descriptions.
Parameters:
search_term(required): Term to search forlimit(optional): Number of results (default: 10, max: 100)sort_direction(optional): "DESCENDING" (newest first, default) or "ASCENDING"
12. get_tasks_for_deal
Get all tasks associated with a specific deal using ID or fuzzy name matching.
Parameters:
deal_id(optional): HubSpot deal IDdeal_name(optional): Deal name for fuzzy search (e.g., "Delta Dental")include_completed(optional): Include completed tasks (default: False)limit(optional): Number of tasks to retrieve (default: 100)
Examples:
- Get pending tasks:
deal_name="Delta Dental" - Get all tasks by ID:
deal_id="38702133148", include_completed=True
13. get_tasks_for_contact
Get all tasks associated with a specific contact using ID, name, or email.
Parameters:
contact_id(optional): HubSpot contact IDcontact_name(optional): Contact name for fuzzy searchcontact_email(optional): Contact email for exact matchinclude_completed(optional): Include completed tasks (default: False)limit(optional): Number of tasks to retrieve (default: 100)
Examples:
- Get tasks by name:
contact_name="John Smith" - Get tasks by email:
contact_email="john@example.com" - Get all tasks by ID:
contact_id="122794298695", include_completed=True
Configuration
Environment variables (set in .env file):
HUBSPOT_ACCESS_TOKEN: Your HubSpot private app access token (required)LOG_LEVEL: Logging level (default: INFO, options: DEBUG, INFO, WARNING, ERROR)
Error Handling
The server handles common API errors:
- 401 Unauthorized: Invalid or expired token
- 403 Forbidden: Insufficient permissions
- 404 Not Found: Resource doesn't exist
- 429 Rate Limited: Automatic retry with exponential backoff
- 500+ Server Errors: Automatic retry with exponential backoff
Troubleshooting
Server Not Appearing in Claude Desktop
- Check the config file path is correct
- Verify Python path in config (try
which pythonin terminal) - Check that all dependencies are installed
- Look for errors in Claude Desktop logs (Help → View Logs)
API Errors
- 401 Unauthorized: Check your
.envfile has the correctHUBSPOT_ACCESS_TOKEN - 403 Forbidden: Verify your private app has the required scopes
- Rate Limit: The server automatically retries, but you may need to reduce request frequency
Meeting Sort Issues
If meetings appear in the wrong order, ensure you've restarted Claude Desktop after updating the code.
Sharing with Others
To share this MCP server:
- Package the project: ZIP the entire
hubspotMcpExtendedfolder - Share the ZIP with other users
- Provide instructions: Share this README with installation steps
- Note: Each user needs their own HubSpot private app token
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。