Basecamp MCP Server
Basecamp 3 project data and task management for MCP-compliant AI assistants.
Data last scanned today · Reviewed today
Overview
The Basecamp MCP Server provides an interface for interacting with Basecamp 3 accounts through the Model Context Protocol. Built with the FastMCP Python SDK, it exposes 79 distinct tools designed to manage project workflows, including todo lists, message boards, campfire chats, and document handling. Users can search across projects, create comments, manage card tables, and handle uploads or attachments directly through AI clients like Cursor or Claude Desktop. The server requires an OAuth application registered via 37signals Launchpad and includes scripts to simplify the configuration process for supported AI development environments.
Our verdict
This server carries a high risk level with a security score of 40/100. Because it is a community-maintained project rather than an official vendor release, its OAuth implementation and read-only access capabilities have not been independently reviewed. It is suitable for users comfortable managing their own Basecamp OAuth credentials and auditing the code, but should be avoided in environments requiring verified, enterprise-grade software.
- Supports 79 different tools for broad Basecamp feature coverage
- Zero direct dependencies reduces supply-chain attack surface
- Includes configuration scripts for popular AI IDEs
- Maintains an active repository with recent commits
- Security score of 40/100 indicates significant unreviewed risk
- Community-maintained with no official support from 37signals
- Authentication and read-only mode logic lack editorial review
Tools
| Tool | Description | Risk |
|---|---|---|
| get_projects | Browse Basecamp projects and project details. | low |
| get_project | Browse Basecamp projects and project details. | low |
| search_basecamp | Search across projects, todos, messages, campfire lines, comments, uploads, and schedules. | low |
| global_search | Search across projects, todos, messages, campfire lines, comments, uploads, and schedules. | low |
| get_todolists | Read and manage todolists, todos, todo groups, and completion state. | low |
| get_todolist | Read and manage todolists, todos, todo groups, and completion state. | low |
| create_todolist | Read and manage todolists, todos, todo groups, and completion state. | medium |
| update_todolist | Read and manage todolists, todos, todo groups, and completion state. | medium |
| trash_todolist | Read and manage todolists, todos, todo groups, and completion state. | high |
| get_todos | Read and manage todolists, todos, todo groups, and completion state. | low |
Compatibility
| Client | Local | Docker | Remote | Read-only |
|---|---|---|---|---|
| Claude Desktop | ||||
| Cursor | ||||
| VS Code | ||||
| Windsurf | ||||
| ChatGPT |
Frequently asked questions
›What are the core requirements to run this server?
You need Python 3.10 or newer, a Basecamp 3 account, and a Basecamp OAuth application configured via their integrations page.
›How is authentication handled?
The server uses OAuth; you must run the provided oauth_app.py script to authenticate and generate a token, which is then stored locally in oauth_tokens.json.
›Does the server provide tools for configuration?
Yes, it includes helper scripts like generate_codex_config.py, generate_cursor_config.py, and generate_claude_desktop_config.py to automatically generate the necessary MCP client configuration files.
›How do I create draft messages instead of publishing them immediately?
You can call the create_draft_message tool directly, or call the create_message tool with the publish parameter set to false.
›What should I do if my system's Python version is older than 3.10?
You should use uv to create a virtual environment, which allows you to run the project using a newer Python version than the one installed on your system.
Alternatives
Badge
Maintain this server? Add the live badge to your README.