Basecamp MCP Server
Enables AI assistants to interact with Basecamp projects through natural language commands. Supports managing projects, to-do lists, messages, and creating tasks with full content rendering capabilities.
README
Basecamp MCP Server (TypeScript)
A Model Context Protocol (MCP) server that exposes Basecamp tools to AI assistants (e.g., VS Code Chat in Agent mode).
Supports projects, to‑do sets/lists, messages (with rendered content), and creating to‑dos.
Features
-
Projects
list_projects— rich listing (status, timestamps, purpose, client flags, bookmarks, enabled dock tools)- Pagination via HTTP
Linkheader (exposesnextPage) - Optional raw dock objects
-
To‑do sets / lists
- Resolves a project’s to‑do set from the project dock
get_todoset— returns to‑do set id/url; optional listslist_todolists— lists to‑do lists for a project’s to‑do set
-
Messages
list_messages— resolves the project Message Board from dock, lists messages with pagination- Optional full content rendering (
markdown/html/text) get_message— fetch a single message (rendered or raw JSON)
-
To‑dos
create_todo— create a to‑do in a list (title, description, optionaldue_on)
Project layout
src/
├─ lib/
│ └─ basecamp.ts # Basecamp client (auth refresh, requests, pagination)
├─ tools/
│ ├─ projects.ts # list_projects
│ ├─ todosets.ts # get_todoset, list_todolists
│ ├─ messages.ts # list_messages, get_message
│ └─ todos.ts # create_todo
└─ basecamp-mcp.ts # entrypoint (register tools, start stdio server)
Quick start
# 1) Install dependencies
npm install
# 2) Build
npm run build
# 3) (Optional) sanity check
node build/basecamp-mcp.js
# (MCP servers sit and wait on stdio; no output is expected here)
Environment variables
Add your required values in a .env file at the project root.
This file is referenced by the MCP server configuration below.
Do not commit your
.envfile to version control.
Configure in VS Code (MCP)
Create .vscode/mcp.json in this project:
{
"servers": {
"basecamp": {
"type": "stdio",
"command": "node",
"args": ["${workspaceFolder}/build/basecamp-mcp.js"],
"envFile": "${workspaceFolder}/.env"
}
}
}
Now restart the server from VS Code:
- Command Palette → “MCP: List Servers” → basecamp → Start/Restart
- Open Chat, switch to Agent, enable the basecamp tools.
Example tool calls (Agent mode)
Projects (JSON with dock):
Run basecamp.list_projects with { "format": "json", "dock_detail": true, "limit": 5 }
or natural language
retrieve all project information
Resolve to‑do set and lists:
Run basecamp.get_todoset with { "project_id": 123456789 }
Run basecamp.list_todolists with { "project_id": 123456789 }
Natural language example
show todo sets and lists for project "{{prpjectName / Id}}"
Create a to‑do:
Run basecamp.create_todo with { "project_id": 123456789, "todolist_id": 222222222, "content": "Kickoff", "due_on": "2025-09-30" }
Natural language example
create a to-do in the "{to do list name}", titled "Update README.md"
Messages (rendered content):
Run basecamp.list_messages with { "project_id": 123456789, "page": 1, "format": "markdown", "include_body": true }
Run basecamp.get_message with { "project_id": 123456789, "message_id": 987654321, "render": "markdown" }
or
get all messages from TestMCP project
get content for this message "{messageId / title"
Pagination: responses include
nextPagefrom the RFC5988Linkheader. Call again with{ "page": <nextPage> }.
Troubleshooting (quick)
- ESM import errors: make sure your project is configured for ESM/NodeNext (standard TS/Node setup).
- 401/429 from API: verify tokens and reduce request volume; the client retries once on
429usingRetry-After. - No to‑do set / message board: that project may not have those tools enabled in its dock.
License
Acknowledgments
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。