Todoist Meeting MCP
Connects Claude to Todoist for transforming meeting notes into actionable tasks with inferred due dates and priorities. It enables full task lifecycle management, including creating subtasks, listing projects, and completing tasks through natural language.
README
Todoist Meeting MCP
A production-ready Model Context Protocol (MCP) server that connects Claude with Todoist for intelligent meeting note processing. Parse meeting notes into actionable Todoist tasks with inferred due dates and priorities, create and complete tasks by name, and list projects and tasks—all through natural language in Claude.
What this does
This MCP server exposes Todoist as tools to Claude: you can paste meeting notes and have action items turned into tasks, create tasks with natural-language due dates (e.g. "by Friday"), add subtasks, list projects and tasks, update tasks, and complete them by name. Tasks are routed to projects/sections by keyword matching or an optional target project hint.
Prerequisites
- Node.js 18+
- Claude Desktop (or another MCP client)
- A Todoist account and API token
Quick setup
-
Get your Todoist API token
Todoist → Settings → Integrations → Developer → copy your API token. -
Configure Claude Desktop
Edit (or create) the MCP config file. If it doesn't exist, create theClaudefolder and the file:- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
Add (or merge in) the following, replacing
your_todoist_api_token_herewith your token:{ "mcpServers": { "todoist-meeting-mcp": { "command": "npx", "args": ["-y", "todoist-meeting-mcp"], "env": { "TODOIST_API_TOKEN": "your_todoist_api_token_here" } } } } - macOS:
-
Restart Claude Desktop
The server will start when Claude needs it. EnsureTODOIST_API_TOKENis set or the server will exit with a clear error.
Example prompts
-
Meeting notes → tasks:
"Here are my standup notes: [paste notes]. Create Todoist tasks." -
Single task:
"Create a task called 'Review PR #42' in my Engineering project due Friday." -
Subtasks:
"Add subtasks to my 'Q2 Planning' task: research competitors, draft slides, schedule review." -
List & complete:
"List my tasks due today." / "Complete the task 'Review PR #42'."
Tool reference
| Tool | Description |
|---|---|
parse_meeting_notes |
Parse raw meeting note text; extract action items, deadlines, priorities; create Todoist tasks (optional target project). |
create_task |
Create a single task with content, description, project/section by name, natural-language due date, priority, labels, optional parent (subtask). |
create_subtasks |
Add multiple subtasks to a parent task (parent found by name). |
list_projects |
List all projects with IDs, names, and section names (for routing context). |
list_tasks |
List tasks with optional filters: project_name, section_name, priority, due_today. |
update_task |
Update a task (found by name): content, description, due date, priority, project, section. |
complete_task |
Mark a task complete (found by name, fuzzy match). |
GitHub release / self-hosted usage
Releases are built and published on GitHub when you push a tag v*.*.* (e.g. v1.0.0). The workflow uploads todoist-meeting-mcp.js as a release asset.
-
Run with npx (no download): use the config block above; npx will run the published package.
-
Run from a downloaded asset:
- Download
todoist-meeting-mcp.jsfrom the Releases page. - Point Claude at it:
"mcpServers": { "todoist-meeting-mcp": { "command": "node", "args": ["/absolute/path/to/todoist-meeting-mcp.js"], "env": { "TODOIST_API_TOKEN": "your_todoist_api_token_here" } } } - Download
Build and test locally
From the project root:
# Install dependencies
npm ci
# Typecheck and lint (same as CI)
npm run typecheck
npm run lint
# Build (single file to dist/index.js)
npm run build
# Run the server (reads MCP over stdio; set token first)
export TODOIST_API_TOKEN=your_token_here
npm start
For development with auto-reload:
npm run dev
Code quality (CI)
On every push and pull request to main/master, the CI workflow runs:
- Typecheck —
tsc --noEmit - Lint — ESLint (TypeScript) on
src/ - Build —
npm run build
The badges at the top show the status of the CI and Release workflows once you set YOUR_ORG to your GitHub org or username.
License
MIT
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。