Taskflow MCP

Taskflow MCP

Automates Notion task workflows with AI-powered tag processing, enabling Claude to interrogate, expand, critique, rewrite, and implement tasks through intelligent tag-based commands.

Category
访问服务器

README

taskflow-mcp

AI-powered Notion task workflow automation for Claude Code

taskflow-mcp is a Model Context Protocol (MCP) server that automates Notion task workflows with AI-powered tag processing. Simply add MCP tags to your Notion tasks, and Claude Code will intelligently interrogate, expand, critique, rewrite, and even implement them.


🎯 Features

Full MCP Implementation

  • 4 Tools - Direct task processing, querying, and updates
  • 3 Resources - Quick access to Ready, In Progress, and all tagged tasks
  • 6 Prompts - AI-powered workflows (interrogate, expand, critique, user stories, rewrite, prepare for coding)
  • Sampling - LLM-powered analysis integrated via prompts

Intelligent Tag Processing

  • Priority-based - Tags processed in optimal order automatically
  • Persona support - "think like X" tags add expert context to all analysis
  • Automatic tag removal - Tags cleared after successful processing
  • Dry-run mode - Preview what would happen without making changes

AI-Powered Workflows (with Automatic Actions)

  • interrogate - Claude asks clarifying questions → Saves as comment
  • expand - Adds technical detail, APIs, edge cases → Saves as comment
  • rewrite - Rewrites task for clarity → Updates task description
  • estimate - Estimates effort and refactoring → Saves as comment
  • critique - Provides constructive feedback → Saves as comment
  • user stories - Generates user stories → Appends to description
  • to-do - Adds to todo list → No code written
  • code - Actually implements the feature (WRITES CODE)
  • confirm - Verifies implementation complete → Saves verification as comment

🚀 Quick Start

1. Install

git clone https://github.com/themightyboosh/taskflow-mcp.git
cd taskflow-mcp
pnpm install

2. Configure

Create .env file in your project directory:

NOTION_TOKEN=ntn_xxxxxxxxxxxxxxxxxxxxxxxxxxxxx
NOTION_DATABASE_ID=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Get your Notion token: https://www.notion.so/my-integrations

3. Setup Claude Code

Add to your .mcp.json (or Claude Code config):

{
  "mcpServers": {
    "taskflow-mcp": {
      "type": "stdio",
      "command": "node",
      "args": ["/absolute/path/to/taskflow-mcp/packages/mcp/index.js"]
    }
  }
}

4. Add MCP Tags in Notion

Add a multi-select property called "MCP" to your Notion database, then tag tasks with:

  • think like [persona] - e.g., "think like a security engineer"
  • interrogate - Claude asks clarifying questions
  • expand - Add technical detail
  • rewrite - Rewrite for clarity
  • critique - Get constructive feedback
  • user stories - Generate user stories
  • to-do - Add to todo list only
  • code - Implement the task automatically

5. Use in Claude Code

Via Resources:

Show me notion://tasks/ready

Via Tools:

Process all my Notion tasks

Via Prompts:

Interrogate task [task-id]

📖 Documentation


🏗️ Architecture

taskflow-mcp/
├── packages/mcp/
│   ├── index.js              # Main MCP server
│   ├── notion-client.js      # Notion API wrapper
│   ├── utils/
│   │   ├── config-loader.js  # Per-project .env loading
│   │   ├── tag-manager.js    # Tag priority & processing
│   │   └── image-downloader.js  # Image download for vision
├── .env.example
├── .mcp.json.example
└── README.md

🎨 Tag Processing Order

Tags are processed in this priority order:

  1. think like [X] - Sets persona context for all subsequent processing
  2. interrogate - Asks clarifying questions
  3. rewrite - Rewrites description for clarity
  4. estimate - Estimates level of effort and refactoring needs
  5. expand - Adds technical detail
  6. critique - Provides feedback (with persona if set)
  7. user stories - Generates user stories (appended to task description)
  8. to-do - Adds to todo list (no implementation)
  9. code - Triggers implementation (ONLY tag that writes code)
  10. confirm - Verifies implementation is complete

Each tag is removed after successful processing.


💡 Usage Examples

Example 1: Requirements Gathering

Notion Task: "Add user authentication" MCP Tags: interrogate, expand, user stories

What happens:

  1. interrogate → Claude asks questions about auth method, session management → Saves as comment
  2. expand → Adds technical detail (JWT vs session, password requirements) → Saves as comment
  3. user stories → Generates user stories for different auth scenarios → Appends to task description
  4. All three tags removed after processing
  5. All actions performed automatically - no manual intervention needed

Example 2: Expert Review

Notion Task: "Implement payment processing" MCP Tags: think like a security engineer, critique, expand

What happens:

  1. think like → Sets persona: "security engineer"
  2. critique → Critiques from security perspective (PCI compliance, data protection) → Saves as comment
  3. expand → Expands with security-focused technical details → Saves as comment
  4. All tags removed
  5. All comments automatically added to Notion task

Example 3: Full Implementation

Notion Task: "Add dark mode toggle" MCP Tags: interrogate, expand, code

What happens:

  1. interrogate → Claude asks about implementation approach, storage → Saves as comment
  2. expand → Adds technical detail (CSS variables, localStorage, toggle component) → Saves as comment
  3. code → Reviews requirements → WRITES THE ACTUAL CODE (creates files, edits components, etc.)
  4. All tags removed after processing
  5. Feature is fully implemented - not just analyzed

🔧 MCP Capabilities

Tools (4)

Tool Description
process_tasks Process all tasks with MCP tags
query_tasks Query tasks by status or MCP tags
add_comment Add comment to Notion task
update_task Update task properties

Resources (3)

Resource Description
notion://tasks/ready Ready tasks with MCP tags
notion://tasks/in-progress In Progress tasks with MCP tags
notion://tasks/with-mcp-tags All tasks with MCP tags

Prompts (6)

Prompt Description
interrogate_task Ask clarifying questions
expand_task Add technical detail
critique_task Provide constructive feedback
generate_user_stories Generate user stories
rewrite_task Rewrite for clarity
prepare_for_coding Analyze and implement

See MCP_CAPABILITIES.md for full details.


🛠️ Development

# Install dependencies
pnpm install

# Start MCP server (for testing)
pnpm start

# Development mode (auto-reload)
pnpm dev

🐛 Troubleshooting

"No .env file found"

  • taskflow-mcp requires per-project .env file
  • Run from directory containing .env with NOTION_TOKEN and NOTION_DATABASE_ID

"Property MCP does not exist"

  • Add a multi-select property called "MCP" to your Notion database
  • Configure your Notion integration to have access to the database

MCP server not connecting

  • Verify absolute path in .mcp.json is correct
  • Ensure Node.js 20+ is installed
  • Check Notion token is valid
  • Restart Claude Code after config changes

📝 Requirements

  • Node.js 20+
  • pnpm 10+ (or npm)
  • Notion integration with database access
  • Claude Code with MCP support

🤝 Contributing

Contributions welcome! Please open an issue or PR.

Development Setup

  1. Fork and clone
  2. pnpm install
  3. Create a branch: git checkout -b feat/my-feature
  4. Make changes and test
  5. Commit: git commit -m "feat: add my feature"
  6. Push: git push origin feat/my-feature
  7. Open a PR

📧 Support


📜 License

MIT © Monumental


🎯 Roadmap

  • [ ] Multiple database support
  • [ ] Custom tag definitions via config
  • [ ] Background polling mode (watch for new tags)
  • [ ] Webhook support for real-time processing
  • [ ] Analytics dashboard
  • [ ] Slack/Discord notifications
  • [ ] Task templates
  • [ ] Batch operations

Made with ❤️ by Monumental

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选