AI-powered GitHub Projects V2 and Issues management with intelligent automation.

🔗 App URL: https://github-projects.deco.page/

A comprehensive MCP (Model Context Protocol) server for managing GitHub Projects V2 and Issues through AI. Built for teams that want to automate their GitHub workflow with intelligent task management, issue tracking, and project organization.

1. Go to your Deco workspace at admin.decocms.com
2. Navigate to Apps in your project
3. Search for "GitHub Projects Manager" or install directly from:
   • App URL: https://github-projects.deco.page/
4. Configure during installation:
   • GitHub Token (required): Personal Access Token with repo, project, and read:org scopes
   • Default Organization (optional): Your main GitHub organization (e.g., deco-cx)

1. Go to GitHub Settings → Developer settings → Personal access tokens → Tokens (classic)
2. Click "Generate new token (classic)"
3. Select scopes:
   • ✅ repo - Full control of private repositories
   • ✅ project - Full control of projects
   • ✅ read:org - Read org and team membership
4. Generate token and copy it
5. Paste it during app installation in deco

This MCP server provides powerful GitHub management capabilities:

• 📊 Projects V2 Management: Full CRUD operations for GitHub Projects V2 using GraphQL API
• 🎯 Complete Issues Control: Create, update, close, label, assign, and comment on issues
• 📝 Database Tracking: Track specific repositories and projects for focused monitoring
• 🤖 AI-Native Tools: All operations exposed as MCP tools for AI agent automation
• 🔍 Tool Introspection: Built-in UI to explore all available tools and their schemas
• ⚡ Real-time Updates: Manage tracked repos and projects through intuitive web interface

With this MCP server, you can:

• Automate Task Creation: Let AI create and organize issues based on requirements
• Smart Project Management: AI can manage project boards and track progress
• Intelligent Labeling: Automatically categorize and label issues based on content
• Team Coordination: Assign issues and manage workflows programmatically
• Status Tracking: Monitor project health and issue states across repositories

After installing the app, you'll have access to these views in your deco workspace:

1. 📊 Projects Browser - Browse GitHub Projects V2 from any organization
2. 🐛 Issues Browser - Explore issues with full details, comments, and labels
3. 📌 Tracking Manager - Configure which repos and projects to monitor
4. 🔧 Tools Inspector - Explore all 25+ available MCP tools

This repository uses Specstory to track the history of prompts that were used to code this repo. You can inspect the complete development history in the .specstory/ folder.

• 🤖 MCP Server: Cloudflare Workers-based server with typed AI tools
• ⚛️ React Frontend: Modern React app with Vite, TanStack Router, and Tailwind CSS
• 🎨 UI Components: Pre-configured shadcn/ui components for rapid development
• 🔧 Type Safety: Full TypeScript support with auto-generated RPC client types
• 🚀 Hot Reload: Live development with automatic rebuilding for both frontend and backend
• ☁️ Ready to Deploy: One-command deployment to Cloudflare Workers
• 🔗 API Integrations: Connect to thousands of services through deco marketplace
• 🛡️ Built-in Auth: OAuth integration with popular providers (Google, GitHub, etc.)
• 🗄️ Integrated Database: SQLite database with 10GB storage included in every project

We're finishing our visual workflow editor where you'll be able to:

• 🎨 Drag & Drop: Compose AI tools visually with an intuitive interface
• 🤝 Real-time Collaboration: Work together on workflows with your team
• 🧠 AI-Generated Workflows: Let AI help you build complex automation flows
• ⚡ Live Preview: See your workflows in action as you build them

This will revolutionize how you create AI-powered automation and data processing pipelines.

# Create a new AI app from this template
npm create deco@latest my-ai-app

# Navigate to your project
cd my-ai-app

# Configure your app (sets name, scope, and workspace)
npm run configure

# Start development server
npm run dev

You can also create a new app by using this repository as a GitHub template:

1. Click "Use this template" on GitHub to create a new repository
2. Clone your new repository
3. Install dependencies and configure:

git clone <your-new-repo-url>
cd <your-new-repo>

# Install dependencies
npm install

# Configure your app (REQUIRED - sets name, scope, and workspace)
npm run configure

# Start development server
npm run dev

The server will start on http://localhost:8787 serving both your MCP endpoints and the React frontend.

• Node.js ≥22.0.0
• Deco CLI - Required for development and deployment:

  npm install -g deco-cli

• Deco Account - Sign up at admin.decocms.com

After creating your app (via npm create or GitHub template), you must:

1. Install Deco CLI globally:

   npm install -g deco-cli

2. Login to Deco:

   deco login

3. Configure your app (REQUIRED before first use):

   npm run configure

   This configures:

   • App name: Determines your deployment URL (<app-name>.deco.page)
   • Scope: Maps to your deco organization/project
   • Workspace: Maps to your deco organization/project

4. Deploy to enable OAuth authentication:

   npm run deploy

Note: The template comes with default values (name = "deco-create", scope = "kmute"). You must run npm run configure to set your own values before development or deployment.

├── server/           # MCP Server (Cloudflare Workers + Deco runtime)
│   ├── main.ts      # Server entry point with AI tools
│   ├── tools/       # Domain-organized AI tools
│   └── deco.gen.ts  # Auto-generated integration types
└── view/            # React Frontend (Vite + Tailwind CSS)
    ├── src/
    │   ├── lib/rpc.ts    # Typed RPC client for server communication
    │   ├── routes/       # TanStack Router routes
    │   └── components/   # UI components with Tailwind CSS
    └── package.json

• npm run dev - Start development with hot reload
• npm run gen - Generate types for external integrations
• npm run deploy - Deploy to production
• npm run configure - Configure your app and workspace

The template includes a fully-typed RPC client that connects your React frontend to your MCP server:

// Typed calls to your AI tools
const result = await client.MY_AI_TOOL({ input: "data" });
const analysis = await client.ANALYZE_DATA({ dataset: data });

Connect to thousands of services instantly through your deco project:

• 🤖 AI Services: OpenAI, Anthropic, Cohere, local models
• 💳 Payments: Stripe, PayPal, Square
• 📧 Communication: SendGrid, Twilio, Slack
• 🗄️ Databases: PostgreSQL, MongoDB, Redis, Supabase
• 📊 Analytics: Google Analytics, Mixpanel, Segment
• And thousands more...

All apps installed in your deco project can be used to create new tools or views:

1. Install apps in your deco project dashboard
2. Run integration: npm run configure or bun run configure
3. Auto-generated types: MCP types are downloaded to shared/deco.gen.ts
4. Start building: Use the typed clients in your tools immediately

All integrations come with auto-generated TypeScript types and zero-config setup.

Every deco project includes a SQLite database with 10GB storage:

• Zero setup: Database is ready to use out of the box
• Type-safe ORM: Built-in Drizzle ORM with TypeScript support
• Automatic migrations: Schema changes are applied automatically
• Production ready: Scales with your application needs

Perfect for storing user data, application state, and business logic without external database setup.

Build powerful AI applications like:

• 📊 Data Analysis Tools: Connect to databases and APIs to analyze business data with AI
• 🤖 Customer Support Bots: Integrate with CRM and support systems for intelligent assistance
• 📝 Content Generation: Combine AI with content management systems and social media APIs
• 💼 Business Automation: Create workflows that connect multiple services and automate tasks
• 🔍 Research Assistants: Build tools that gather and analyze information from multiple sources
• 📈 Analytics Dashboards: Create AI-powered insights from your business data

• 🚀 Get Started: Create your first AI app
• 📚 Documentation: docs.deco.page
• 💬 Community: Join our Discord
• 🐙 Source Code: deco-cx/deco

Consider using SpecStory in your AI editor to track your development history and prompts as you build your AI application. It's a great way to document your development process and share knowledge with your team.

Ready to build the next generation of AI applications?

🚀 Start Building Now • 📖 Read the Docs • 💬 Join Community