Noverload MCP Server

Connect your Noverload saved content to AI assistants like Claude, Cursor, and Windsurf using the Model Context Protocol (MCP).

🎯 Beta Release v0.6.0 - Enhanced stability, improved error handling, and better reliability based on user feedback.

🚀 Powered by Noverload API v2 for advanced search, content synthesis, and intelligent token management.

Quick Start

1. Get your token: Generate a Personal Access Token from your Noverload dashboard
2. Copy the config: Use the zero-install configuration below with your token
3. Add to your AI tool: Paste into Claude Desktop, Cursor, or Windsurf settings
4. Start using: Ask your AI about your saved content!

{
  "mcpServers": {
    "noverload": {
      "command": "npx",
      "args": ["-y", "noverload-mcp@latest"],
      "env": {
        "NOVERLOAD_CONFIG": "{\"accessToken\":\"your-token-here\",\"apiUrl\":\"https://www.noverload.com\",\"readOnly\":true}"
      }
    }
  }
}

Features

• 📚 Access all your saved content (YouTube, X posts, Reddit, articles, PDFs)
• 🔍 Advanced search with multiple modes (smart, semantic, fulltext)
• ⚠️ Token warnings for large content (prevents context overflow)
• ✅ View and complete action items
• 🎯 Goals tracking
• 🧠 Content synthesis and insights generation
• 🔒 Secure access with personal access tokens
• 📝 Read-only mode by default for safety

Installation

For Users

Recommended: Zero-Install with NPX

No installation needed! NPX automatically downloads and runs the latest version:

{
  "command": "npx",
  "args": ["-y", "noverload-mcp@latest"]
}

Alternative: Global Install

For faster startup (but requires manual updates):

npm install -g noverload-mcp

Then use:

{
  "command": "noverload-mcp",
  "args": []
}

For Development

git clone https://github.com/drewautomates/noverload-mcp.git
cd noverload-mcp
npm install
npm run build

Configuration

Step 1: Get Your Personal Access Token

1. Log in to Noverload
2. Go to Settings → API Access
3. Click "Generate Personal Access Token"
4. Copy the token (you won't be able to see it again)

Step 2: Configure Your AI Tool

Claude Desktop

Edit your Claude configuration file:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json
• Linux: ~/.config/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "noverload": {
      "command": "npx",
      "args": ["-y", "noverload-mcp@latest"],
      "env": {
        "NOVERLOAD_CONFIG": "{\"accessToken\":\"YOUR_ACCESS_TOKEN_HERE\",\"apiUrl\":\"https://www.noverload.com\",\"readOnly\":true}"
      }
    }
  }
}

Cursor

1. Open Cursor Settings
2. Navigate to Features → Model Context Protocol
3. Add configuration:

{
  "noverload": {
    "command": "noverload-mcp",
    "args": [],
    "env": {
      "NOVERLOAD_CONFIG": "{\"accessToken\":\"YOUR_ACCESS_TOKEN_HERE\",\"apiUrl\":\"https://www.noverload.com\",\"readOnly\":true}"
    }
  }
}

Windsurf

Add to your Windsurf MCP configuration:

{
  "mcpServers": {
    "noverload": {
      "command": "npx",
      "args": ["-y", "noverload-mcp@latest"],
      "env": {
        "NOVERLOAD_CONFIG": "{\"accessToken\":\"YOUR_ACCESS_TOKEN_HERE\",\"apiUrl\":\"https://www.noverload.com\",\"readOnly\":true}"
      }
    }
  }
}

Available Tools

Once configured, your AI assistant can:

Reading Content

• list_saved_content - Browse your saved content library
• get_content_details - Get full details including summaries and insights
• search_content - Search through your content by keywords
• list_actions - View action items extracted from content
• list_goals - See your Health, Wealth, and Relationships goals

Writing (when read-only is disabled)

• save_content - Save new URLs to your Noverload library
• complete_action - Mark action items as completed

Security Recommendations

1. Use Read-Only Mode: Keep readOnly: true in your configuration unless you specifically need write access
2. Protect Your Token: Never share your personal access token
3. Revoke When Needed: You can revoke tokens anytime from Noverload settings
4. Scope Appropriately: Consider creating separate tokens for different use cases

Self-Hosting

If you prefer to run your own instance:

Option 1: Local Development Server

git clone https://github.com/yourusername/noverload-mcp.git
cd noverload-mcp
npm install
npm run build

# Run directly
node packages/mcp-server/dist/index.js '{"accessToken":"YOUR_TOKEN","readOnly":true}'

Option 2: Deploy to Your Infrastructure

The MCP server can be deployed to any Node.js hosting platform:

1. Vercel/Netlify Functions: Deploy as a serverless function
2. Docker Container: Package and run anywhere
3. VPS: Run on your own server with PM2

Example Dockerfile:

FROM node:20-alpine
WORKDIR /app
COPY package*.json ./
RUN npm ci --production
COPY dist ./dist
EXPOSE 3000
CMD ["node", "dist/index.js"]

Option 3: Private NPM Registry

Host on your own NPM registry for team distribution:

# Build the package
npm run build

# Publish to your registry
npm publish --registry https://your-registry.com

API Endpoints Required

For self-hosting, your Noverload API needs these endpoints:

• GET /api/user - Validate access token
• GET /api/content - List saved content
• GET /api/content/:id - Get content details
• POST /api/content - Save new content
• GET /api/content/search - Search content
• GET /api/actions - List actions
• POST /api/actions/:id/complete - Complete action
• GET /api/goals - List goals

Development

Project Structure

noverload-mcp/
├── packages/
│   ├── mcp-server/       # Main MCP server implementation
│   │   ├── src/
│   │   │   ├── index.ts      # Entry point
│   │   │   ├── client.ts     # Noverload API client
│   │   │   ├── tools/        # MCP tools (actions)
│   │   │   └── resources/    # MCP resources
│   │   └── package.json
│   └── mcp-utils/         # Shared utilities
└── package.json           # Workspace root

Testing Locally

# Install dependencies
npm install

# Run in development mode
npm run dev

# Build for production
npm run build

# Type check
npm run typecheck

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

Troubleshooting

"Invalid access token"

• Ensure your token is correctly copied
• Check if the token has expired
• Verify you're using the correct API URL

"Client not initialized"

• Restart your AI assistant after configuration changes
• Check the configuration JSON syntax

Tools not appearing

• Ensure the MCP server is properly configured
• Check your AI assistant's MCP logs
• Try reinstalling the package

License

MIT

Support

• Issues: GitHub Issues
• Documentation: Noverload Docs
• Discord: Join our community