Brinn MCP Server

A Model Context Protocol (MCP) server that connects your AI assistants (Claude, Cursor, Windsurf, etc.) with your Brinn workspace. This integration allows AI agents to read, search, and manage your research notes, memos, and goals directly.

✨ Features

• 📝 Memo Management: Create, read, search, and update research notes.
• 📉 Daily Summaries: Get AI-generated summaries of your recent memos.
• 🏷️ Tags & Goals: Organize search and track your research progress.
• 🛠️ Developer Friendly: Built with TypeScript and the Model Context Protocol SDK.

Quick Setup

Connect Brinn to your favorite AI assistant in seconds using the built-in setup CLI:

npx brinn-mcp setup <provider> <BRINN_API_KEY>

Supported Providers:

• claude - Claude Desktop App
• cursor - Cursor Code Editor
• windsurf - Windsurf Editor
• gemini - Google Gemini / AI Studio
• roo-cline - VS Code Roo-Cline extension
• antigravity - Antigravity AI Assistant
• mcp-cli - Official MCP CLI for debugging

🔐 OAuth & Official AI Integration

If you are using the official Claude.ai or Cursor integration, the following redirect URIs must be allowlisted in your Brinn server settings (or configured via npm run mcp-setup on the server):

• http://localhost:6274/oauth/callback
• http://localhost:6274/oauth/callback/debug
• https://claude.ai/api/mcp/auth_callback
• https://claude.com/api/mcp/auth_callback

[!IMPORTANT] Your Brinn API Key must start with mclk_. You can find this in your Brinn settings.

🌥️ Cloudflare Workers Deployment

You can deploy Brinn MCP as a serverless Cloudflare Worker to provide a persistent, multi-tenant SSE (Server-Sent Events) endpoint for your AI assistants.

Quick Deploy:

1. Clone the repository and install dependencies:

   git clone https://github.com/naumanch969/memolink-mcp.git
   cd brinn-mcp
   npm install

2. Build and deploy to your Cloudflare account:

   npm run deploy

Setup on Cloudflare:

Once deployed, configure the environment variable:

• BRINN_API_URL: https://your-brinn-api.com/api (Optional).

Connecting to SSE:

The worker provides an MCP-compliant SSE endpoint at: https://brinn-mcp.<your-subdomain>.workers.dev/mcp/sse

Authentication: The SSE endpoint requires your Brinn API Key. You can provide it in two ways:

1. Authorization Header: Bearer mclk_SECRET
2. Query Parameter: ?apiKey=mclk_SECRET

Example for claude_desktop_config.json (SSE Mode):

{
  "mcpServers": {
    "brinn-cloud": {
      "url": "https://brinn-mcp.your-subdomain.workers.dev/mcp/sse?apiKey=mclk_your_secret_key"
    }
  }
}

🛠️ Available Tools

The following tools are exposed to your AI assistant:

| Tool | Description | Safety Hints | |------|-------------|--------------| | create_entry | Captures a new journal entry, research note, or personal memory. | destructive: false | | list_entries | Searches, filters, and retrieves past journal entries and memories. | readOnly: true | | read_entry | Fetches the full, detailed content of a specific entry. | readOnly: true | | update_entry | Refines, corrects, or updates an existing journal entry. | destructive: false | | delete_entry | Permanently deletes a specific journal entry or memory. | destructive: true | | list_collections | Retrieves a list of the user's organizational collections. | readOnly: true | | create_collection | Creates a new organizational category. | destructive: false | | read_collection | Gets details and metadata for a specific collection. | readOnly: true | | update_collection | Renames or modifies collection metadata. | destructive: false | | delete_collection | Permanently removes a collection category. | destructive: true | | get_user_tags | Retrieves all existing tags and their usage frequency. | readOnly: true |

Manual Configuration

If you prefer to configure your environment manually, ensure the following environment variables are set:

• BRINN_API_KEY: Your private Brinn API Key.
• BRINN_API_URL: (Optional) Defaults to https://api.brinn.app/api.

Example for claude_desktop_config.json:

{
  "mcpServers": {
    "brinn": {
      "command": "npx",
      "args": ["-y", "brinn-mcp"],
      "env": {
        "BRINN_API_KEY": "mclk_your_secret_key",
        "BRINN_API_URL": "https://api.brinn.app/api"
      }
    }
  }
}

🚀 Usage Examples

1. Research Journaling

User: "Create a new entry about my findings on the impact of GPT-5 on the developer market from my research today. Tag it as 'research' and 'ai-trends'." AI Goal: Use create_entry with content and tags.

2. Context Retrieval

User: "What was my mood last time I talked about the project launch in my entries?" AI Goal: Use list_entries with search terms, then read the related mood fields.

3. Organization

User: "Create a new collection called 'Project Brinn' to organize my research." AI Goal: Use create_collection with name: "Project Brinn".

🔒 Privacy & Security

Brinn MCP acts as a bridge between your AI assistant and your private Brinn workspace.

• Data Ownership: Your notes and data remain exclusively on Brinn servers.
• Encryption: All communication is performed over HTTPS.
• Privacy Policy: View our full privacy policy at https://mcp.brinn.app/privacy.
• Firewall Configuration: If you use a firewall (e.g., AWS Security Groups, ufw), allowlist Anthropic's outbound IP range 160.79.104.0/21 to enable Claude to connect to your MCP server.

🆘 Support & Community

• GitHub Issues: Report bugs or request features on our Issue Tracker.
• Email Support: Contact us at [email protected].
• Website: https://brinn.app.

🛠️ Development

Build

npm install
npm run build

Run Dev Server

npm run dev

Built by Opstin Technologies Team