matt-vst-lfr

mattdowney/matt-vst-lfr

3.2

If you are the rightful owner of matt-vst-lfr and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

This server implements the Model Context Protocol (MCP) to deliver Matt's voice, style, and tone preferences.

Matt's Voice, Style, and Tone MCP Server

HTTP server implementing the Model Context Protocol (MCP) to serve Matt's voice, style, and tone preferences.

Installation (Private Package)

Prerequisites

You need access to this private repository and GitHub authentication configured:

# Authenticate with GitHub (if not already done)
gh auth login

# Add GitHub Package Registry configuration
echo "@mattdowney:registry=https://npm.pkg.github.com" >> ~/.npmrc
echo "//npm.pkg.github.com/:_authToken=$(gh auth token)" >> ~/.npmrc

Global Installation (Recommended)

npm install -g @mattdowney/matt-vst-lfr-mcp

Use in Cursor IDE

Add to your ~/.cursor/mcp.json:

{
  "mcpServers": {
    "Voice and Tone": {
      "type": "stdio",
      "command": "matt-vst-lfr-mcp",
      "args": [],
      "env": {}
    }
  }
}

Setup on New Machines/Projects

  1. Install GitHub CLI and authenticate:

    gh auth login
    
  2. Configure npm for GitHub Package Registry:

    echo "@mattdowney:registry=https://npm.pkg.github.com" >> ~/.npmrc
    echo "//npm.pkg.github.com/:_authToken=$(gh auth token)" >> ~/.npmrc
    
  3. Install globally:

    npm install -g @mattdowney/matt-vst-lfr-mcp@latest
    
  4. Add to your project's mcp.json and restart Cursor

Local Development

git clone https://github.com/mattdowney/matt-vst-lfr.git
cd matt-vst-lfr
npm install
npm run mcp

HTTP server (optional) runs on http://localhost:3000

MCP Endpoints

Capabilities

  • GET /mcp/capabilities - Server capabilities

Prompts

  • GET /mcp/prompts/list - List available prompts
  • POST /mcp/prompts/get - Get specific prompt

Available prompts:

  • voice-style-tone - Complete voice profile
  • writing-guidelines - Core writing rules
  • banned-phrases - What to avoid

Tools

  • GET /mcp/tools/list - List available tools
  • POST /mcp/tools/call - Call specific tool

Available tools:

  • apply-voice-style - Apply Matt's voice guidelines to improve text
  • get-voice-guidelines - Get quick reference of voice guidelines

Resources

  • GET /mcp/resources/list - List available resources
  • POST /mcp/resources/read - Get specific resource

Available resources:

  • voice://matt/full-profile - Complete JSON profile
  • voice://matt/signature-moves - Signature writing moves
  • voice://matt/banned-elements - Banned phrases/patterns

Usage Example

# Get capabilities
curl http://localhost:3000/mcp/capabilities

# List prompts
curl http://localhost:3000/mcp/prompts/list

# Get voice profile prompt
curl -X POST http://localhost:3000/mcp/prompts/get \
  -H "Content-Type: application/json" \
  -d '{"name": "voice-style-tone"}'

Health Check

GET /health - Server health status

Deployment

Vercel

  1. Install Vercel CLI: npm i -g vercel
  2. Run: vercel
  3. Follow prompts
  4. Disable Protection: Go to your Vercel dashboard → Project Settings → Security → Disable "Vercel Authentication"
  5. Your MCP server will be available at: https://your-project.vercel.app

Note: By default, Vercel Pro accounts enable authentication protection. You must disable this in the dashboard for public API access.

GitHub Setup

# Create repository on GitHub, then:
git remote add origin https://github.com/yourusername/matt-vst-lfr.git
git push -u origin master