Music Media MCP Server

🎵 Music Media MCP Server

An MCP (Model Context Protocol) server that generates AI-powered music videos. Give it an image or video and it will analyze the visual content, compose a matching soundtrack using Google's Lyria 3 model, merge everything with FFmpeg, and return a playable video artifact.

Pipeline

Source Media (image/video URL)
  → Gemini Vision analyzes the visual content (if no prompt given)
  → Lyria 3 generates a 30-second AI music track
  → FFmpeg merges audio + media into a single .mp4
  → Uploads to Google Cloud Storage
  → Returns an HTML artifact with an inline video player

Features

• Auto music prompting — If no music description is provided, Gemini Vision analyzes the image/video and generates a fitting music prompt automatically
• Multiple media types — Supports images (.jpg, .png, .webp) and videos (.mp4, .mov)
• Smart video handling — Images loop for 30s, short videos loop to fill, long videos trim to 30s
• HTML artifact output — Returns a styled video player that MCP-compatible chatbots render inline
• Cloud Run ready — Deploys to Google Cloud Run with a single command

Prerequisites

• Python 3.10+
• FFmpeg installed and on PATH

  # macOS
  brew install ffmpeg
  # Ubuntu/Debian
  sudo apt install ffmpeg
  

• Google Cloud project with:
  • Vertex AI API enabled (Lyria lyria-002 + Gemini gemini-2.0-flash-001)
  • A GCS bucket for output storage (with public read access or signed URLs)
  • Application Default Credentials:

    gcloud auth application-default login
    

Setup

1. Clone and install:

   git clone https://github.com/joshndala/music-media-mcp.git
   cd music-media-mcp
   python -m venv .venv
   source .venv/bin/activate
   pip install -e .
   

2. Configure environment:

   cp .env.example .env
   # Edit .env with your GCP project ID and GCS bucket name
   

3. Set up GCS CORS (required for video playback in chatbot artifacts):

   # Create cors.json
   echo '[{"origin":["*"],"method":["GET"],"responseHeader":["Content-Type","Content-Length","Range"],"maxAgeSeconds":3600}]' > cors.json
   gsutil cors set cors.json gs://YOUR_BUCKET_NAME
   

Running Locally

# stdio transport (for Claude Desktop and other MCP desktop clients)
python server.py

# SSE transport (for web-based MCP clients)
python server.py --transport sse --port 8000

# Test with MCP Inspector
npx @modelcontextprotocol/inspector
# Then connect to http://localhost:8000/sse

Deploying to Cloud Run

# Build the container
gcloud builds submit \
  --tag us-central1-docker.pkg.dev/YOUR_PROJECT/YOUR_REPO/music-media-server \
  --project YOUR_PROJECT

# Deploy
gcloud run deploy music-media-server \
  --image us-central1-docker.pkg.dev/YOUR_PROJECT/YOUR_REPO/music-media-server \
  --region us-central1 \
  --platform managed \
  --allow-unauthenticated \
  --set-env-vars "GCP_PROJECT_ID=YOUR_PROJECT,GCS_BUCKET_NAME=YOUR_BUCKET,GCP_LOCATION=us-central1" \
  --memory 2Gi \
  --timeout 300 \
  --project YOUR_PROJECT

Your SSE endpoint will be at: https://YOUR_SERVICE_URL/sse

MCP Client Configuration

Claude Desktop (`claude_desktop_config.json`)

{
  "mcpServers": {
    "music-media": {
      "command": "/path/to/.venv/bin/python",
      "args": ["/path/to/server.py", "--transport", "stdio"],
      "env": {
        "GCP_PROJECT_ID": "your-project-id",
        "GCS_BUCKET_NAME": "your-bucket-name",
        "GCP_LOCATION": "us-central1"
      }
    }
  }
}

Web/Chatbot (SSE)

Point your MCP client to your deployed Cloud Run URL:

https://your-service-url.run.app/sse

Tool Reference

`generate_and_merge_media`

Returns: A complete HTML document with an inline video player.

Example prompts:

• "Upbeat electronic dance music with synth arpeggios"
• "Calm ambient piano piece evoking a misty morning"
• "Cinematic orchestral score with soaring strings"
• (omit for automatic AI analysis)

Environment Variables

License

MIT