AVC MCP Server

MCP server for AVC video enhancement service.

What is this?

This is an MCP (Model Context Protocol) server that provides video enhancement capabilities to MCP clients like Claude Desktop.

• MCP Client: Software that calls MCP tools (like Claude Desktop)
• MCP Server: This project - provides video enhancement tools via MCP protocol
• AVC Service: The remote video enhancement service that this server calls

Claude Desktop (MCP Client) → AVC MCP Server → Remote AVC Service

Installation

# Using uv (recommended)
uv pip install -e "."

# Using pip
pip install -e "."

Usage

Stdio Mode (for Claude Desktop)

# Run with default settings
avc-mcp-server

# Or with custom AVC service URL
avc-mcp-server --avc-url http://192.168.0.7:8001 --avc-api-key your-key

Configure in Claude Desktop (claude_desktop_config.json):

{
  "mcpServers": {
    "avc-video-enhancement": {
      "command": "avc-mcp-server",
      "env": {
        "AVC_BASE_URL": "http://192.168.0.7:8001",
        "AVC_API_KEY": "your-api-key"
      }
    }
  }
}

SSE Mode (HTTP)

# Run SSE server on port 8000
avc-mcp-server --transport sse --port 8000

# With custom AVC service URL
avc-mcp-server --transport sse --port 8000 --avc-url http://192.168.0.7:8001

The SSE endpoint will be available at: http://localhost:8000/sse

Available Tools

enhance_video_sync

Enhance a video synchronously (waits until complete).

Input:

• video_url (string, required): Video URL or local file path
• resolution (string, optional): Target resolution - "480p", "540p", "720p", "1080p", "2k" (default: "720p")
• type (string, optional): Source type - "url" or "local" (default: "url")

Output:

{
  "success": true,
  "task_id": "abc123",
  "status": "completed",
  "video_url": "https://...",
  "message": "Enhancement completed"
}

enhance_video_async

Start video enhancement asynchronously (returns task ID immediately).

Input: Same as enhance_video_sync

Output:

{
  "success": true,
  "task_id": "abc123",
  "status": "wait",
  "message": "Task created"
}

get_task_status

Query the status of a video enhancement task.

Input:

• task_id (string, required): Task ID from enhance_video_async

Output:

{
  "success": true,
  "task_id": "abc123",
  "status": "completed",
  "video_url": "https://...",
  "message": "Enhancement completed"
}

Environment Variables

Development

# Install dev dependencies
uv pip install -e ".[dev]"

# Run tests
uv run pytest -v

# Code formatting
ruff format .
ruff check .

License

MIT