Skip to content

VSCodeMCP.md

Latest commit

 

History

History
121 lines (97 loc) · 2.75 KB

VSCodeMCP.md

File metadata and controls

121 lines (97 loc) · 2.75 KB
Raw

VS Code MCP Integration

OARC-Crawlers provides a Model Context Protocol (MCP) server that integrates seamlessly with Visual Studio Code's Copilot Chat and other MCP-compatible tools.

Quick Start

# Install the package
pip install oarc-crawlers

# Install and register with VS Code
python -m oarc_crawlers.mcp_api install --name "OARC Crawlers"

Configuration

The MCP server uses WebSocket transport and is configured by default to:

  • Run on port 3000
  • Support streaming responses
  • Use VS Code-compatible error formatting
  • Provide tool documentation through introspection

VS Code Settings

To configure VS Code to use the OARC-Crawlers MCP server:

  1. Open VS Code Settings (Ctrl+,)
  2. Search for "MCP Server"
  3. Add a new server configuration: 
    {
    "mcp.servers": [
        {
            "name": "OARC Crawlers",
            "port": 3000,
            "transport": "ws"
        }
    ]
}

Available Commands

The server exposes the following tools to VS Code:

YouTube Operations

  • Download videos and playlists
  • Extract video captions
  • Search for content
# Example in Copilot Chat
Human: Download this video in HD: https://youtube.com/...
Assistant: I'll help you download that video in high resolution using the OARC Crawlers tools.

GitHub Operations

  • Clone and analyze repositories
  • Find similar code patterns
  • Generate repository summaries
# Example in Copilot Chat
Human: Show me the structure of tensorflow/tensorflow
Assistant: I'll analyze the TensorFlow repository structure for you using OARC Crawlers.

Web Crawling

  • DuckDuckGo search integration
  • Documentation extraction
  • ArXiv paper fetching
# Example in Copilot Chat
Human: Find recent papers about transformers
Assistant: I'll search ArXiv for recent transformer papers using OARC Crawlers.

Error Handling

The server provides VS Code-compatible error responses:

{
    "error": {
        "code": "TOOL_ERROR",
        "message": "Failed to download video",
        "details": {
            "url": "https://youtube.com/...",
            "reason": "Video unavailable"
        }
    }
}

Common error codes:

  • TRANSPORT_ERROR: Connection issues
  • CLIENT_ERROR: Invalid parameters
  • TOOL_ERROR: Tool execution failed
  • AUTH_ERROR: Authentication failed

Development

To run the server in development mode:

from oarc_crawlers import OARCCrawlersMCP

# Initialize with custom configuration
mcp = OARCCrawlersMCP(
    data_dir="./data",
    name="OARC Crawlers (Dev)",
    port=3001  # Use different port for development
)

# Configure for VS Code
mcp.mcp.configure_vscode(
    server_name=mcp.mcp.name,
    port=mcp.port,
    supports_streaming=True
)

# Run server
mcp.run()