💻 VS Code / GitHub Copilot Installation

Prerequisites

• VS Code 1.85+ installed
• GitHub Copilot extension installed and active
• Python 3.10+ installed
• uv or pip package manager

1Install Code Scalpel

# Test that uvx works
uvx codescalpel --version

pip install codescalpel
codescalpel --version

2Configure VS Code

Option A: User Settings (All Projects)

1. Open VS Code
2. Press Ctrl+Shift+P (Windows/Linux) or Cmd+Shift+P (macOS)
3. Type "Preferences: Open User Settings (JSON)"
4. Add the MCP server configuration:

{
  "mcp.servers": {
    "code-scalpel": {
      "command": "uvx",
      "args": ["codescalpel", "mcp"]
    }
  }
}

Option B: Workspace Settings (Project-Specific)

For project-specific configuration, create or edit .vscode/settings.json:

{
  "mcp.servers": {
    "code-scalpel": {
      "command": "uvx",
      "args": ["codescalpel", "mcp"]
    }
  }
}

Option C: Using pip Installation

If you installed with pip, use the full Python path:

{
  "mcp.servers": {
    "code-scalpel": {
      "command": "/path/to/python",
      "args": ["-m", "code_scalpel"]
    }
  }
}

# macOS/Linux
which python

# Windows
where python

3Reload VS Code

After saving the configuration:

1. Press Ctrl+Shift+P / Cmd+Shift+P
2. Type "Developer: Reload Window"
3. Press Enter

4Verify Installation

Check MCP Server Status

1. Open the Output panel (Ctrl+Shift+U / Cmd+Shift+U)
2. Select "MCP" from the dropdown
3. Look for "code-scalpel" server status

Test with Copilot

Open Copilot Chat and try:

You should see Copilot invoke the analyze_code tool.

Advanced Configuration

With Environment Variables

{
  "mcp.servers": {
    "code-scalpel": {
      "command": "uvx",
      "args": ["codescalpel", "mcp"],
      "env": {
        "CODE_SCALPEL_LICENSE_PATH": "${workspaceFolder}/.code-scalpel/license.jwt"
      }
    }
  }
}

With Working Directory

{
  "mcp.servers": {
    "code-scalpel": {
      "command": "uvx",
      "args": ["codescalpel", "mcp"],
      "cwd": "${workspaceFolder}"
    }
  }
}

Multiple Workspaces

For multi-root workspaces, add to each workspace's settings:

{
  "folders": [
    { "path": "frontend" },
    { "path": "backend" }
  ],
  "settings": {
    "mcp.servers": {
      "code-scalpel": {
        "command": "uvx",
        "args": ["codescalpel", "mcp"]
      }
    }
  }
}

Using Code Scalpel in VS Code

In Copilot Chat

Ask Copilot to use specific Code Scalpel tools:

@workspace Use Code Scalpel's security_scan to check src/auth.py for vulnerabilities

@workspace Extract the UserService class using Code Scalpel

In Inline Chat

Select code and press Ctrl+I / Cmd+I:

Analyze this function with Code Scalpel and suggest improvements

Common Workflows

Troubleshooting

VS Code Variables

You can use these variables in your configuration:

Next Steps