Copilot Init

A VS Code extension that exposes GitHub Copilot chat capabilities as a local API server for programmatic use.

Features

• Exposes Copilot Chat models through a RESTful API
• Access to multiple AI models including GPT-4, Claude, and more
• Streaming and non-streaming response options
• Simple Python client for easy integration

Installation

1. Install the VS Code extension
2. Make sure GitHub Copilot Chat is installed and enabled in your VS Code

Usage

Starting the Server

1. Use the command palette (Ctrl+Shift+P / Cmd+Shift+P) and select "Start Copilot API Server"
2. The server will run on a local port (default is 3000, or next available port)
3. A status bar item will display the current port

Python Client

The easiest way to use the API is with the provided Python client:

pip install crimson-copilot-init-client

Example Usage

from crimson.copilot_init_client.client import CopilotClient

# Initialize the client (connects to the local API server)
client = CopilotClient()

# List available models
models = client.list_models()
for model in models:
    print(f"- {model['name']} ({model['vendor']}/{model['family']})")

# Get a completion response
client.chat("Write me a simple Python function to calculate factorial")

# Get a streaming response
for fragment in client.chat_stream("Write me a simple Python function to calculate factorial"):
    print(fragment, end='', flush=True)  # Print fragments as they arrive

Direct API Endpoints

If you prefer to use the API directly, the following endpoints are available:

GET /

Check if the API server is running.

GET /models

List all available Copilot models.

POST /chat

Get a complete (non-streaming) response.

Request body:

{
  "message": "Write a function to calculate factorial",
  "history": [
    {
      "content": "I need help with Python",
      "isUser": true
    },
    {
      "content": "I'd be happy to help with Python. What do you need?",
      "isUser": false
    }
  ],
  "vendor": "copilot",
  "family": "gpt-4"
}

POST /chat/stream

Get a streaming response.

Request body: Same as /chat endpoint.

Configuration

You can configure the extension to auto-start the API server when VS Code starts by adding the following to your settings.json:

{
  "copilotApi.autoStart": true
}

Requirements

• VS Code 1.85.0 or higher
• GitHub Copilot Chat extension installed and enabled

License

MIT