INFO

This guide is for those who have already installed and are using monacopilot with a version older than v1 and want to upgrade to monacopilot's stable new major v1 release. For newcomers, you can skip this.

Upgrading to Monacopilot v1.0.0

With the release of v1.0.0, Monacopilot now delivers faster, more accurate code completions powered by Mistral's Codestral model. This guide will help you quickly update your existing implementation.

Breaking Changes

• Provider Optimization: Previous general-purpose chat models have been replaced with Mistral's specialized codestral model, which is specifically designed for code completion. This model leverages Fill-in-the-Middle (FIM) technology to deliver significantly faster and super accurate completions.

• Enhanced Customization API: The interfaces for custom prompts and custom models have been completely redesigned to provide more flexibility and better performance. These changes enable more sophisticated integrations while maintaining a clean developer experience.

Update Steps

1. Update your dependencies

npm install monacopilot@1.0.0

2. Update your server-side code

Replace your existing CompletionCopilot configuration:

// Before
const copilot = new CompletionCopilot(process.env.OPENAI_API_KEY, {
    provider: 'openai',
    model: 'gpt-4o',
});

// After
const copilot = new CompletionCopilot(process.env.MISTRAL_API_KEY, {
    provider: 'mistral',
    model: 'codestral',
});

For more information about available providers and models, see the Copilot Options documentation.

3. Update your environment variables

Create or update your API key:

# Before
OPENAI_API_KEY=your_openai_api_key_here

# After
MISTRAL_API_KEY=your_mistral_api_key_here

You can obtain a Mistral API key from the Mistral AI Console.

4. Update Custom Prompt Implementation (If Used)

If you use the custom prompt feature, the interface has changed:

// Before
copilot.complete({
    options: {
        customPrompt: metadata => ({
            system: 'Your custom system prompt here',
            user: 'Your custom user prompt here',
        }),
    },
});

// After
copilot.complete({
    options: {
        customPrompt: metadata => ({
            context: 'Information about the codebase context',
            instruction: 'Instructions for code completion',
            fileContent: 'File content with cursor position',
        }),
    },
});

For detailed information about the new custom prompt format, see the Custom Prompt API documentation.

5. Update Custom Model Integration (If Used)

If you use a custom model configuration, update to the new interface:

// Before
const copilot = new CompletionCopilot(API_KEY, {
    model: {
        config: (apiKey, prompt) => ({
            endpoint: 'https://your-api-endpoint.com',
            body: {
                inputs: prompt.user,
                // other parameters
            },
            // headers
        }),
        transformResponse: response => ({text: response.generated_text}),
    },
});

// After
const copilot = new CompletionCopilot(API_KEY, {
    model: {
        config: (apiKey, prompt) => ({
            endpoint: 'https://your-api-endpoint.com',
            body: {
                // The prompt now has context, instruction, and fileContent
                inputs: `${prompt.context}\n\n${prompt.instruction}\n\n${prompt.fileContent}`,
                // other parameters
            },
            // headers
        }),
        transformResponse: response => ({text: response.generated_text}),
    },
});

For complete details on implementing custom models, see the Custom Model Integration documentation.