Working with AI

Stellify is designed for AI-assisted development. You can use AI to build new features, refactor existing code, or get help understanding your codebase.

MCP Integration

MCP (Model Context Protocol) lets AI assistants like Claude interact directly with your Stellify project.

What MCP Enables

With MCP connected, AI can:

• Read your project structure, files, and methods
• Create new files, routes, and components
• Modify existing code at the method level
• Run database migrations and seeders
• Build complete features from descriptions

Setting Up MCP

1. Install the Stellify MCP server:

npm install -g @stellify/mcp

2. Configure your AI client (Claude Desktop, Cursor, etc.) to use the MCP server

3. Authenticate with your Stellify API key

4. The AI now has access to your project

Example Prompts

Creating a feature:

"Create a blog posts feature with a Post model, PostController with CRUD methods, and API routes"

Adding to existing code:

"Add a search method to the UserController that filters by email"

Building UI:

"Create a dashboard page with a sidebar, header, and main content area using Tailwind"

Editor Prompt

The Stellify editor includes a built-in prompt interface for quick AI assistance.

Using the Prompt

1. Open a file or route
2. Click the AI button (or use the keyboard shortcut)
3. Describe what you want
4. Review and apply the suggestions

Good Prompts

Be specific about what you want:

• "Add validation for email and password fields"
• "Create a method that calculates the total price including tax"
• "Add error handling for when the user is not found"

Context Awareness

The editor prompt knows:

• Which file you're viewing
• What methods and variables exist
• Your project's models and routes

This context helps AI make relevant suggestions.

AI Workflows

Building New Features

1. Describe the feature at a high level
2. Let AI scaffold the structure (models, controllers, routes)
3. Review and adjust the generated code
4. Add business logic manually or with AI help
5. Test and iterate

Refactoring Existing Code

1. Select the code you want to change
2. Describe the refactoring goal
3. Review the proposed changes
4. Apply or modify as needed

Code Review

Ask AI to review your code:

• "Check this controller for security issues"
• "Suggest improvements for this method"
• "Is there a more efficient way to write this query?"

Tips for AI-Assisted Development

Start small: Build features incrementally rather than asking for everything at once.

Review everything: AI-generated code should be reviewed like any other code contribution.

Be specific: Vague prompts get vague results. Include details about parameters, return types, and expected behavior.

Iterate: First results are rarely perfect. Refine your prompts based on what you get back.

Use structured data: Stellify's structured code storage means AI can make precise, targeted changes rather than rewriting entire files.