SQL AI — Natural Language to SQL, Inside VS Code

Write plain English. Get production-ready SQL. Without leaving your editor.

SQL AI is a VS Code extension that converts natural language prompts into SQL queries via a custom sidebar. Choose your dialect, describe what you need, and get back a query, an explanation, and performance hints — instantly.

Features

• Natural language → SQL — describe your query in plain English
• Multi-dialect support — PostgreSQL, MySQL, and SQLite
• Inline explanations — understand what the generated query does
• Performance hints — optional suggestions to optimize your query
• One-click copy — send the SQL straight to your clipboard
• Local history — revisit past queries without re-typing

How It Works

1. Click the SQL AI icon in the VS Code activity bar
2. Type a plain-English prompt — e.g. "Show the top 5 customers by revenue this month"
3. Select your SQL dialect
4. Hit Generate
5. Review the SQL, explanation, and any performance hints
6. Copy the result or pull it from History later

Demo

Prompt:   "Show all users who signed up in the last 30 days and haven't placed an order"

SQL:      SELECT u.*
          FROM users u
          LEFT JOIN orders o ON o.user_id = u.id
          WHERE u.created_at >= NOW() - INTERVAL '30 days'
            AND o.id IS NULL;

Hint:     Consider adding an index on users.created_at for large tables.

Requirements

• VS Code 1.85.0 or later
• A running SQL AI backend (default: http://localhost:3000)
• A valid JWT token stored in the webview's localStorage under sqlai_token

Getting Started (Local Dev)

1. Clone the repo and install dependencies

git clone https://github.com/aishwaryaKurakula/sql-ai.git
cd sql-ai
npm install

2. Start the backend

npm run dev:backend

3. Build the extension

npm --workspace=extension run build

4. Launch in VS Code

• Open the repo in VS Code
• Press F5 to launch an Extension Development Host
• Click the SQL AI icon in the activity bar

For watch mode during development:

npm run dev:extension

Available Command

Current Limitations

• No built-in login flow — requires a valid JWT token to already be present
• Backend URL is hardcoded; no settings UI yet to change it
• Query history is local to the webview, not synced from the backend
• Schema auto-detection is in the codebase but not yet wired into the sidebar

Roadmap

• [ ] Auth flow directly inside the extension
• [ ] Configurable backend URL via VS Code settings (sqlai.apiUrl)
• [ ] Auto-detect SQL dialect from workspace .env files
• [ ] Send schema context with prompts for more accurate generation
• [ ] Sync history with the backend

Extension Info

Contributing

Pull requests are welcome. For major changes, open an issue first.

License

MIT