SQL with AI — Database Explorer + AI Query Assistant

Explore your databases, understand schemas quickly, and turn natural language into runnable SQL — right inside VS Code.

SQL with AI is a lightweight SQL client for SQL Server, PostgreSQL, and MySQL with an AI assistant that uses the currently selected database schema (including PK/FK relationships) to generate accurate queries.

Why you’ll love it

• One-click SQL Explorer: Connections → databases → tables/views/procs/functions
• Schema-aware AI: ask questions like “number of orders by product last month” and get executable SQL
• Context follows you: selection in the explorer becomes the AI’s context automatically
• Fast results viewer: pretty grid, execution time, NULL highlighting

Screenshots (placeholders)

Replace these image paths with your actual screenshots.

SQL Explorer

Ask in plain English → Get SQL → Run

Query results

Key features

🌲 SQL Explorer

• Browse databases, tables, views, procedures, and functions
• Inspect table details: columns, indexes, and constraints
• PK/FK relationships are surfaced and used by the AI assistant to join tables correctly

🧠 AI SQL Assistant (Copilot-powered)

In Copilot Chat, talk to the participant @ai-sql:

• Ask for data (it generates SQL, opens a query editor, and runs it)
  • “Give me the number of appointments by siteslug”
  • “Bring product names by number of orders last month”
• Ask for understanding (it explains the database and report ideas)
  • “Explain this database to me. What kind of useful reports can I get?”

The assistant first classifies your intent (query vs metadata question) and responds accordingly.

📝 Query editor + execution

• Create a new SQL document from the explorer
• Run the selected text or the whole file
• Results show up in a dedicated panel

Quickstart

1. Open SQL Explorer from the Activity Bar
2. Click Add Connection
3. Connect to your database
4. Expand Databases, pick one, and start:
   • Right-click → New Query
   • or ask @ai-sql a question in Copilot Chat

What it’s great for

• Quickly onboarding into an unfamiliar database
• Building one-off reports and sanity checks
• Exploring PK/FK relationships without leaving the editor
• Turning business questions into SQL without memorizing table names

Requirements

• Visual Studio Code (recent versions recommended)
• Network access to your database
• For AI features: GitHub Copilot Chat enabled

Notes & limitations

• The results grid shows up to 1000 rows for performance.
• If your schema is very large, the AI assistant summarizes it (it’s optimized for accuracy, not dumping everything).

License

MIT