MyDBA - AI-Powered Database Assistant

MyDBA is an AI-powered VSCode extension that brings database management, monitoring, and optimization directly into your development environment. Built for developers and database administrators who want intelligent insights without leaving their IDE.

🚀 Features

Phase 1 MVP ✅ (Complete)

• Multi-Database Support: MySQL 8.0+, MariaDB 10.6+ (GA versions only)
• AI-Powered Query Analysis: Multi-provider support (VSCode LM, OpenAI, Anthropic, Ollama)
• Visual EXPLAIN Plans: Interactive tree diagrams with pain point highlighting
• Query Profiling: Performance Schema integration with waterfall charts
• Database Explorer: Tree view with databases, tables, indexes, and processes
• Enhanced Process List: Transaction detection with grouping by user/host/query
• Security-First Design: Credential isolation, production safeguards, query anonymization
• Documentation-Grounded AI: RAG system with MySQL/MariaDB docs to reduce hallucinations
• Editor Compatibility: Works in VSCode, Cursor, Windsurf, and VSCodium
• Comprehensive Testing: Integration tests with Docker, 70%+ code coverage

Metrics Dashboard

Real-time monitoring dashboard showing:

• Server information (version, uptime)
• Connections over time (current vs max)
• Queries per second with slow query detection
• Buffer pool hit rate
• Thread activity (running vs connected)

📋 Requirements

Editor Requirements

• VSCode: 1.85.0 or higher (also supports Cursor, Windsurf, VSCodium)
• Node.js: 18.x or higher (for development)

Database Requirements

Supported Versions

• MySQL: 8.0+ (LTS and Innovation releases)
• MariaDB: 10.6+, 10.11 LTS, 11.x+ (GA versions only)

Performance Schema (Required)

MyDBA requires Performance Schema to be enabled for monitoring features like:

• Query profiling and execution analysis
• Slow query detection
• Queries without indexes detection
• Transaction monitoring
• Process list with transaction details

Enable Performance Schema:

For MySQL 8.0+:

# In my.cnf or my.ini
[mysqld]
performance_schema = ON

For MariaDB 10.6+:

# In my.cnf or mariadb.conf.d/50-server.cnf
[mysqld]
performance_schema = ON
performance-schema-instrument = '%=ON'
performance-schema-consumer-events-statements-current = ON
performance-schema-consumer-events-statements-history = ON

Restart your database server after making configuration changes.

Verify Performance Schema is enabled:

SHOW VARIABLES LIKE 'performance_schema';

Should return: performance_schema | ON

User Permissions (Required)

The database user needs the following privileges for full functionality:

-- Process monitoring (required for process list)
GRANT PROCESS ON *.* TO 'mydba_user'@'%';

-- Database listing
GRANT SHOW DATABASES ON *.* TO 'mydba_user'@'%';

-- Metadata access (for schema information)
GRANT SELECT ON mysql.* TO 'mydba_user'@'%';

-- Performance Schema access (for monitoring and profiling)
-- UPDATE privilege is needed to configure instruments and consumers
GRANT SELECT, UPDATE ON performance_schema.* TO 'mydba_user'@'%';

-- Optional: Replication monitoring
GRANT REPLICATION CLIENT ON *.* TO 'mydba_user'@'%';

-- Database-specific access (adjust as needed)
GRANT SELECT, INSERT, UPDATE, DELETE ON your_database.* TO 'mydba_user'@'%';

-- Apply changes
FLUSH PRIVILEGES;

Note: MyDBA automatically configures Performance Schema instruments and consumers when you use profiling features. The UPDATE privilege on performance_schema.* is required for this auto-configuration.

Quick Links:

• 📖 Database Setup Guide - Detailed setup instructions
• ⚡ Quick Reference - Quick setup checklist and commands
• 🧪 Testing Guide - Docker setup for development

🛠️ Installation

From Marketplace (Coming Soon)

# Install from VSCode marketplace
code --install-extension mydba.mydba

From Source

# Clone repository
git clone https://github.com/your-org/mydba.git
cd mydba

# Install dependencies
npm install

# Compile TypeScript
npm run compile

# Package extension
npm run package

# Install locally
npm run install-extension

🚀 Quick Start

Prerequisites

Before connecting, ensure:

• ✅ Performance Schema is enabled in your database (see Database Setup Guide)
• ✅ User has required permissions (PROCESS, SELECT on performance_schema.*, etc.)
• ✅ Database version is supported (MySQL 8.0+ or MariaDB 10.6+)

Quick verification:

-- Check Performance Schema
SHOW VARIABLES LIKE 'performance_schema';  -- Must return 'ON'

-- Check your permissions
SHOW GRANTS FOR CURRENT_USER();

Getting Started

1. Connect to Database

   • Open Command Palette (Ctrl+Shift+P)
   • Run MyDBA: New Connection
   • Enter connection details (host, port, username, password)
   • Choose environment (dev/staging/prod)

2. Explore Database

   • View databases and tables in the MyDBA sidebar
   • Click to expand and explore schema
   • Right-click for context actions

3. View Metrics Dashboard

   • Click "Metrics Dashboard" in the sidebar
   • Monitor real-time database performance
   • Track connections, queries, buffer pool, and threads

4. Analyze Queries

   • Select SQL query in editor
   • Run MyDBA: Analyze Query or use @mydba /analyze
   • View AI-powered insights and recommendations

5. Visual EXPLAIN Plans

   • Run MyDBA: Explain Query
   • See interactive tree diagram
   • Identify performance bottlenecks
   • Apply one-click fixes

6. Profile Slow Queries

   • Run MyDBA: Profile Query
   • View execution stages with waterfall chart
   • Identify time-consuming operations

🔧 Configuration

AI Provider Setup

MyDBA supports multiple AI providers for maximum compatibility:

Provider	Editors	Cost	Privacy	Setup Required
GitHub Copilot (VSCode LM API)	VSCode only	Included with Copilot subscription	Data sent to GitHub	Automatic (no config)
OpenAI	All editors	Pay-per-use (~$0.0015/1K tokens)	Data sent to OpenAI	API key required
Anthropic Claude	All editors	Pay-per-use (~$0.003/1K tokens)	Data sent to Anthropic	API key required
Ollama	All editors	Free	100% local, no data sent	Local installation

Auto-Detection (Recommended)

Set mydba.ai.provider to "auto" (default) and MyDBA will automatically use the best available provider in this order:

1. VSCode Language Model (if available)
2. Configured API keys (OpenAI → Anthropic)
3. Ollama (if running locally)
4. Prompt to configure

VSCode with GitHub Copilot

No configuration needed! If you have GitHub Copilot, MyDBA will automatically use it.

OpenAI Setup

1. Get API key from platform.openai.com
2. Open Command Palette (Ctrl+Shift+P)
3. Run MyDBA: Configure AI Provider
4. Select "OpenAI" and enter your API key
5. Choose model (default: gpt-4o-mini)

Anthropic Claude Setup

1. Get API key from console.anthropic.com
2. Open Command Palette (Ctrl+Shift+P)
3. Run MyDBA: Configure AI Provider
4. Select "Anthropic" and enter your API key
5. Choose model (default: claude-3-5-sonnet)

Ollama Setup (Local & Private)

1. Install Ollama from ollama.ai
2. Pull a model: ollama pull llama3.1
3. MyDBA will auto-detect Ollama running on localhost:11434
4. Configure custom endpoint in settings if needed

Note for Cursor/Windsurf Users: VSCode Language Model API is not available in VSCode forks. Use OpenAI, Anthropic, or Ollama instead.

AI Settings

{
  "mydba.ai.enabled": true,
  "mydba.ai.provider": "auto",
  "mydba.ai.anonymizeQueries": true,
  "mydba.ai.includeSchemaContext": true,
  "mydba.ai.openaiModel": "gpt-4o-mini",
  "mydba.ai.anthropicModel": "claude-3-5-sonnet-20241022",
  "mydba.ai.ollamaEndpoint": "http://localhost:11434",
  "mydba.ai.ollamaModel": "llama3.1"
}

Security Settings

{
  "mydba.confirmDestructiveOperations": true,
  "mydba.warnMissingWhereClause": true,
  "mydba.safeMode": true,
  "mydba.preview.maxRows": 1000,
  "mydba.dml.maxAffectRows": 1000
}

Performance Settings

{
  "mydba.refreshInterval": 5000,
  "mydba.slowQueryThreshold": 1000,
  "mydba.queryTimeout": 30000,
  "mydba.maxConnections": 10
}

💬 VSCode Chat Integration

Use @mydba in VSCode Chat for natural language database assistance:

@mydba /analyze SELECT * FROM users WHERE email = 'test@example.com'
@mydba /explain SELECT * FROM orders WHERE created_at > '2024-01-01'
@mydba /profile SELECT COUNT(*) FROM large_table
@mydba how do I optimize this slow query?
@mydba what indexes should I add to the users table?

🔒 Security & Privacy

Data Privacy

• Local Processing: Schema metadata and query results stay local
• AI Anonymization: Queries templated (<table:name>, <col:name>, ?) before sending to AI
• Credential Security: Stored in OS keychain (Keychain/Credential Manager)
• No Data Collection: Telemetry disabled by default

Production Safeguards

• Destructive Operations: Confirmation required for DROP/TRUNCATE/DELETE/UPDATE
• Missing WHERE Warnings: Alerts for UPDATE/DELETE without WHERE clause
• Row Limits: Previews capped at 1,000 rows, DML operations blocked if > 1,000 rows
• Environment Awareness: Stricter rules for production connections

🏗️ Architecture

Core Components

• Service Container: Dependency injection for testability
• Database Adapters: Pluggable architecture for multi-database support
• AI Service Coordinator: VSCode LM API integration with RAG
• Webview Providers: Interactive EXPLAIN and profiling viewers
• Security Layer: Credential management and safety validators

Technology Stack

• Core: TypeScript, Node.js, VSCode Extension API
• Database: mysql2 (MySQL/MariaDB), pg (PostgreSQL), ioredis (Redis)
• AI: VSCode Language Model API, VSCode Chat API
• UI: Webview UI Toolkit, D3.js, Plotly.js, Chart.js
• Security: VSCode SecretStorage, SSL/TLS, SQL injection prevention

🧪 Testing

# Run unit tests
npm test

# Run integration tests (requires Docker)
npm run test:integration

# Run linting
npm run lint

# Run all tests
npm run pretest

Test Environment

# Start test databases
docker-compose -f docker-compose.test.yml up -d

# Run tests
npm run test:integration

# Cleanup
docker-compose -f docker-compose.test.yml down

📊 Performance

Targets

• Extension activation: < 500ms
• Connection test: < 2s
• Tree view refresh: < 200ms
• EXPLAIN visualization: < 300ms
• AI analysis: < 3s

Resource Limits

• Memory (idle): 50MB
• Memory (active): 200MB
• CPU (idle): 0%
• CPU (active): 5%
• Concurrent connections: 10

🤝 Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines.

Development Setup

# Clone and install
git clone https://github.com/your-org/mydba.git
cd mydba
npm install

# Start development
npm run watch

# In another terminal, open VSCode
code .
# Press F5 to launch Extension Development Host

Code Style

• TypeScript with strict mode
• ESLint configuration
• Prettier formatting
• Comprehensive tests (unit + integration)

📄 License

Licensed under the Apache License, Version 2.0. See LICENSE for details.

⚠️ Disclaimer

IMPORTANT: This software is provided "as is" without warranty of any kind. Users are responsible for:

• Testing in non-production environments first
• Understanding the impact of suggested optimizations
• Backing up data before applying changes
• Monitoring database performance after changes

See SECURITY.md for security policies and supported versions.

📞 Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Documentation:
  • Database Setup Guide
  • Quick Reference
  • Testing Guide

🙏 Acknowledgments

• Inspired by vscode-kafka-client
• Built with VSCode Extension API
• AI powered by VSCode Language Model API

---

Made with ❤️ for the database community