Natural language database interactions powered by AI

Transform your questions into SQL, visualize data, and get instant insights — all through conversation.

Quick Start • Features • Documentation • Examples

</div>

Overview

myquery is an AI-powered CLI and web interface that lets you interact with databases using natural language. Ask questions, get answers, visualize data, and export results — without writing SQL.

Built on LangChain and OpenAI, myquery bridges the gap between business questions and database queries.

You: "Show me top 10 customers by revenue"
myquery: [Generates SQL, executes query, displays table, provides AI analysis]

Features

Core Capabilities

Natural Language to SQL — Ask questions in plain English (or Indonesian)
Interactive Chat — Conversational interface with context memory
Smart Visualizations — Auto-generates charts when you ask
AI Analysis — Automatic insights and summaries from query results
Data Export — Export to CSV, JSON, Excel with one command

Advanced Features

Multi-Database Support — PostgreSQL, MySQL, SQLite
Cross-DB Merge/Join — Combine data from multiple databases
Query Optimization — AI-powered performance suggestions
Schema Discovery — Automatic analysis and exploration
Web UI — Beautiful interface with real-time chat
MCP Server — REST API for external integrations

Quick Start

Installation

Option 1: Quick Install (Recommended)

Linux & macOS:

curl -sSL https://install.myquery.dev | bash

Windows (PowerShell):

# Download installer dari GitHub Releases
# https://github.com/zakirkun/myquery/releases

Option 2: Package Manager

PyPI:

pip install myquery

Homebrew (macOS):

brew tap myquery/tap
brew install myquery

Debian/Ubuntu:

# Download dari Releases
wget https://github.com/zakirkun/myquery/releases/latest/download/myquery-setup-linux.deb
sudo dpkg -i myquery-setup-linux.deb

Option 3: From Source

# Clone the repository
git clone https://github.com/zakirkun/myquery.git
cd myquery

# Install dependencies
pip install -r requirements.txt

# Install in development mode
pip install -e .

Untuk panduan instalasi lengkap, lihat INSTALL.md

Configuration

Create a .env file:

# Required
OPENAI_API_KEY=your_openai_api_key_here

# Optional: Auto-connect database
DB_TYPE=postgresql
DB_NAME=mydb
DB_USER=postgres
DB_PASSWORD=your_password
DB_HOST=localhost
DB_PORT=5432

First Query

# Start interactive chat (auto-connects using .env)
myquery chat start

# Ask a question
You: "Show me all tables"
You: "List top 10 products by sales"
You: "Show me a chart of revenue by month"

That's it! No manual connection, no SQL needed.

Usage

Interactive Chat

myquery chat start

Chat naturally with your database:

You: show me all customers
→ Displays formatted table + AI analysis

You: show me a bar chart of sales by region
→ Table + Analysis + Interactive chart (opens in browser)

You: export this to excel
→ Exports last result to Excel file

Single Queries

# Natural language
myquery query execute "Find all orders from last month"

# Raw SQL
myquery query sql "SELECT * FROM customers LIMIT 10"

# With debug mode
myquery query execute "Show revenue" --debug

Data Export

# Export query results
myquery export query "Show all customers" --format csv
myquery export query "Top products" --format excel --filename report
myquery export query "Revenue data" --format all  # CSV + JSON + Excel

Visualizations

# Auto-detected chart type
myquery visualize chart "Sales by month"

# Specific chart type
myquery visualize chart "Revenue by region" --type bar
myquery visualize chart "Category distribution" --type pie

Multi-Database Operations

# Add database connections
myquery multidb add prod --type postgresql --name proddb --user admin
myquery multidb add staging --type postgresql --name stagingdb --user admin

# Query all databases
myquery multidb query "SELECT COUNT(*) FROM users"

# Merge results (union)
myquery multidb query "SELECT * FROM products" --merge --merge-type union

# Join by key
myquery multidb query "SELECT id, revenue FROM sales" \
  --merge --merge-type join --merge-key id

Web Interface

# Start web server
myquery web start

# Access at http://localhost:8000
# Features: Interactive chat, schema explorer, visualizations

MCP Server

# Start MCP server for external integrations
myquery server start --port 7766

Examples

Business Analytics

# Revenue analysis
"What's our total revenue by month for this year?"
"Show top performing products"
"Which customers haven't ordered in 90 days?"

# Visualize trends
"Line chart of monthly sales trends"
"Pie chart showing revenue by category"

Data Exploration

# Schema discovery
"What tables do we have?"
"Show me the structure of the orders table"
"Find tables with foreign key relationships"

# Sample data
"Show me a sample of the users table"
"What are the distinct values in the status column?"

Cross-Database Comparison

# Compare environments
myquery multidb add prod --type postgresql --name prod_db
myquery multidb add staging --type postgresql --name staging_db

# Compare data
myquery multidb query "SELECT COUNT(*) FROM users"
myquery multidb compare

Export & Share

# Generate reports
myquery export query "Monthly sales summary" --format excel
myquery export query "Customer list" --format csv
myquery export query "Analytics data" --format all

Documentation

Command Reference

| Command | Description | |---------|-------------| | myquery chat start | Interactive chat session | | myquery query execute "<question>" | Execute natural language query | | myquery query sql "<sql>" | Execute raw SQL | | myquery export query "<question>" | Execute and export results | | myquery visualize chart "<question>" | Create visualization | | myquery multidb add <name> | Add database connection | | myquery multidb query "<sql>" | Query multiple databases | | myquery web start | Start web UI | | myquery server start | Start MCP server | | myquery connect db | Connect to database | | myquery --help | Show all commands |

Environment Variables

| Variable | Description | Default | |----------|-------------|---------| | OPENAI_API_KEY | OpenAI API key (required) | - | | OPENAI_MODEL | Model to use | gpt-4-turbo-preview | | DB_TYPE | Database type | - | | DB_HOST | Database host | localhost | | DB_PORT | Database port | 5432 | | DB_NAME | Database name | - | | DB_USER | Database user | - | | DB_PASSWORD | Database password | - | | MCP_PORT | MCP server port | 7766 | | LOG_LEVEL | Logging level | INFO |

Supported Databases

PostgreSQL — Full support with all features
MySQL — Full support with all features
SQLite — Full support, perfect for local development

Export Formats

CSV — Universal format, Excel/Google Sheets compatible
JSON — Structured data with metadata
Excel (.xlsx) — Formatted spreadsheet with auto-adjusted columns

Chart Types

Auto — Smart detection based on data
Bar — Comparisons and rankings
Line — Trends and time series
Pie — Distributions and proportions
Scatter — Correlations and relationships
Table — Interactive data table

Architecture

myquery/
├── cli/              # CLI interface (Typer)
│   ├── main.py       # Entry point
│   └── commands/     # Command modules
├── core/             # Core logic
│   ├── agent.py      # Main orchestrator
│   └── *.py          # Specialized modules
├── tools/            # LangChain tools
│   ├── connect_db_tool.py
│   ├── execute_query_tool.py
│   ├── visualize_data_tool.py
│   ├── export_data_tool.py
│   └── *.py
├── web/              # Web UI (FastAPI)
│   └── main.py
├── mcp/              # MCP Protocol
│   └── server.py
└── config/           # Configuration
    └── settings.py

Query Flow

User Input → Natural language question
Schema Analysis → Understand database structure
SQL Generation → LLM creates SQL using schema context
Validation → Safety checks (no destructive operations)
Execution → Run query on database
Formatting → Display results in rich tables
Analysis → AI-powered insights
Visualization (optional) → Generate charts
Export (optional) → Save to file

Safety Features

Read-Only by Default — Blocks DROP, DELETE, TRUNCATE, UPDATE, ALTER
Connection Validation — Tests connections before executing
Query Timeout — Configurable timeout for long queries
Password Security — Credentials never saved in session files
Error Handling — Graceful failures with helpful messages

Myquery

Install / Use

README