sqlsaber 0.61.0

SQLsaber - Open-source agentic SQL assistant

SQLsaber

SQLsaber is an open-source agentic SQL assistant. Think Claude Code but for SQL.

Ask questions about your data in plain English. SQLsaber writes SQL queries, executes them, and explains the results.

SQLsaber reads your schema, writes SQL, executes it, and explains the results—all from a single natural language prompt.

$ saber "How many orders were placed last month?"

# SQLsaber will:
# 1. Discover relevant tables (orders, order_items, etc.)
# 2. Analyze their schema
# 3. Generate and run SQL
# 4. Return results with explanation

Why SQLsaber?

• No context switching — Stay in your terminal, ask questions, get answers
• Schema-aware — Automatically introspects your database; no manual setup
• Safe by default — Read-only queries; won't modify your data
• Works with your stack — PostgreSQL, MySQL, SQLite, DuckDB, and CSV files

Install

# Recommended
uv tool install sqlsaber

# Or with pipx
pipx install sqlsaber

Then run:

saber

On first launch, SQLsaber walks you through connecting a database and setting up authentication.

Quick Examples

# Interactive mode
saber

# Single query
saber "show me users who signed up this week"

# Pipe from stdin
echo "top 10 customers by revenue" | saber

# Use a specific database
saber -d mydb "count active subscriptions"

Features

Feature	Description
Schema introspection	Discovers tables, columns, and relationships automatically
Conversation memory	Add context like "dates are stored as Unix timestamps"
Structured knowledge base	Store searchable KPI definitions, SQL patterns, and domain notes per database
Thread history	Resume past conversations with saber threads resume
Extended thinking	Enable --thinking for complex analytical queries
Multiple AI providers	Anthropic, OpenAI, Google, Groq (Claude Opus 4.5 by default)

Configuration

Add a database

saber db add mydb
# Prompts for connection details

Or pass a connection string directly:

saber -d "postgresql://user:pass@localhost:5432/mydb" "count users"

Change AI model

saber models set

Add memory (context for your database)

saber memory add "customer_id in orders table references users.id"
saber memory add "all timestamps are UTC"

Add knowledge (searchable KPI/query references)

# Add a structured knowledge entry
saber knowledge add "Revenue KPI" "Recognized revenue from shipped orders only" --sql "SELECT SUM(amount) FROM orders WHERE status = 'shipped'" --source "finance-wiki"

# Long description/SQL from files
saber knowledge add "Revenue definition" "$(cat ./knowledge/revenue_definition.md)" --sql "$(cat ./sql/revenue.sql)"

# Search saved knowledge
saber knowledge search "revenue shipped orders"

How It Works

1. Discovery — Lists tables and identifies relevant ones based on your question
2. Schema analysis — Introspects only the tables needed
3. Query generation — Writes SQL tailored to your database dialect
4. Execution — Runs read-only queries with safety checks
5. Results — Formats output with explanations

Documentation

Full docs at sqlsaber.com:

• Installation
• Getting Started
• Database Setup
• Knowledge Base
• Command Reference

Contributing

Contributions welcome! Please open an issue first to discuss changes.

If you find SQLsaber useful, a ⭐ on GitHub helps others discover it.

License

Apache-2.0 — see LICENSE