opendbpylot (Node.js)

Node.js / TypeScript bindings for opendbpylot — turn a natural-language question into SQL, run it on your database, and get the results. Uses Retrieval-Augmented Generation (RAG) with a self-repairing SQL loop.

npm install opendbpylot

Prebuilt native addons ship for macOS (x64/arm64), Linux (x64/arm64, glibc), and Windows (x64) — no compiler needed.

This installs both a Node library and a fully-working dbpylot command — the whole CLI runs in-process from the native addon, so no Rust toolchain is needed:

npm install -g opendbpylot
dbpylot init        # setup wizard
dbpylot             # chat with your database
dbpylot serve       # web UI
dbpylot ask "how many orders per country?"
dbpylot mcp         # serve as an MCP server for agent hosts (OpenPylot, Claude Desktop, …)

dbpylot mcp exposes the engine over the Model Context Protocol (stdio). Secrets can be set non-interactively for scripted setups: printf '%s' "$KEY" | dbpylot config set-key openai and dbpylot config set-db sqlite /data/app.db.

Usage (library)

First configure an LLM provider + a database. The bindings share the same config as the dbpylot CLI (install it with cargo install opendbpylot), so run the wizard once:

dbpylot init

Then:

import { OpenDbPylot } from "opendbpylot";

const bot = new OpenDbPylot();
const result = bot.ask("how many orders per country?");

console.log(result.sql);        // the generated (and self-repaired) SQL
console.log(result.columns);    // column names
for (const row of result.rows)  // result rows
  console.log(row);

// Teach it about your schema / business rules:
bot.trainDocumentation("Revenue excludes cancelled and refunded orders.");
bot.trainQuestionSql("top products", "SELECT name FROM products ORDER BY price DESC LIMIT 10;");

Fully self-contained web UI

You don't even need the CLI — launch the embedded web app straight from Node and configure everything (LLM + database) in the browser Settings panel:

import { OpenDbPylot } from "opendbpylot";
OpenDbPylot.serve();   // opens http://127.0.0.1:8080, runs until interrupted

API

• new OpenDbPylot() — load the engine from your saved configuration.
• .ask(question) → AskResult — { sql, columns, rows, repairsUsed, answer? }.
• .trainDdl(ddl), .trainDocumentation(doc), .trainQuestionSql(q, sql).
• OpenDbPylot.serve() — run the embedded web UI in-process (no CLI binary needed).
• OpenDbPylot.init() / .doctor() — convenience shims that call the dbpylot CLI (cargo install opendbpylot); serve() + Settings does the same setup.

Licensed under Apache-2.0.