wannabespace/conar

Tamery.app – AI‑assisted SQL helper

What it is – Tamery is an open‑source desktop application (Electron + React) that lets you store database credentials securely in the cloud and then ask an LLM to write or optimise SQL for you. It currently supports PostgreSQL, MySQL, MSSQL and ClickHouse, with SQLite and MongoDB slated for later.

Key AI features

• Prompt‑driven SQL generation/optimisation via the Vercel AI SDK (OpenRouter backend).
• You can swap the underlying model, so you’re not locked to a single provider.

How it works

1. Add a connection (encrypted, password‑protected).
2. Open the Tamery UI, type a natural‑language request (e.g., “Give me the top 10 customers by revenue last month”).
3. The app sends the prompt to the chosen LLM, receives a SQL string, and can run it against the selected database.

Tech stack – React + TypeScript UI, Electron for a native desktop feel, Tailwind + shadcn/ui for styling, Vite for bundling, and a backend built with Bun, Hono, Drizzle ORM, and oRPC. Supporting services (Postgres, Redis) are run locally via Docker; deployment uses Railway and Stripe for licensing.

Getting started – Clone, run pnpm install, start Docker services (pnpm run docker:start), migrate the DB (pnpm run drizzle:migrate), then launch all parts with pnpm run dev. The CLI includes optional browser‑automation helpers (agent‑browser or terminal‑browser) for debugging UI‑driven agents.

Who it’s for – Developers, data analysts, or anyone who wants a quick, AI‑augmented way to craft SQL without memorising dialect nuances, while keeping credentials encrypted.

---

All details are taken directly from the repository’s README.