GoNavi

GoNavi is an open‑source, cross‑platform database client that bundles a SQL editor, result viewer and an AI assistant in a single lightweight package. The installer is between 20 MB and 26 MB and runs on Windows, macOS and Linux without relying on Electron, which keeps the download size and runtime memory lower than many comparable tools.

The client supports more than 50 data sources, including MySQL, PostgreSQL, Oracle, Redis, Kafka, MongoDB, ClickHouse and a range of Chinese and open‑source databases. Drivers for the most common sources are built in; additional drivers can be installed on demand or added via a custom driver + DSN configuration. You create connections by selecting a category, searching by name, or pasting a URI, and the same configuration can be exported or imported as JSON for easy sharing between machines.

In the main workspace you write SQL in a Monaco‑based editor that offers autocomplete for schemas, tables and columns. After executing a query, results appear in a virtual‑scroll table that remains responsive even with large result sets. Cells can be edited directly, with changes batched into a transaction that you can commit or roll back. Export options include CSV, XLSX, JSON and Markdown. An execution history panel lets you filter past statements by status or time and re‑inject them into the editor with a single click. All executed statements are recorded in an audit log that is masked by default; you can configure retention periods and export the log if needed.

The built‑in AI assistant can read the current database schema and generate, explain or optimise SQL statements. It works with OpenAI, Gemini, Claude and any service that implements the OpenAI‑compatible API. The AI can insert the generated SQL into the editor for preview before execution, and it never receives the database password, which stays on the local machine. GoNavi also includes an MCP service that lets external coding agents query table structures and, when enabled, run SQL through the same security settings as the AI assistant. A lightweight Web Server version is available for Docker, Kubernetes or Helm deployments, and automatic update checks provide the latest x64 and ARM64 builds.

Key features

• Supports 50 + data sources across relational, NoSQL, time‑series, vector and message‑queue categories
• Native UI built with Go and system WebView, avoiding Electron bloat