VibeDB

Overview  •  Roadmap  •  Releases  •  License

Desktop database management for developers who want speed, clarity, and production-aware safety.

Product Story

VibeDB is a desktop database manager built for real working sessions: browsing tables, shaping schemas, editing rows, running SQL, visualizing relationships, generating SQL from plain English, and moving carefully when a connection points at production.

It is designed to feel like a sharp devtool, not a bloated admin panel. Dense layouts, direct interactions, and visible safety rails are part of the product.

Core Values

• Move fast, stay sharp - Production access should stay usable, but never casual.
• Local-first trust - Connections, preferences, and credentials stay in native desktop storage flows instead of living in the browser.
• Dense developer workflow - Keyboard-first navigation, multi-tab work, and compact surfaces keep the tool focused on throughput.
• Safety as product - Guardrails are visible, deliberate, and part of everyday usage instead of being buried behind configuration.

Why Teams Reach For VibeDB

• Production-aware query policy - VibeDB sits between unrestricted SQL editors and fully read-only modes. Production-tagged connections remain usable, while destructive statements are blocked or explicitly confirmed.
• Transactional editing - Multi-row changes are staged and committed as a single atomic operation instead of becoming a fragile sequence of ad hoc updates.
• Visual schema workflows - Tables and views can be created through a polished builder with real-time SQL preview, while the schema canvas helps explore relationships quickly, adjust zoom, reset the view, and capture shareable screenshots.
• Table-aware AI SQL assistant - The side panel turns natural-language prompts into SQL using the currently selected table schema, then lets you copy the result or insert it into a query tab immediately.
• Fast table navigation - Large results, logs, and sidebar lists stay responsive through virtualization and compact UI patterns.
• Pin important tables - Right-click any table in the sidebar to pin it, keeping critical tables at the top of the list for faster access.
• Built-in query workflow - SQL editing, saved queries, execution results, and schema refresh behavior live in one place instead of being scattered across modal-heavy flows.
• Local secret handling - Sensitive credentials are stored with desktop-native protection primitives rather than plain browser storage.

Feature Highlights

Query With Confidence

• DROP and TRUNCATE are blocked in the query editor for production-tagged connections.
• ALTER, CREATE, DELETE, INSERT, REPLACE, and UPDATE require explicit confirmation before they run against production-tagged connections.
• Read queries remain fast and direct.
• Query policy is visible in the editor so developers know when the tool is treating a connection differently.
• Guided flows such as create table, create view, edit table, row delete, and table truncate remain available with purpose-built confirmations instead of being treated like ad hoc SQL.

Edit Data Like A Devtool

• Inline editing and row inspection are built for quick, repeated changes.
• Selected row deletes use confirmation guardrails for sensitive environments.
• Schema-changing flows refresh the surrounding context so the app stays aligned with the database.

See The Schema Clearly

• The schema visualization view renders tables and relationships on a dedicated canvas for fast exploration.
• Toolbar controls support zoom in, zoom out, and reset view so large schemas stay navigable.
• Capture creates a schema screenshot preview in memory, and the result can be copied directly to the clipboard without exporting a permanent file first.

Generate SQL From Plain English

• The AI Assistant opens alongside an active data tab and scopes generation to the currently selected table.
• Default AI model is GwenCoder3, free and unlimited for all users.
• The panel shows the current table name, column count, and primary-key hints before generation.
• Generated SQL can be copied or inserted into the current query tab, or into a new AI Query tab when no query tab is active.

Stay In Flow

• Multi-tab workflow for switching between queries, tables, and saved work.
• Sidebar table context menu supports Pin Table / Unpin Table.
• Pinned tables are shown at the top of the table list.
• Keyboard-first query execution and editing.
• Compact, dark-first UI tuned for long working sessions.

Work Across Engines

• SQLite
• Turso / libSQL
• PostgreSQL
• MySQL is planned next

Query Policy

Query policy is one of VibeDB's clearest product values.

Many tools push teams toward only two extremes: fully unrestricted SQL access or fully read-only access. VibeDB takes a more practical approach for day-to-day engineering work. A production-tagged connection stays usable, but the editor applies statement-aware guardrails where mistakes are most expensive.

Current behavior:

• DROP and TRUNCATE are blocked in the query editor on production-tagged connections.
• ALTER, CREATE, DELETE, INSERT, REPLACE, and UPDATE require confirmation there before execution.
• Query parsing supports multi-statement SQL and ignores semicolons inside strings and comments.
• Blocked editor statements are re-checked in Rust before execution.
• Backend enforcement distinguishes between query editor execution and guided app flows, so query-editor restrictions do not accidentally break structured schema tools.
• Guided destructive actions such as row deletion and table truncation keep their own production confirmation flows.

This is an application-level safety rail. It complements database roles and permissions; it does not replace them.

Trust Model

• Credentials and sensitive tokens are stored with tauri-plugin-stronghold.
• App state and preferences are stored with tauri-plugin-store.
• VibeDB is built to reduce accidental mistakes in live environments, especially during direct SQL work.

Designed For

• Solo developers working across local, staging, and production environments
• Small teams that want a fast desktop client with clear operational guardrails
• Developers who prefer direct database access without giving up visible safety cues

Project Links

• Roadmap
• Releases
• License