SQL Editor

The SQL Editor is the heart of SoftDB. Built on Monaco Editor — the same engine that powers VS Code — it gives you a full-featured coding environment tuned specifically for database work.

Writing Queries

Type SQL directly into the editor. The editor supports standard SQL syntax with full highlighting for keywords, strings, numbers, comments, and operators.

Multiple statements are supported. Separate them with semicolons and SoftDB will execute them in sequence:

CREATE TABLE users (id SERIAL PRIMARY KEY, name TEXT);
INSERT INTO users (name) VALUES ('Alice'), ('Bob');
SELECT * FROM users;

Selection execution lets you run just part of your query. Highlight any text before pressing the run shortcut and only the selected portion executes. This is handy when you have a long file with multiple queries and want to test one at a time.

Autocomplete

The editor provides schema-aware completions as you type. No configuration needed — it reads your live schema automatically.

What gets completed:

• Tables — all tables in the current connection appear as suggestions
• Views — listed alongside tables with a distinct icon
• Functions — database functions with auto-inserted parentheses
• SQL keywords — SELECT, FROM, WHERE, JOIN, and the rest
• Snippets — common query patterns like SELECT ... FROM ... WHERE

Column completion works lazily. When you type tablename., the editor fetches that table’s columns on demand and shows them with their types and primary key markers. Columns are sorted so primary keys appear first.

Tip

Column suggestions load the first time you type tablename. for each table. After that they’re cached for the session, so subsequent completions are instant.

Query Execution

Run your query with Ctrl+E (or Cmd+E on Mac). Results appear in the data grid below the editor.

Explain runs EXPLAIN (ANALYZE, FORMAT JSON) on your query and shows the execution plan. This is available for PostgreSQL and Redshift connections. Use it to understand index usage, join strategies, and where time is being spent.

Optimize with AI sends your current query to the AI Assistant with a prompt asking for optimization suggestions. The AI panel opens automatically with the query pre-loaded.

Cancel stops a running query. Useful when you accidentally run a slow query against a large table.

Tip

If you have multiple statements in the editor and want to explain just one, select it first before pressing Ctrl+Shift+E.

Snippets

Snippets let you save queries you run often. They’re stored locally and available across sessions.

Saving a snippet: Press Ctrl+S (Cmd+S on Mac) while in the editor. Give it a name and optionally assign it to a folder.

Scope: Snippets can be global (available in every connection) or connection-scoped (only visible when connected to a specific database). Connection-scoped snippets are useful for queries that reference tables specific to one project.

Folders: Organize snippets into folders using a path like reports/monthly or admin/cleanup. The snippet browser shows them in a tree.

Built-in snippet templates appear in autocomplete alongside your saved ones. Start typing SELECT, INSERT INTO, JOIN, or CREATE TABLE to see the templates.

SQL Formatting

The editor can automatically uppercase SQL keywords as you type. When you finish a word like select and press space, it becomes SELECT. This keeps your queries consistent without any manual effort.

You can toggle this behavior in Settings > Editor > Auto-uppercase keywords.

Themes

The editor respects your app theme. Switch between Dark, Light, Nord, and Dracula in Settings > Appearance and the editor updates immediately — no restart needed.

Keyboard Shortcuts

Action	Windows / Linux	macOS
Execute Query	Ctrl+E	Cmd+E
Explain Query	Ctrl+Shift+E	Cmd+Shift+E
Optimize with AI	Ctrl+Shift+O	Cmd+Shift+O
Save Snippet	Ctrl+S	Cmd+S
Trigger Autocomplete	Ctrl+Space	Ctrl+Space

Tip

Autocomplete triggers automatically as you type. If the suggestion list doesn’t appear, press Ctrl+Space to open it manually.

MongoDB and Redis

When connected to MongoDB, the editor switches to JSON mode. Write queries as JSON objects:

{ "collection": "users", "action": "find", "filter": { "active": true }, "limit": 100 }

Autocomplete provides MongoDB-specific snippets for find, count, insert, delete, and common filter operators like $gt, $in, and $regex.

Redis connections use plain text mode. Type Redis commands directly:

HGETALL session:abc123