A small change, with a before and after
You need to fix a few rows. First, save a backup. Make the change, then check what actually moved. Here’s that whole workflow in dbterm, using a small SQLite database you can throw away afterwards.
This is useful when you’re checking a seed script, testing an app action, or cleaning up development data. If you spend your day in a terminal, you can keep the query, results and backup in the same place.
1. Open a practice database
Install dbterm, then run dbterm. On the Dashboard, press N, choose SQLite, and use a new file called dbterm-demo.sqlite in a writable folder. Save and connect. Use a separate demo file for this walkthrough.
Press Alt+Q to focus the query editor. Run these statements one at a time with Enter:
CREATE TABLE tasks (id INTEGER PRIMARY KEY, title TEXT NOT NULL, status TEXT NOT NULL);INSERT INTO tasks VALUES (1, 'Write the guide', 'todo'), (2, 'Check the release', 'todo'), (3, 'Fix the typo', 'done');SELECT * FROM tasks ORDER BY id;You should see three tasks. Shift+Enter adds a new line without running your query. Ctrl+Space opens SQL suggestions.
2. Get around without reaching for the mouse
Use Alt+T to focus Tables. Type tasks, then press Enter to open it. Press Space on the table to pin it for next time.
Alt+R focuses Results; Alt+Q takes you back to SQL. If you forget where something lives, Ctrl+P searches commands, objects and recent queries. These are the default shortcuts; your saved settings may differ.
3. Take a backup
Press Alt+B for an instant backup. Choose an absolute destination folder, such as C:\dbterm-demo-backups on Windows or a folder under your home directory on Linux/macOS. Wait for the run to finish successfully before continuing.
SQLite snapshot backups are built in. PostgreSQL and MySQL backups need their native dump tools installed. For recurring work, Alt+K opens Backup Center, where you can create a schedule and configure the background agent. The backup guide covers that setup and restore checks.
4. Set a checkpoint
Press Alt+W to open Change Profiler, then N to create an anchor. Call it Before task update, include the tasks table in the plan, and capture the baseline.
An anchor is your reference point for comparing data. Keep this first one small: capture and scan both read the selected rows.
5. Make a change and check it
Return to the query editor and run:
UPDATE tasks SET status = 'done' WHERE id = 1;INSERT INTO tasks VALUES (4, 'Record the demo', 'todo');Open Change Profiler again with Alt+W. Select your anchor and press S to scan. Open the report with Enter.
You should find one updated row—task 1 changed from todo to done—and one inserted row, task 4. The other two tasks should be unchanged. Use Tab to focus the row list, select the updated row with the arrow keys, and press Enter to see its full before/after values. Close that view and return to the anchor list with Esc, then press F for the final scan and to finish the anchor; the compact report stays available.
The report shows observed changes, not who made them. It isn’t an undo button or a continuous audit log. Your backup is the separate recovery artifact.
Where this earns its place
- Debugging: capture an anchor, exercise a feature in your app, then inspect the affected rows.
- Small data fixes: find the table, back up, run a precise update, and review the result.
- Repeat visits: saved connections, pinned tables and query history save you setting everything up again.
For a PostgreSQL or MySQL server, you can also save the login without a database name and browse the databases that account can access. SQLite keeps this first walkthrough free of server setup.
That’s enough for a first session. Install dbterm, try the four-row example, and use the full guide when you need more detail.