Browsing

The workspace has the connection’s databases, schemas and tables in a tree on the left (pin the ones you use), and tabs on the right. Ctrl/⌘ K opens the command palette anywhere.

A table opens in a grid that loads as you scroll, on tables of any size.

• Filter builds conditions joined by AND or OR (nested groups on paid plans). Sort by clicking headers, or several columns in the Sort panel. Fields hides columns. The view lives in the address bar, so a link opens the same rows.
• Search (Ctrl/⌘ F) looks in the text columns. On SQL databases, All in the search box also looks in numbers, dates, JSON and the rest, read as text: 1960-01-02 finds a birth date, 124.5 a price. It’s slower on big tables, since no index helps.
• Edit a cell by double-clicking it (Enter saves, Esc cancels). JSON and array values open in a code editor that colors and checks the JSON as you type; there ⌘/Ctrl Enter saves. Each save shows a toast with Undo. If someone changed the row since you loaded it, Wharfbase shows their value and yours and lets you pick.
• Space on a row opens the row form: every field, the rows it points to (foreign keys), and the rows that point back.
• Add row is in the toolbar; select rows to Delete them (or use the row form). Tables without a primary key (or another way to address a row) are read-only, and the grid says why.
• Structure shows columns, indexes, foreign keys and check constraints.
• Export writes CSV (up to 10,000 rows on Community; JSON and NDJSON on paid plans), with the current filter, sort and visible columns.

MongoDB collections open in the same grid, and whole documents can be edited as JSON. Edits keep every BSON type (ObjectId, dates, decimals). The console takes shell syntax (db.orders.find({ status: "paid" })).

Pipeline (MongoDB collections and views) builds an aggregation one stage at a time:

• Add stage starts from a template ($match, $group, $lookup …). You write each stage in shell syntax.
• The output after any stage shows next to the stages: its play button, or ⌘/Ctrl Enter in the stage. Turn a stage off to see the pipeline without it, and move stages up and down.
• It only reads. $out and $merge are refused there. Open in console puts the pipeline in a console tab, where writes follow the usual rules (read-only connections, production, approvals).
• Masked fields turn the pipeline off for that collection, since a pipeline can reach fields without naming them.

Redis opens a key browser: search with a pattern (user:*), filter by type, and edit strings, hashes, lists, sets, sorted sets and streams. TTLs can be set or removed. The console runs any command, except admin commands for non-admins and writes on read-only connections.