Clover Scheduling needs a small workload dashboard. The fictional fixture records contain an id, label, amount and status. Deliver a semantic table with filtering, numeric sorting, visible totals, an explicit empty state and a complete reset.
A dashboard becomes unreliable when its table, total, and controls each use a different copy of the data. Build one visible-row pipeline instead. Filter the source rows, sort a copy when needed, and derive both the table and its total from the result. This keeps a narrow Open filter from showing a total for the entire workload.
You'll carry the same table through six steps. The first step establishes semantic column and row headers. Later steps add independent controls and their combined boundaries. Sorting must be numeric, searches must handle a query with no matches, and reset must undo every control. The last step returns to the original fixture order rather than leaving a hidden sort active.