Crawl table
Work the crawl as rows: select pages, stack filters, and jump to the inspector or map.
After a crawl you land on the table. Each row is a URL plus the fields Silkra extracted. A default set of columns is visible. The rest are still in the crawl, just hidden.
Click a row to open the inspector. ⌘K focuses the command bar when you want to narrow the set.
Select pages
Click one row for that page. Shift-click or drag across rows to select a range. ⌘ click toggles one row without clearing the rest. ⌘A selects every visible row.
A bar appears on a multi-select: inspect the set, attach it to the agent, or deselect. ⌘L attaches the current selection to the agent without opening that bar.
Right-click a row for the same moves, plus a jump to the map on that page.
Three ways rows disappear
They are easy to mix up.
Filters are the current question. They show as chips. ⌘⌫ clears them. Command bar →
The issues tray (alert icon next to the bar) lists every finding. Select one issue filter at a time: missing titles, redirect chains, thin pages. Its chip stacks with the others. Issues →
Exclusions are different. They are workspace rules. Matching pages drop out of the table, the counts, the export, and the agent, and they do not come back when you clear filters. Exclusions →
When anything is hiding rows, the bar shows visible versus total.
Columns, export, map
⌘O opens the column picker. Show or hide fields by group. Hidden columns can still be filtered. Columns →
⌘E exports the visible columns as CSV, JSON, or a Google Sheet if Drive is connected. When filters are active, choose the filtered view or full table. Build custom makes a different field set, including full page content. Exporting →
Switch to the map from the title bar and the current filter stays on. Matching pages spotlight. The rest dim. ⌘T brings you back to the table.
Ask the agent to narrow the table in a sentence if you would rather not type the chips yourself.