Exclusions
Persistent rules that hide pages from this workspace — the table, counts, issues, export, and the agent.
A filter is the question you are asking right now. Clear it and the rows come back. An exclusion is different: it is a standing rule for this workspace. Matching pages disappear from the table, the row counts, issues, the export, and anything the agent queries. They stay gone until you turn the rule off.
Nothing is deleted. Exclusions hide, they do not remove. Toggle a rule off in the dialog and every page it hid comes back immediately.
Why this matters
Say the client site has a few thousand tag pages you never want to think about. Filtering them out gets them off the table for this session — until you clear filters, or open the workspace tomorrow, and they are back. Excluding them means they stay gone: out of your issue counts, out of the export, and out of what the agent sees when it answers a question about this crawl. That is the difference. Filters answer today's question. Exclusions decide what this workspace is even about.
Open the dialog
Click the exclusions icon next to Columns on the command bar. It dots when anything is hidden, and the tooltip shows the count.
Quick presets cover the usual junk: noindexed pages, URLs with a query string, redirects, error pages. Flip one on and the count updates.
Custom rules take a field, an operator, and a value — same shape as a filter, but it persists. Save a set as a named preset if you will reuse it on another workspace.
The dialog footer shows how many pages are hidden. Clear all resets the workspace to nothing excluded.
Ask the agent
You can ask the agent to hide a class of pages in a sentence instead of building the rule yourself. It proposes the exclusion and waits for you to approve before anything changes — this is not a silent action.