How do you decide which URL patterns should be crawlable?

For every pattern, determine:

• Should Google discover it?
• Should Google crawl it?
• Should it be indexed?
• Should it have a self-canonical?
• Should it appear in XML sitemaps?
• Does it represent unique search value?
• Can the URL space grow uncontrollably?

For example, core product and category URLs are normally designed to be crawlable and indexable, while internal-search and session URLs are commonly controlled.

The decision should come from the site's information architecture and search intent, not from arbitrary crawler-tool scores.