Glossary · Data & retrieval

Crawler

A bounded, breadth-first reader that follows links from a start URL under path, depth and page limits.

What is crawler?

A crawler starts from one URL and walks the links it finds, subject to a page ceiling, a depth ceiling, path globs, allow and deny patterns, and a same-host rule. Each page it reads is billed, which is why the scope is a set of explicit bounds rather than a suggestion, and why the page cap also caps the credit reserve.

Why it matters

The failure mode of crawling is not technical, it is economic: an unbounded crawl of a large site is a bill. Scoping a crawl from a map of the site first — a flat-fee discovery pass — is the difference between reading the two hundred pages you wanted and the two thousand you did not.

Related terms

See it in context

Back to the full glossary

Put your agent to work. Keep the decision.

Give your agent a real browser, authorize the accounts it may use, and keep control of every step. Repeated runs compile into skills that replay without a model.