Runnable Anthropic engineering practice

Build a same-host web crawler

Edit the Python starter, run three browser-based checks, and rehearse URL normalization, cycle safety, and bounded concurrency.

Run the free exercise

Practice workspaceOpen

1 Read question2 Draft response3 Review notes

Anthropic · Coding

Web Crawler

Must priorityMedium

The question

Discover and visit pages within a domain boundary. Executable browser contract: Implement crawl_same_host(start_url, pages). pages maps an absolute URL to the links found on that page. Visit the start URL and every reachable HTTP(S) URL whose hostname matches the start hostname. Resolve relative links, remove URL fragments, avoid duplicate work and cycles, and return URLs in deterministic breadth-first discovery order. Reject a start URL without an HTTP(S) scheme or hostname.

Try a first passNo account required. Nothing you type here is sent or saved.

Run the baselineMake the crawler survive real checks.

Python runs privately in this browser.

Loading the Python editor…

No account required · code is not uploaded or saved

Keep your crawler workSave this run and open 12 free Anthropic questions.

A free workspace keeps this code and your progress. Upgrade only when you want every Anthropic question and contextual AI review.

Save run & open 12 free questions Unlock all Anthropic questions · $44.10 first month

Free needs no card. Premium is then $49/month; cancel anytime. Secure checkout follows account creation.