Skip to main content
cloudflare.createCrawl starts an asynchronous crawl from one URL.
automations/create-cloudflare-crawl.automation.ts

Inputs

accountId and url are required. render defaults to true; set it to false for static fetching, which rejects browser-only page controls. Cloudflare defaults formats to html, source to all, depth to 100,000, limit to 10, contentUse to full, and crawlPurposes to all three supported purposes. Optional inputs also set cache age, modified-since timestamp, JSON extraction, include/exclude patterns, external links, and subdomains. cacheTtl defaults to 5 seconds. Rendered crawls accept the common Browser Run page controls.

Output

Returns { jobId }. Creating a crawl starts metered provider work and is not replay-safe. Use cloudflare.getCrawl to read progress and paginated records.