Ask the verifier to check a fix on the live site

Ask the verifier to check an issue on the production site after your fix is deployed. It answers at once with the verification it started; the verdict comes later. Without urls it checks the issue's affected pages; urls narrows the check to some of them (at most 50, each one of the issue's own affected URLs). A rule judged over the whole crawl, or an issue wider than 50 pages, is checked by the next finished crawl instead: waiting_on.reason is crawl_needed. Pass the version you last read as expected_version. With an idempotency_key a retry answers the first call's verification (replayed: true) instead of starting another. issues_report_deployed starts the check itself: while that check is pending, this answers it (replayed: true) instead of starting another, when it names no urls or only pages that check holds. Next: wait for the verdict with verifications_get, and accept it, failed and inconclusive included.