{"version":"2.1.286","anchor":"domain-check-rate-limit-error-tells-the-model-to-stop-retryi","canonical_anchor":"domain-check-rate-limit-error-tells-the-model-to-stop-retryi","heading":"Claude no longer loops when the web page safety check is rate-limited","tier":"notice","area":"Web Fetch","scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.286\/e\/domain-check-rate-limit-error-tells-the-model-to-stop-retryi","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.286","markdown":"### Claude no longer loops when the web page safety check is rate-limited\n\nWhen the safety check before fetching a web page is rate-limited, Claude is now told to carry on without the page instead of retrying\n\n**What**\n\nBefore fetching a web page, Claude Code checks the page's domain for safety. When that check is refused because too many checks were made too quickly, the message Claude receives has changed. It used to say to retry after about a minute. It now tells Claude:\n\n- not to retry in a loop or wait and try again repeatedly,\n\n- to continue the task without the page,\n\n- to tell you that the safety check was rate-limited,\n\n- that one later attempt is allowed.\n\n**Why**\n\nThis changes how Claude behaves when web fetching is rate-limited: rather than stalling your task on repeated retries, it moves on and tells you the page was skipped.\n\n- Area: Web Fetch\n- Tier: You'll notice\n- Useful: 2\/5\n- Signal: 1\/5"}