User agent
RankLatchBot/0.1 (+https://ranklatch.com/bot; SEO monitoring)
RankLatch uses the same user agent for customer monitoring and its public one-page SEO tools. Requests do not carry a customer login, CMS credential, or browser cookie.
Public pages only
RankLatch accepts HTTP and HTTPS URLs on standard web ports, rejects local and private network destinations, and follows at most five server redirects. It reads the server response without executing JavaScript and stops after 12 seconds or 2 MB of page data.
For an active monitor, RankLatch discovers eligible public URLs from the homepage and common XML sitemap locations, then checks only the number of pages included in the customer's plan. Scheduled monitoring normally runs once per day.
robots.txt is enforced before page requests
Before requesting a page, RankLatch reads the origin's robots.txt and evaluates the path for the RankLatchBot group, falling back to the wildcard group when no specific group exists. The same check runs before following a redirect to another origin.
To block all RankLatch requests, publish:
User-agent: RankLatchBot
Disallow: /A more specific Allow rule can keep selected paths available. RankLatch also respects blocks applied at a firewall, CDN, or origin.
No bypass and no speculative fetch
A missing robots.txt file or an ordinary 4xx response other than 429 is treated as having no crawler rules. If robots.txt times out, cannot be verified, is rate-limited, or returns a server error, RankLatch fails closed and does not request the target page. Customers see a fetch error when a monitored page is blocked.
RankLatch does not attempt to bypass authentication, paywalls, access controls, rate limits, or bot protection. For unexpected traffic or abuse reports, email abuse@ranklatch.com with the domain and approximate request time.