I posted to HN, and then my website was flooded with AI bot traffic

The site belongs to sprinklz.io, a customizable feed reader that can deprioritize topics such as LLM news. The earlier HN post drew roughly a hundred upvotes. Load kept climbing for days afterward, which did not fit a normal traffic spike. Per-endpoint logging showed most requests never tried to fetch articles, and because the site is a single-page app, they never ran the JavaScript needed to render it. NGINX logs showed that nearly half of all traffic came from one client. It describes itself as a service that watches sites for changes and turns them into an AI-written newsletter. The rest was mostly known crawlers, including ones from AI labs, plus scanners probing for WordPress files. The fix was robots.txt rules, a block on the worst offender, and rate limits. Commenters mostly voiced frustration with scraping. Many said that blocking tools such as Cloudflare challenges no longer work reliably. For AI builders, the lesson is about reputation and cost. A product that polls sites on a schedule, such as a monitoring, summarization or agent tool, can look like abuse to small operators. Honest user agents, published IP ranges, robots.txt compliance, caching and backoff are cheap to build. Skipping them invites blocklists and bad press.