My Blog's Bot Detection Without JavaScript: 74.5% of Browser Requests Were Not Human

Bot Detection Without JavaScript: What My Blog Measured

A Cloudflare Worker-based classifier on a personal blog used network metadata, Fetch Metadata, and header rules to separate bots from readers without JavaScript. Over two days, it moved 277 of 372 browser-User-Agent requests (74.5%) out of the Browsers category, revealing a stark gap between edge counts and Cloudflare Web Analytics. The author explains the rules, the importance of retaining classification reasons, and the limits of the evidence: client identity and readership require different signals, and the remaining disagreement is a measured problem, not a resolved one.