Notes from the bypass trenches.

A curl request and a browser request used to look the same on the wire. They do not anymore. Here is what actually changed, why your scripts started returning 403s, and what the response means.

Why your spoofed User-Agent does not matter, how JA3 and JA4 actually work, and what curl-impersonate does under the hood. A field guide to the slowest-moving arms race in scraping.

Headless browsers solve every scraping problem and cost ten times what they need to. Here is the decision framework we use to pick between recurl and a Playwright/Puppeteer pipeline.