In my last post, i shared that we can use Fiddler to bypass cloudflare protection. I used it for a few days and realized that Fiddler is not stable. I decided to do more research. It turned out that Cloudflare use some signature in TLS protocol to identify the client. Chrome/Firefox/Edge using different library. Luckily there are some library developed to simulate these TLS protocol. The tool name is curl-impersonate , check it out here https://github.com/lwthiker/curl-impersonate#Advanced-usage
I have a thought about this, Cloudflare spends a lot of efforts on this to detect the bots. But now, we know where cloudflare get the data and we can trick it, so the protection isn’t worth much. It can only protect for normal user, but for anyone who writing bots, they will know this sooner or later and they can overcome this easily.
I guess when cloudflare sell this feature to the client, the client normally don’t know this trick and they still think that their site is fully protected. They should think about some other solution.