My setup is as follows:
AdsPower browser
Mobile phone with SIM card, with rotating IPs.
Automated scripts which run certain workflows with puppeteer with extra-stealth
Pupeteer navigates x.com with CDP. I also added additional code for even more stealth and CDP masking, like creating many synthetic events emulating mouse movement, etc. always using CDP actions for everything, not JS (like there is no element.click()).
So, i pretty much have everything covered:
* IPs - mobile IP which change after session. like 200000 pool of IPs
* No proxying - I literally connect my PC which is running the automation to mobile internet through USB cable
* Everything is automated: Ip rotation, connecting to internet, opening browser, etc. There is 0 chance of IP or fingerprint overlapping between different accounts
* AdsPower fingerprint - tested through various tools and benchmarks, seems consistent and unique. maintaining the fingerprint per account
* Pretty good automation scripts i guess. idle periods, simulating erratic mouse moves, smooth human-like scrolling and similar stuff.
* Irregular patterns. No particular account order, random execution time, even actions are randomized and different.
I do account registration manually, yet I already run into problems. sometimes getting try again later, always getting Arkose FunCaptcha challenges with 5 steps.
Still, can register few accounts per day with different email providers (my IP and fingerprint are always different).
So okay, I get it. Maybe similar typing patterns, maybe same country/IP pool, maybe unoriginal email account names, maybe providers like proton.me getting flagged.
BUT,
after running 10 accounts simultaneously with all of the mentioned security measures, and doing ONLY scrolling the feed, with some random post openings, reading comments, 10-20 minutes per day, I get suspended after few days. reason - https://help.x.com/en/rules-and-policies/authenticity
I do no get why.
The only possible reasons i see:
* AdsPower browser - maybe it is not consistent? maybe they leak something, or spoofing is detected?
* CDP. even if using adsPower browser, with stealth plugins, additional masking, etc., it may be still detectable. but another option is using some library to actually move the mouse (which would happen out of the browser scope - undetectable), which makes PC unusable, not so good option
* patterns, recurrences x.com detection AI model matches, but I do not see any.
Any thoughts, ideas?
AdsPower browser
Mobile phone with SIM card, with rotating IPs.
Automated scripts which run certain workflows with puppeteer with extra-stealth
Pupeteer navigates x.com with CDP. I also added additional code for even more stealth and CDP masking, like creating many synthetic events emulating mouse movement, etc. always using CDP actions for everything, not JS (like there is no element.click()).
So, i pretty much have everything covered:
* IPs - mobile IP which change after session. like 200000 pool of IPs
* No proxying - I literally connect my PC which is running the automation to mobile internet through USB cable
* Everything is automated: Ip rotation, connecting to internet, opening browser, etc. There is 0 chance of IP or fingerprint overlapping between different accounts
* AdsPower fingerprint - tested through various tools and benchmarks, seems consistent and unique. maintaining the fingerprint per account
* Pretty good automation scripts i guess. idle periods, simulating erratic mouse moves, smooth human-like scrolling and similar stuff.
* Irregular patterns. No particular account order, random execution time, even actions are randomized and different.
I do account registration manually, yet I already run into problems. sometimes getting try again later, always getting Arkose FunCaptcha challenges with 5 steps.
Still, can register few accounts per day with different email providers (my IP and fingerprint are always different).
So okay, I get it. Maybe similar typing patterns, maybe same country/IP pool, maybe unoriginal email account names, maybe providers like proton.me getting flagged.
BUT,
after running 10 accounts simultaneously with all of the mentioned security measures, and doing ONLY scrolling the feed, with some random post openings, reading comments, 10-20 minutes per day, I get suspended after few days. reason - https://help.x.com/en/rules-and-policies/authenticity
I do no get why.
The only possible reasons i see:
* AdsPower browser - maybe it is not consistent? maybe they leak something, or spoofing is detected?
* CDP. even if using adsPower browser, with stealth plugins, additional masking, etc., it may be still detectable. but another option is using some library to actually move the mouse (which would happen out of the browser scope - undetectable), which makes PC unusable, not so good option
* patterns, recurrences x.com detection AI model matches, but I do not see any.
Any thoughts, ideas?
