Undetectable account automation

Lets say I want to start automating and growing a couple of accounts. How do I make sure I don't give myself away through obvious boting tech. I understand that there are behaviour patterns.

• I am currently trying to automate logging in to a different site which uses google recapcha. used Chromium and Camoufox
• I launch a browser for manual logging in.
• I have 2 envs, one on server one local, local one logs in, the one on server doesn't. Same behaviour from me, same proxy ip, but somehow when run from server it doesn't work.
  
  Before any behaviour, how do I make sure I set myself up correctly with right tools?

xelex said:

Lets say I want to start automating and growing a couple of accounts. How do I make sure I don't give myself away through obvious boting tech. I understand that there are behaviour patterns.

• I am currently trying to automate logging in to a different site which uses google recapcha. used Chromium and Camoufox
• I launch a browser for manual logging in.
• I have 2 envs, one on server one local, local one logs in, the one on server doesn't. Same behaviour from me, same proxy ip, but somehow when run from server it doesn't work.
  
  Before any behaviour, how do I make sure I set myself up correctly with right tools?

Click to expand...

it might have less to do with ur actions nd more to do with the server environment, to be honest. compare ur browser versions, headers nd network conffigurations both locally nd on the server side first. captchas might detect environmental differences, so i would solve this issue first.

this issue usually comes down to environment detection and canvas fingerprints. if u run scripts on a server, always pair it with a residential proxy nd use an antidetect framework, much better than straight line automation

Don't overlook network-level fingerprints. If your User-Agent claims Windows/Chrome, but your TCP/IP packet profile reveals a Linux kernel, the mismatch is obvious. DNS leaks must also be prevented, ensuring no direct requests expose your actual location. Furthermore, if the proxy IP is registered to a datacenter ASN, the connection will likely be flagged immediately. And list goes on.
If you want to do it then it's not just tools that you need - it is infrastructure.

It's probably software rendering on the server, of theres no real gpu linux usually falls back and google might find that a bit suspicious
also most virtual servers setups barely have any desktop fonts installed, check cookies too. If your local chrome has months of browsing history but the server is totally clean that could be whats triggering the bot check

Impossible to guarantee undetectable automation; running Selenium with the --headless=new flag and adding random 12-30 second delays keeps actions below LinkedIn's 100-per-day limit, but their anti-bot system still flags scripted behavior