[andrewmcwatters]

Yes, but the fault lies with the banks who do not allow their own customers access to their data. Plaid, Intuit, and other private companies scrape financial institutions unless they provide more secure methods to obtain customer data, and most of them do not do this.

So the state of the art to connect to banks... is Selenium with stealth modifications.

I own a business which does the same work as Plaid, Intuit, et al.

[langcss]

An open source repo of selenium scripts for different banks would be a decent thing.

[andrewmcwatters]

I’m going to risk speaking for other people in the space and say that’s not actually a good idea since at the time of writing there are detection strategies that can shut all of us out of every financial institution if they so chose to do so due to recent changes with specific browsers.

It’s unsolved at the moment, and may not be for some time. So it’s a matter of time before the current scraping approaches break. We basically have to recompile the browsers from scratch to stop leaking information that distinguishes automation.

[zburatorul]

https://github.com/jbms/finance-dl