It's quite light on specifics. It should have been straightforward for the author to excerpt some of the prompts he was submitting, to show how innocent they are.
For all I know, the author was asking Claude for instructions on extremely sketchy activity. We only have his word that he was being honest and innocent.
Maybe the problem was using automation without the API? You can do that freely with local software using software to click buttons and it's completely fine, but with a SAAS, they let you then ban you.