Support

I cannot reproduce this issue but I know what can cause this. Akeeba Solo requires that session support is correctly set up in your PHP. The security token is saved in the session when you access the edit page. When you submit the new user form it includes a token. That token is compared to the one in the session. If the tokens do not match an error is issued. If the PHP session doesn't work correctly then the tokens will, indeed, not match as a new session will be opened on every page load. This means that a new token will be generated in each page load, hence the issue.

Check your php.ini to determine where there PHP session save path points to. This path must exist and must be writable by the user under which PHP runs. Moreover, the contents of that folder should not be periodically deleted (as is typically the case with /tmp and its subfolders).

Other issues which might also cause this problem is having a caching proxy or CDN in front of your Akeeba Solo installation. In this case you are served an old copy of the edit form. The old copy of the form also includes an old token which is no longer valid. Therefore the token check will fail. Since this depends on your server setup I can't tell you how to check for it.

If all else fails and you want to go the manual route, please note that the password in the table is stored hashed using bcrypt. If you search for "bcrypt online" you will find several pages. I haven't found one which doesn't need to send the password to the server to hash it, therefore I cannot trust them with real world passwords. If you're comfortable with CLI I'd recommend using this instead: https://www.npmjs.com/package/bcrypt-cli It's worth noting that every time you create a hash of the same password you get a different result. This is by design of bCrypt. It creates a new, truly random salt on every invocation. Also, if you are wondering about the number of rounds: use 10. The more rounds you use the harder it is for someone to "break" the hash and find a password which collides with it. With modern hardware a value of 5 is enough, whereas a value of 10 should be just about right for the next decade (barring any unforeseen breakthrough in computing).

I hope this information helps!