Your agent learns how a refund should work and who can invite a teammate. Keep those rules with references to the source, screens and flows that check them.
Run a user journey, capture the screen and compare it with an earlier snapshot. Your agent can check the visible difference against the change you intended.
Run your application’s own commands and keep their output for comparison. An exit code tells you the command finished; its result tells you what happened.
Test signups, email codes and magic links with a remote inbox. Capture the address, code or link as a flow variable and continue the journey in the browser.