This message was deleted.

We do have the concept of client "overrides" where you can turn tests on/off or adjust variation weights. If you load those overrides from a database or API, you can then manage experiments remotely without changing the code. Still working on streamlining that process

Gotcha. And I see that info is already in the readme