Should a trader reduce risk after reaching a good profit?

Protecting progress can be sensible, but changing your risk after every winning trade may also interfere with a tested plan.

Do you keep risk constant throughout an evaluation, or follow a predefined scaling rule?