Most test frameworks follow a "runner mentality"; you write test functions and hand them to a runner that controls discovery, execution order, and flow. The moment you need something dynamic (loop a test N times based on a runtime value, skip a test based on another's result), you're hunting for plugins. We present an alternative that is, in a way, a return to basics: treat your tests as a regular program. Dependencies become if statements. Retries become while loops. Parallel execution becomes a function argument. No plugins — just code, just Python. We use TestFlows (pip3 install testflows) to illustrate the idea, but the core argument is framework-independent: test code should have the same expressive power as production code. Have you hit the ceiling of your test runner? How do you handle dynamic test flows today?