You accept a Tab suggestion that changes a function's signature. The same function is called in three more places further down the file, and each one now needs the same treatment. Without touching the mouse, you press Tab once more straight after accepting. What should you expect?

The principle — Accepting a suggestion is not the end of the interaction. Cursor documents that after you accept a Tab suggestion you can press Tab again, and Tab predicts your next editing location and jumps there. The second press is navigation, not insertion.

Why the key is correct — A signature change is exactly the shape of edit that repeats: every call site needs the same treatment, and the editor has already worked out where those sites are. Pressing Tab again moves the cursor to the next one, so the scrolling and clicking you would otherwise do by hand is handled for you. The real benefit is not typing speed but coverage — the sites you would have missed while scrolling are the ones that break the build later.

Why the others are wrong — Falling back to inserting a tab character is what the key does in ordinary typing, but immediately after an accepted suggestion it is bound to the jump instead, which is the whole point of the feature. Regenerating the same completion would be useful in a tool that cycles alternatives; here the prediction is about where you are heading next, not about other ways to write what you already accepted. Waiting for more typing misreads the trigger: the jump is offered on the key press, so a user who assumes nothing will happen simply reaches for the mouse and does the work manually.

Remember this — After accepting, press Tab again and let it take you to the next place the change has to land.

Sources — Cursor's Tab completion documentation.