Designing a ladder teaches more about the graph than solving one.
The deepest form of practice is building a ladder instead of solving one. Start with a word, choose a target three or four rungs away, and then check that every step is legal and that the route is actually reachable. Building a valid ladder is harder than it looks, and the difficulty is the lesson.
The first thing building teaches is how the graph is shaped. A route that looks obvious when solving can be impossible when constructing, because the intermediate words have to work in both directions. Players who build a few ladders stop believing that any promising first move must be reachable.
The second thing it teaches is where difficulty comes from. A ladder can be made hard by choosing a target in another family, by choosing a starting word with few exits, or by leaving only one legal route through the middle. Each choice produces a different kind of puzzle, and seeing the choices explains why some daily puzzles feel unfair.
The third thing is humbling. Most first attempts break, and the breaks are informative. A construction that fails at the fourth rung is a map of the place where the vocabulary ran out.
This exercise is best done on a notebook rather than in the browser, and it needs no tooling at all. When a constructed ladder is finished, checking it manually is part of the exercise. Then have a go at the same route as a solver to see whether the construction was actually fair, and check the daily ladder the next day to compare a designer's choices with the game's own.
One warning: building can become a hobby of its own. The exercise is a practice tool, not a replacement for play, and a player who spends all their time designing ladders gets very good at design and no better at the game. A few builds per month is plenty, and the rest of the time should be spent on real boards.