The usual heuristics for problem solving should apply. Solve easier similar problems, then put additional constraints on them and solve them anyway. Lift constraints until the problem gets solved, try to reapply them back.
For “consider why I might be wrong”, steelmanning opposing positions as views held in their own right seems like a better framing, rather than constructing something in opposition to your own views (someone might disagree with your views not for the sake of disagreement, but because their position just happens to be different). Which requires being comfortable with holding many contradictory views sufficiently seriously that their individual development isn’t damaged by not being actually endorsed (while not losing track of what is actually endorsed, when that’s relevant at all, which is not always).
The usual heuristics for problem solving should apply. Solve easier similar problems, then put additional constraints on them and solve them anyway. Lift constraints until the problem gets solved, try to reapply them back.
For “consider why I might be wrong”, steelmanning opposing positions as views held in their own right seems like a better framing, rather than constructing something in opposition to your own views (someone might disagree with your views not for the sake of disagreement, but because their position just happens to be different). Which requires being comfortable with holding many contradictory views sufficiently seriously that their individual development isn’t damaged by not being actually endorsed (while not losing track of what is actually endorsed, when that’s relevant at all, which is not always).