When you're stuck, the problem usually isn't the answer. It's the question. You're solving the wrong thing perfectly. The right question gets you out of the problem entirely.
Filter
"How do I stop X?"
You're building a dam on a river that shouldn't exist. Ask instead: "Why does X exist at all? Can I kill it at the source?"
Constraint
"How do I do X within Y?"
You're treating a limit as fixed when it might not be. Ask instead: "Is Y actually a real constraint? What if it didn't exist?"
Symptom
"Why isn't X working?"
You're assuming X should exist. Ask instead: "Should X exist at all? Am I fixing the right thing?"
Tool
"Which tool for X?"
You've locked in on a task before checking the goal. Ask instead: "Is X even the right task? What's the goal behind it?"
Permission
"Is X possible?"
You're waiting for permission that nobody needs to give. Ask instead: "What's the actual goal? What's the simplest path?"
Before you try to fix anything, run these three questions. If you can fix at the source, the filter becomes unnecessary.