• Why is there 1000 random tests? Is this many really necessary?
• All random tests expect answer of 0, 1, 2, or sometimes 3. Fixed tests show that the domain can be much larger. Why are random tests so limited w.r.t. expected answers?

What are the constraints for subtask 6? (It seems to be much harder, than stress test.)

You should write the constraints in the description.

It would be nice to see more Sample Tests and also Fixed Tests, one that don't jump right to being so large.

Missing imports in sample tests

• Missing imports in sample tests

• The input range is not sparse enough, so solutions from subtask 2 can easily pass the tests

Initial code is incorrect: the actual function needs to be implemented is jump_to_zero, not sum_dig