trying to make small test cases and solving them manually works for me most of the time.

Actually there isn't. Firstly if your solution has a lot of edge cases, then you should probably look for some better solution, because the correct ones mostly don't have many edge cases.

Secondly, you should emphasize on thinking the solution thoroughly, before implementing. This is a good practise and will save you a lot of time later.

Thirdly, once you solve a lot of problems, the bugs and edge cases are mostly similar. You'll eventually become good at identifying them.

You can also use these two links https://codeforces.net/blog/entry/18087 https://codeforces.net/blog/entry/44493#comment-290139