With this you could make sure that eg. PR's will pass all the tests - and then if you also use some code coverage stats you could track those too - so eg. if PR is made and tests passes but code coverage goes down - that is failure, because that PR didn't contain necessary tests to cover those new stuff.
With this you could make sure that eg. PR's will pass all the tests - and then if you also use some code coverage stats you could track those too - so eg. if PR is made and tests passes but code coverage goes down - that is failure, because that PR didn't contain necessary tests to cover those new stuff.