• Python new test framework should be used

• JS node 14 should be enabled

• Should have sample tests on output not limited to 0 only

• Description does not mention what does it count as having their backs covered by another pawn?????

• a pawn attack(and defend as well) only the 2 square on the sides in front of him. This is not the case for side-most squares. You should probably mention that attacks and defends can only be applied to diagonal squares in front of the white pawns, but bear not to exceed the boundaries of the chessboard or something equivalent

• NO example tests

• Other relevant information that is crucial in solving the kata should not be provided with a wiki link only

the random tests generation needs to be constrained, to be sure they generate every type of cases. Currently, there is clos to no chances an input generates True (hence the number of tests, I guess, btu that's not the way to go to handle this)

Hi,

• no need to do 1000 random tests. 100 should be enough
• the problem will have matrices of dimensions from 3 to 10 -> this is ambiguous. Prefer size 3 to 10, considering the context.
• it may happen in the random tests that no arrow is present in the grid:
  • either the random generation is wrong and needs to be fixed
  • or that case mus be in the description and the fixed/sample tests
• about the description, considering the saize of the example, you should put all the relevant information before them. You could use a 3x3x3 matrix instead of one of the 4x4x4 to make a point (visually) about the different sizes... and to reduce a bit the size of the description itself x)

Cheers

PS: I'm not sure either it's interesting, compared to the 2D version... Make it N-dimensioned? (the directions would need to be given in a different way, then... Signed numbers? as the indexe of the dimension to update, for example? Not sure that's actually better... The recursive part could be, tho)

Duplicate of katas with array manipulation based on indices

Forseeable possible problem:

Values like n = 17 and n = 1 (which the result consists of only 1 element) are indistinguishable to each other after handshaking.