Can you solve this seriesmate?

Here is a (probably flawed) seriesmate I recently composed. Hope you like it!

It would be greatly appreciated if you could share any duals or cooks that you found

Edit: The proof is slightly wrong. This should fix it:

a seriesmate is...? I forgot.

A seriesmate is a mate where white makes a series of moves to mate black with no reply. In other words, a directmate but only white is moving.

That's a most interesting concept! There is one obvious pretty little thing wrong (last diagram) which is that there are 9 white pawns on the board when you castle during the solution. Because of course then Rh6 could not have come from h1 before the king moved. When you fixed it I'll look at the other parts!

Btw, there is no standard concept of "past" for problemtypes for which no game rules exist (how would you play a seriesmover game?) but most solvers would assume you played the past by orthodox chess rules just as you did. There are however many fairy types for which game rules are easily derivable and their past can be played with those rules. The latter is gaining popularity while the field is under development. Think for instance of the simple cases of Chess960 or Capturing Chess histories.

How are my three solutions?

@KMMCS88 and @Arisktotle thanks for finding those cooks! Definitely not what I intended!

Btw, I assume that 1 of KMMCS88's cooks is no cook for the reason I gave in my cook. That won't help you though to get rid of the other 2.

Nope, none of my cooks relied on illegality.

Didn't you castle in one of your 3 cooks? 

Oh, I didn't notice that problem. Yeah, one of mine was that one.

Is this better? Seriesmate in 3

Did I get it?

Take the last move back and ask "how did that bishop get to f8?".

@Arisktotle What do you mean? 

@KMMCS88 Wow, you are incredibly good at finding cooks !

As long as you can find cooks (undesirable solutions) you need not get what the author intended. But to get the authors idea you need a good understanding of retrograde logics.

The moves were in this order: 1 ... g6 2 Bh6 3 Bf8 g5 4 ... Rh6

I suppose you still want g7-g5 to be blacks last move. How then did the bishop arrive on f8? Or did you change the whole solution?

But then the seriesmate would be in two, not three.

Aha, you are approaching mark X! This is the point where you need an understanding of retrograde logics. I can attempt to explain it but not until Graywing13  has ended up with a workable version.

I think this works