Post your own problem compositions!

I have a number of atomic chess endgame studies (see my lichess studies... studies of studies?), some orthodox proof games and retros, and some slightly odder ones. Here's one of my better attempts.

Phénix 287 (July-August 2018)
Add a black queen for a PG 9.5 (add a black queen somewhere, then reach the resulting position after white's 10th move)
lichess.org/editor/rnb1kbnr/p3pppp/8/8/2P3P1/2N2N2/PP2PPB1/R2QR1K1_w_-_-_0_1

Made this puzzle from one of my games. :]

@Groove_And_Chess
These are neat studies, especially the second one looks familiar to me. It is funny (although not surprising) that even in the end, the engine still thinks it's winning for Black.

@Skittle-Head
Nice! I think this is even a mate in 2, since 2....Qd1# releases the White King even faster from his struggles. If you remove the pawn on c3, your solution should be correct :)

Here is another stupid little one which I have just found in my old documents.

This is one of the few puzzles I've composed. I would recommend trying to calculate this one from start to finish.

Here's a Mate in 3 puzzle I've composed. lichess.org/study/Q1q8IcI7

This is a simple one I composed to show that stalemate is useful sometimes.

Mate in 4:

lichess.org/editor/k7/P7/1K6/P7/6p1/6N1/RP6/8 w - -

You all make nice composers! :)

This is a Crazyhouse problem that highlights one of lichess's shortcomings regarding the variant.
Specifically, you can't tell the difference between a real queen and a promoted pawn until you capture it.

lichess.org/analysis/6bk/5prp/4p1pK/2BpPp2/3P1P2/8/8/8_w_-_-_0_1#0

White to DRAW, not win