I personally think that it would be easier to just have a list of all the ways sleep clause can be broken. Going off the list Hip made, here is what I would classify as "breaking sleep clause"
Magic coat: legal (eh, could go either way on this one)
Metronome: illegal
Assist: illegal
Sleep Talk: illegal
Encored sleep: illegal
Encored unavoidable sleep: legal (i.e. you are trapped by dugtrio or mean look)
Trapped Choice Sleep: legal
Last pp sleep move:illegal
Effect Spore: Legal
Any move that you choose that has a chance of breaking sleep clause should count as breaking it. That is why I am iffy on Magic Coat, as it does preset the chance to sleep, but it all depends on what move they choose. Encore Trapped Sleep and Trapped Choice sleep moves are the only situation where the user has no ability to choose their next move, and effect spore is an ability which only activates when contact moves hit, and even then only sleeps 10% of the time, so I think that would be acceptable.
i'll use this for the base of my post because i agree with most of it. for me this all boils down to intent. i dont not agree that metronome should break sleep clause if it puts two pokemon to sleep (or one if you'd previously put another pokemon to sleep), because you did not intend to do so. it's not the strongest argument, but if people really want to state that metronome should result in a loss if it results in two intentionally slept pokemon, i will wonder aloud where the metronome hate is when it concerns OHKOs or evasion moves. i dont know how shoddy currently handles the usages of those moves if accessed through metronome and you're playing with ohko/evasion clause on though so if someone could shed light on that it would be cool.
i mean whatever, if i wanted to be a purist i could say that this actually should be illegal because "lol you should have known that you had a like 1-in-100 chance of sleeping that second foe pokemon and remembered that you intentionally slept a previous foe pokemon)", so i guess i'll chance my stance and say that this should indeed be illegal if only because it will "never happen" and that if you're using metronome in a serious battle anyway and already slept a foe pokemon you deserve to lose if you're really that lucky (unlucky?) to conjure up a second sleep move with metronome that hits
magic coat is very interesting, but shouldn't break sleep clause because the magic coat user did not intend to intentionally sleep two pokemon, and this is something that the sleep user could abuse more than the magic coat user. the magic coat user would have to know that the sleep-move pokemon both has the move, plans on using it, but most importantly that
the sleep-move pokemon does not also have another move that can be bounced back by magic coat. the magic coat user shouldn't be penalized for wanting to bounce back a possible twave or stun spore. and the sleep user should know full well that one of his pokemon has already been put to sleep, so that if he DOES use the sleep move, he could get fucked over by magic coat.
now the sleep user doesn't necessarily know that the magic coat pokemon has magic coat, but i dont think the magic coat user should be penalized for using magic coat on a sleep move if the user's trainer already intentionally slept a foe pokemon, especially when there is virtually no way of knowing that this sleep pokemon doesn't have another move that can be bounced by by magic coat. so i think that ideally, if:
1) the magic coat user's trainer has already intentionally put one foe pokemon to sleep
2) the magic coat user knows without question that the sleep-move pokemon has a sleep move
3) the magic coat user knows
without question that the sleep-move pokemon has no other moves that can be bounced back by magic coat (incredibly unlikely but this is possible)
4) the magic coat user successfully uses magic coat to sleep a second foe pokemon (meaning the foe used it, the magic coat user uses magic coat, and the sleep move actually hits)
then yes, the magic coat user's trainer should be disqualified due to
intentionally breaking sleep clause. this is the only scenario that seems fair to me, though. the magic coat user very much does have to know that the sleep user has no other moves affected by magic coat, or else we literally cannot state that the magic coat user's trainer
intended to sleep a second foe pokemon. please weigh in on my thoughts on this and let me know if i didnt consider anything
assist and sleep talk should be illegal because you know your own team and you know that there is a chance that you can select the sleep move. same goes for psycho shift as states previously—this is intent.
encored sleep is illegal since you have a choice to switch out, so by not doing so you very much intend to sleep another pokemon (and you can't argue that maybe they would have stayed in or switched in insomnia honchkrow since neither would not break sleep clause even if you're that silly). by the same token, trapped choice is not illegal because you did not intend to sleep TWO pokemon, just one.
last pp sleep is illegal for the same reason encored sleep is, you have the chance to switch to another pokemon. i dont think this should be illegal if this is your last pokemon, though, but lol you are going to lose anyway unless some silly sandstream stuff kills their 5% machamp and 4% vaporeon when all you have left is the faster breloom that only has spore PP left, because breloom still had no choice but to use spore.
finally effect spore should be legal if only because the FOE has to intentionally use contact moves on you. you would have to know that the foe only has contact moves when you stay in in a scenario like the magic coat one above, and even then if you successfully slept a second foe pokemon this is not your fault, as it would be with magic coat if all four of the possibilities i listed above are in effect, or as it would be in a roundabout way when you use metronome and conjure up spore or whatever