A while ago there was a discussion about clients, and I commented that it'd be interesting to embed a mud client as a Facebook application (some of the Facebook applications have literally millions of users). Recently I've seen at least three muds do exactly that - to which someone commented that muds are too complex to appeal to most Facebook users. And I think there's probably a fair amount of truth in that statement.
But it got me thinking about the minigames that many muds offer. In particular, the recent discussions about player-generated content as a minigame, and the parallels I drew between farming games and Tonitrus's area-building system.
It strikes me that it would be relatively straightforward to create a simple dungeon-building game along the same lines as those Facebook farming games. Think of it as a very simple Dungeon Keeper style game, except without the conflict. You have a grid of tiles, and can put down corridors, chambers, etc, then populate them with monsters, traps and treasure. The building effort would be minimal, but as your dungeon grew you'd unlock more tile options, new types of monster, etc. You'd also have to maintain the dungeons by feeding the monsters, cleaning away corpses, etc. Minimal effort, but enough to get people to log on once every day or two (i.e., the same model as used by games like Farm Town).
I've been thinking of creating something along these lines anyway, but in-game. However there's no reason why it couldn't be a Facebook application, and the mud could simply import the dungeon data for the mud players to use. The Facebook dungeons wouldn't be directly effected by the actions of the players, but perhaps they could rack up a score based on their popularity, (mud player) kills, etc.
I could see same approach being used for other minigames as well - perhaps some sort of crafting game (with the demand and pricing based on the needs of the mud, and the end products being sold within the mud), or city building game, etc.