I like the guildmaster having an ulterior motive, maybe even being corrupt. Gives the game some inner conflict, and helps to differentiate it. Maybe no one knows he's dying, and he doesn't want anyone to know that he's forging "fetch rare herb" quests to get ingredients for his life-prolonging alchemy. (Maybe all those rescued princesses are being turned into medicine, too.) Those quests are the stick; every month, you have to have looted a particular combination of items or game over. It doesn't have to be a sharp stick, it could be a minor thing that ramps up slowly as you play, just so long as it gives you a motivation to level your fighters.

But you don't want to send the same adventurer on too many herb quests or they'll figure out something's up. And adventurers that figure out that something's up have to have "accidents". So you don't want every quest to succeed: you want to have most quests succeed (to level up your adventurers for the required quests), but you want some of them to fail, so as to get rid of suspicious adventurers. (But not fail so obviously that it creates MORE suspicious adventurers!)

In terms of a quest mechanic, I like the puzzle idea. Maybe you have a quest-launching area with set of cards/tiles/whatever representing challenges (monsters, obstacles, locks, etc.) and can drag adventurers onto it with their own matching icons (so this adventurer has nine hearts; this one has three hearts but can climb a cliff and has three lockpicks). A few cards are upside-down so you don't have complete knowledge (but DO have plausible deniability). If you match every challenge, they'll come back with the loot (gold, herbs, princesses, etc.). If an obstacle isn't matched, then they'll come home empty handed. But for each *enemy* that isn't matched, it's roulette time and a random adventurer on the team dies. (Hopefully that suspicious meddler! But you can never predict...)

I want to add politics, friendship, morality and many more stuffs but would those be interesting even if the core loop isn't fun? I will consider them later.

I'm not sure why you consider them separate, the whole integrated design is what makes games fun not just one particular mechanic.

From the look of things the game has 3 phases.

The planing phase that the player is in full control, and here you have to cram as much things as possible for the player to do and thus affect the game with.

The testing phase that resolves the mission/dungeon. It is essential to cram as much "oh shit" factors that the players have to predict and account for as well as risk.

The testing phase IS the challenge, the mission/quests should be an absolute disaster waiting to happen.

Maybe make it so you shouldn't complete them in one go so that it takes multiple turns.

For example first you would explore and scout the monsters rather then engage them and gather as much information.

Next maybe you test the waters and engage in some limited encounters and from that feedback you plan better.

And only after you are confident you send to to completely clear it.

And after that maybe you optimize to be a resource extracting expedition.

Dungeons should be deep, dangerous and unpredictable.

This is for dungeons for missions you have different factors to account but still spread out the completion over multiple turns as long as you are not over the time limit.

The third phase is the outcome, heroes killed, what has been looted and how the mission has played out and thus problems you have to manage out of.

IF you send in the wrong heroes on the mission you may have botched the mission irreversibly so you have to mange the fallout.

As long as you have a good first phase where you mange stuff and how good the challenge is in the second phase is how fun, engaging and interesting the game is.

As long as you haven't completed this phases it is pointless to talk about "core loop", got it?

I am still choosing between writing a "oh shit oh shit oh shit simulation game" that you can never beat or have no ending, and a slow pace game that just let you build slowly and enjoy the "story".

They are not mutually exclusive,you can have both as long as they are interesting.

I remember playing this game "Rebuild" where the player send people out on missions etc. They did it quite nice in the sense that the decision of the player matters and there is a overall goal for the player. However, if there is a danger that is threatening the "world" then gold and rewards for the guild don't really make sense, but wait, every single save the world game have "gold" in it =/

Gold doesn't need to be motivation, saving your ass can be just as good.

In fact profitability should be pretty independent, from doing missions and quests. It is up to you in how much ruthless you are to be profitable, most people want to be saved for free and aren't exactly swimming in gold.

Like all profitable things, guilding is a racket. And since you are one of the surviving ones you are pretty ruthless indeed, which ties in to the moral thing.

A guild has taxes and expenses and you have to pretty much be a bitch to a lordling to keep afloat which ties in to the politics and backstabbing between factions.

You pretty much are mercenaries with some polish to keep you shiny.

Some charity missions are nothing but a PR campaign so that you look somewhat legitimate and keep you from everyone strangling you in distrust.

At least that would be the realistic grim view.