This discussion has an example of how I would write out a mission. Listing potential obstacles and twists should take about an hour. My typical approach has been to create one each of the four types (wilderness, weather, animal, mice), and present two as mission hazards and leave two in reserve as twists. You don’t have to introduce twists off failed tests, either–applying conditions is more than acceptable, and keeps the session moving along.
In Mouse Guard, remember that failing tests means that the characters still succeed, but at the cost of conditions or twists. You don’t need to focus your preparation on all the possible outcomes of a situation. It’s not the cleanest example, but when the patrol tests to discover the map in the Find the Grain Peddler mission, they find the map regardless of whether or not they meet the scout obstacle. Perhaps all their searching attracts the attention of a hungry animal (twist), or they become tired following the effort (condition), however they always find the map.
Mouse Guard isn’t a game about whether or not the patrol uncovers a traitorous map or successfully pathfinds their way to the next settlement; it’s about the price they pay to complete their missions and–most importantly–how their Beliefs, Instincts, Traits, and Relationships influence their duty as guardmice. Your campaign’s just getting started, but please let us know if you have further questions on any of this.