Destruction of a dungeon without crown approval, through destruction or theft of the dungeon core or any other means, is a crime punishable by a fine of 10 platinum coins. Destruction of any regulated dungeon is to be treated as an act of treason against the crown. As an exception to this rule, any plague type dungeon or dungeon carrying the errant designation may be destroyed via shattering of the dungeon core without prior permission, but the remains of the core must be turned over to an appropriate authority as soon as realistically possible, for which an appropriate reward will be paid.

- An excerpt from the common law of the Kingdom of Jetosu

Floors must exist on a two dimensional plain. That was a Rule, and great pity was due any dungeon unlucky enough to be born on a narrow hill top. To grow deeper new floors must be purchased with dungeon points. But why? What happened when a new floor was purchased? A staircase got built, and the core got relocated. Both of these were simple things that the dungeon could do with mana. So it did. At the back of its core room, the dungeon constructed a staircase downwards, telling itself that this wasn't really against the Rules because it was going to be a new floor and not an extension of the existing one. After dropping a sufficient distance, it began digging out passageways and rooms, finally constructing a room the same size as its original core room directly beneath. It then opened a temporary hole in the ceiling and dropped its core through. But as soon as the core left the room, all that the dungeon knew was pain.

Error: Dungeon core missing. Working... Core located on second floor.
Incongruity detected: Dungeon has only one floor. Working... Resolving conflict... Second floor registered successfully. New core room established.

The dungeon woke up with a hangover. That was new; it had never had a hangover before. For that matter it had never woken up before either, unless you counted the day it was born. But despite the pain, the dungeon was excited. It had worked! It had a proper second floor now. It could feel it! The dungeon immediately pumped all of its mana into a slime summoning.

Error: Attempted to summon a monster too high level for this floor.

Arggg, more exploding mana was really not what it wanted to feel right now. Perhaps all of its mana was a little too much. The dungeon had felt that thoughts were flowing a little more smoothly after awaking, but apparently it still had a way to go if it was still making stupid mistakes like that. After waiting for mana to regenerate, the dungeon tried again with half.

New monster unlocked: Big slime

The dungeon stared at its new monster as best it could given its lack of eyes. It looked identical to the regular slimes, except that it was larger. It didn't appear any more intelligent than a regular slime, but that didn't matter. The dungeon had proven the concept and knew it could create new monsters without expending dungeon points. It could take its time to expand the number of floors and range of monsters.

The dungeon explored all of its monsters, increasing the mana pumped in one at a time. Giant leeches gave nothing new at 5 mana. There was something at 10 mana, but it again needed a higher floor. At 5 mana, attempting to summon a dog produced a lesser wolf, but there was nothing up to 10 mana. Producing a third floor went a little more smoothly than the second now that the dungeon knew what to expect, allowing an unlock for giant slimes and carrion crawlers.

It took the dungeon a full month to build up and populate the second and third floors. The problem was its low mana regeneration, but time was something the dungeon had in abundance. Unfortunately, maximum mana capacity was not something that could be worked around with time, and it had reached the limits of what monsters could be summoned by overcharging existing summons with only 10 mana. Increasing mana capacity and regeneration required reinforcing the core, which was another option dungeon points could be spent on. But while it was apparent how to construct a new floor manually, how would the dungeon go about reinforcing its core? Furthermore, the core was important. The dungeon could lose monsters, walls, rooms or whole floors without a problem, but the core was vital. The dungeons brain and heart rolled up into one neat crystalline package. Experimenting on the core and getting something wrong would mean death.

Nevertheless, if the dungeon wished to continue to grow, it had to try something new. Reinforcing the core allowed it to store more mana, so perhaps what was required was to try and force more mana into the core than would normally fit. For safeties sake, the dungeon decided to begin its experiments elsewhere.

Building a second pedestal in the core room, it placed a single copper coin on it and tried to force in mana, to no effect. The mana didn't interact with the copper at all. Repeating with a silver coin caused the coin to shatter with a single point of mana. That was not good, and the dungeon was relived it hadn't tried that on its core. The coin was very small so perhaps it needed a smaller amount of mana, but what was smaller than 1? The dungeon produced another coin, starting to move in a point of mana, but part way through struggled to stop and withdraw a part of the mana. It took a few attempts, but in the end the struggles were successful.

New skill unlocked: [Mana Finesse]. Smaller units of mana can now be controlled.
Incongruity detected: [Mana Finesse] not available for purchase until dungeon has ten floors. Resolving conflict... Purchase requirements for [Mana Finesse] reduced to three floors.
Mana: 6.5/10 (regeneration 0.0003/second)

Whatever was displaying text around its core played another longer message, as it tended to do whenever the dungeon was doing something that felt particularly naughty. This time it could feel the steady trickle of mana falling into its core, replacing the step by step increase of before. Nevertheless, the coin had still shattered. It could keep decreasing the amount of mana, but the dungeon was starting to learn that operating outside the framework of its instincts provided the best growth. It had already improved its control over mana, so perhaps rather than decreasing mana it should increase the amount of silver.

Waiting for its mana to regenerate fully, the dungeon spent it all on silver coins, spawning them all on top of one another. The result was a shapeless blob of silver sitting on top of the pedestal.

New loot unlocked: Silver nugget

Another success. Waiting again for its mana to regenerate, the dungeon started to trickle it into the silver nugget. 1 point, 5 points, all 10 points, to no effect. The silver simply ate it all. This was strange. The dungeon wasn't great at maths, but it knew it had put in more mana than was needed to shatter the silver coins the nugget had been made from. Maybe there was a difference between a 'coin' and a 'nugget'. Or maybe there was a difference between trickling in mana and dumping it in all at once. In fact both were true, but watching the flows of mana around the silver and realising the mana trickling in was leaking back out, the dungeon decided it must be the second. This time it dumped all 10 points into the nugget at once. The nugget glowed for a time then settled back down, but it was obviously changed from the experience.

New loot unlocked: Mythril nugget
Incongruity detected: Mythril materials not available for purchase until dungeon has fifty floors. Resolving conflict... Lowering purchase requirements impossible due to creation of further conflicts. Locking feature impossible due to feature already existing in dungeon.
Error: Unable to resolve conflict. Designating dungeon as [Errant].

Another long message! That... should be good, but for some reason the dungeon had a bad feeling about this one. It felt that it had been a little too naughty, and had been scolded. But no matter, the experiment had been successful and the dungeon now had a new type of loot available. It was unable to summon it due to the mana cost being greater than the dungeons maximum mana, but it had needed to wait for mana to regenerate several times during its creation, so that seemed logical. Perhaps rather than risking its core, the dungeon could summon higher level monsters bit by bit in a similar way? The dungeon remembered its dog and leech experiment and decided that was unlikely to work for its more fleshy types of monsters, but what about slimes? It ordered a few slimes to link up, and found that it could indeed produce a big slime that way. Then what about linking together a few of its largest slimes?

New monster unlocked: King slime
Incongruity detected: Monster exists on a floor that is below its minimum level. Resolving conflict... Cannot move monster to lower floor due to no further floors being available. Erasing monster...

The dungeon felt the mana in the king slime start to burst apart. No! Invaders killing its monsters was the natural order of things, but there was no way it would let its newly created monster just fall to bits on its own. The dungeon clamped down on the mana in the king slime with all the strength its newfound control over mana afforded it.

Error: Resistance detected. Unable to erase monster. Unable to resolve conflict. Dungeon already designated as [Errant]. No further action taken.

Phew, its new slime seemed to have stabilised. Nevertheless, the behaviour of the new slime was no different to its first slimes. It seemed that slimes just weren't going to be intelligent creatures no matter how large they got. It seemed that to get further, the dungeon was going to need to risk its core.

With full mana, the dungeon took a fraction of a point of mana, and tried to force it into its core. Nothing happened. Good, better to use too little than too much. It tried again with a little more, then more again, slowly working its way to a full point. Then two, three, five. Still nothing. All ten. Still nothing. The dungeon pondered for a while. Of course, it was being silly! The mana it was trying to force into its core came from the core in the first place. It was taking a point out before putting it back in again. Obviously that wasn't going to achieve anything!

The dungeon watched the mana flows around its core. It spent a point, and watched the regeneration. What was it that stopped the regeneration when it was full? What was it that limited the rate of regeneration? What caused the regeneration in the first place, when it wasn't consciously trying to move mana around? The dungeon spent some mana again, but this time instead of watching the natural regeneration it tried to reinforce it, drawing in more of the ambient mana and moving it towards the core. It found that it could indeed boost the regeneration speed in this way. Once the core was full, it seemed to simply cease naturally attracting mana, but that had no effect on the dungeons conscious efforts to draw in mana. It continued, and felt its available mana increase. It felt... uncomfortable. As it continued, it felt as if it were starting to burn, then suddenly there was a crack, massive pain and a rapid decrease in the dungeons mana supply. The dungeon shifted its attention and instead of pulling in more mana it desperately worked to hold in the flood of mana spewing from the crack in its core. The contained mana oozed around the crack, congealing, and actually crystallizing around the crack. Eventually the leak stopped and the dungeon inspected its core. No longer the perfect round crystal it had been, but misshapen, with new blue crystals jutting out at odd angles where the crack had been.

Dungeon core upgraded. Maximum mana capacity increases to 20. Mana regeneration increases to 0.0006/second.
Warning: Dungeon core unstable.

The dungeon felt itchy. It knew the attempt hadn't even been close to perfect, but it had survived. With the current state of its core, the dungeon felt that manipulating it further would be a terminally foolish plan, but at least it had more mana to play with now. Time for another new floor.