Quinn climbed down to the basement (after which I deleted the ladder), and I directed her to go back to the entrance to my old core room. There I built a hopefully comfortable chair for her so she could sit and watch my work. Quinn had noted–not as a direct complaint, but the implication was there–that my working speed was pretty slow. It wasn’t something I had really paid attention to before, but I guess it was true. Making new things took all of my focus, which meant I didn’t record how much time it took for me to work. Regardless, as long as Quinn was engrossed by watching the changes, then it was fine.

  I planned the new monster template for the basement to actually match the theme of the room. Not only that, but I was curious if I could incorporate trap mechanics into the monster itself. Specifically, whether I could hide the monster as a non-monster. The classic monster for this was, of course, the mimic. The mimic hides as an inanimate object, indistinguishable, silent, patient. When the party gets close seeking treasure after a hard fought battle, the mimic suddenly pounces! The mimic had become such a tropetastic monster that it was a bit of a running joke on the internet that anything could be made a mimic–the treasure chest, mimic! The furniture–mimics! The totally normal person standing in the room–mimic! The mimic–another mimic! Ma’am, It was mimics all the way down.

  Spiraling thoughts aside, my concept was to procedurally hide the monster among the mundane. In this way, even if the delvers know of the monster that lurks below, they won’t be able to preemptively attack it because its spawning location will be different each instance.

  Of course, first it was necessary to set the scene. The features of the room needed to be part of the theme and how the monster worked. So, I removed the large stone spikes in the middle of the room that sat below the shaft. I had originally put in the spikes because I was worried that, in a world of magic and the system, people could easily survive a fall from the bridge. Valid as that was, stone was not thematic.

  I didn’t have just any undead, I had undead that were nought but bones. That and what was likely the original purpose of the basement. I wanted to lean hard into these. So, the material of choice was to be bone.

  I remade the spikes but out of bone this time, and with more attention to the details of their arrangement. The strongest bone in the human body, I somehow remembered, was the femur. I built the first new spike as a pyramid of femurs smashed and fused together, many angled upwards with their ends shaved down to sharp points. It was shorter than the old spike, so it less resembled a deadly stalagmite as a boney bush of pain. I made three more of the same, all of them in the bottom of the bowl of the floor, under the shaft. I made sure to leave sufficient space between each of them so that someone could walk–gingerly–between and around them.

  Next, after leveling the outer floor to the same as the inner floor, I began adding bones. Lots of bones. The process was repetitive such that I had a thought on how Quinn could participate. “Hey Quinn, would you like to make some sculptures or out of bones?” She hesitated about it, worried she wouldn’t be good at it, but agreed to at least give it a try.

  So that was how we worked. Quinn picked an area and collected and stacked bones into precarious arrangements and towers. Meanwhile, I focused on making more bones in the other parts of the room. I created bones of all types, littered across the outer floor. I made human bones and even started adding in goblin and dwarf bones too. Hundreds of bones. Thousands of bones. I made sure, however, to leave some small spots and a few paths around the room clean.

  When I felt there were sufficiently large piles of bones around the room, I stopped and took stock of the room. It was filled with all sorts of bones, most sitting in large piles. All of the bones looked new (insofar as a bone could look new). I just didn’t have the time to modify the bones as I made them to create variated aging or to create partial pieces of bone. Quinn was busy engineering free standing bone structures out of bone, occasionally suffering a catastrophic collapse and having to restart. But she seemed happy to have something to do, so I let her continue, even if what she was doing was not essential to the room’s function.

  I went to work developing the new monster for the room. It was going to be my toughest monster yet, functional service as the final “boss” monster of the first floor. The concept for the monster was simple, a tried-and-true commonly used monster in computer and tabletop role playing games: a bone golem.

  The bone golem’s design was going to be significantly more complicated than anything I had made before, and would require using my skills together, including Sigilmancy, Trap Creation, and Landscaping. I placed several status screens in front of me, each holding a part of the total design relating to each aspect of the monster’s design. This way, I hoped to avoid stretching my will and intent too thin across too many ideas. Now, I only needed it for the final step, actualization.

  The main feature of the monster would be its body; or rather, its lack of one, at least a fixed one. Instead, the monster’s body would be dynamic--changeable, malleable, amorphous. The bones in the basement I had piled up would serve as its body, but not all of them. Only some of them, and which one those would be would change when the instance was generated. As an added benefit, this would also mean that the bone golem’s starting location within the basement would be different.

  Through my experience developing the modular keys and locks system previously, I had learned a little bit about making semi-random systems. So, I started to develop a similar mechanism for the bone golem. But that was when–having vaguely anticipated this problem–I ran into my first roadblock: what bones could and would constitute the bone golem. There were thousands of bones in the room; allocating bones completely randomly would result in a monster scattered across the room, which was completely infeasible. Hence the conspicuous bone piles. I could use the bone piles as discrete sets of bones, having a starting set rather than the individual pieces.

  I tried setting up a base of piles of bones available for monster instancing, with the system picking one randomly when the Holding instanced. However, I received a feeling of negative feedback from the system. Trying similar methods–even one where I only had two piles as options–all yielded failure. Even when I tried to interface Trap Creation into the process, the system still suggested it wouldn’t work.

  The narrative has been taken without permission. Report any sightings.

  I realized the problem. A Holding made the instance based on the template of the actual dungeon, including the monsters. But my bone golem idea wasn’t a monster created in the here and now and then instanced, it was a monster created upon instancing. In other words, I was asking the Holding to create an instance of a monster based purely on a design template of the monster, not the monster itself. The system did not like this.

  So I pivoted to another way to skin this cat.

  This time, I simply designated all the bones in the outer area (excepting the bone bushes in the center) as the starting material for the golem. This time, when the holding instanced, I wanted the system to remove all portions of the golem’s body except the randomly chosen section.

  Unfortunately, the system again said no; although, it wasn’t as strong a denial, which meant I was on the right track.

  Third attempt. I started the same as before, designated all the same bones as part of the golem’s initial body. This time, I left it so that all the bones would still be part of the golem. However, I added a layer over it such that bones would be categorized into two parts: active bones and dormant bones. Active bones were bones the Golem could form into a cohesive body structure and utilize for attacks, defense, and skills. Dormant bones were all the rest that the golem could not use. Finally, as one of its abilities, the golem could convert the latter into the former in order to replenish its lost bone mass.

  The system largely accepted this method as possible, although there were many fiddly bits that I had to adjust. For example, I had to clarify what constituted a ‘bone,’ particularly in terms of whether it could be used or not. I made it so that only fully intact bones–not broken bones, bone pieces, or dust from bones–counted as bones.

  This led to giving clearer definition to the golem’s health pool. A health pool equal one for one to all of the composite bones would be absurd. Forcing a golem with only a few bones to hobble around, and the delvers to chase the monster to the very last bone to claim victory, was inefficient. I sensed that the system would reject this as-as, so I set out to establish some clearer parameters.

  So, I separated the golem’s “health” from its bone mass (both active and dormant), basically health points that mimicked the common system used in role playing games. When the golem lost active bones due to damage, it would replenish its body with nearby dormant bones, but doing so would not replenish its health pool. To add a bit more spice, I also added conditions for some bones to be special–both those that were defensive, harder to destroy and taking less damage to the health pool, and a couple bones that acted as weak points for critical hits. Conversely, destroying dormant bones would reduce its replenishment reserves, but not affect its health pool.

  As I continued to develop the details, the system was giving its quiet, subtle assent. The monster was tough, but fair.

  The final matter to decide was what level I wanted. Definitely, I wanted it to be the highest level monster for the floor. Not only would it serve as a boss, but it was going to fight by itself, which suggested a high level to compensate. Nevertheless, I didn’t want to spend all my mana, as there were things to still fix, not to mention the second floor was still very incomplete.

  Then again, Mr. Crazypants needed to die.

  With all of the design pieces ready, I poured mana into it, more and more. 50. 60. 70. Finally, I stopped at 80, level 8.

  Nothing in the room overtly changed when I was done. Quinn had been continuously playing and stacking bones, having made over a dozen. A lot of time had passed and while she had been occupied, I could tell that she was a little bored of it. I pinged her on the tablet, “Quinn! The monster is done!”

  Quinn read the message and her head whipped around, frantically searching for the new monster in the room. “What? Where?” she exclaimed when she didn’t see anything different.

  “Around you! All around you!” I provided a summary of the new Bone Golem’s status:

  [Amorphous Golemnity] was the heart–and bones–of the golem. The golem could change its own shape, which included moving various pieces around its central body, such as defense bones, weak points, and its attacking arms. The upgrade allowed it to maintain two attacking appendages at the same time.

  [Heap] was the trap skill. The skill made it so all the bones of the golem appeared as if dormant when instanced. In other words, someone looking at the golem, from above, for example, would only see bones and not an identifiable monster. The golem only turned active and formed its main body when someone triggered the trap–in this case, stepping into the donut-shaped area of the basement around the shaft.

  [Absorb Bone] was the challenge enhancer. The golem, when the skill activated, would absorb some nearby dormant bones into its body, bringing it back whole. The upgrade also allowed it to gain some health points back too. A balance to this was that “weak point” bones that were destroyed would also reduce the golem’s maximum health too, so the golem couldn’t just “top up” back to its original health over and over again.

  Finally, [Bonesplosion] was the ultimate. When the golem triggered [Bonesplosion], part of its body would explode, sending bone shrapnel in every direction. It synergized nicely with [Absorb Bone] as the golem could use it to replenish its reserves after using the AOE attack.

  “What do you think?” I felt good about the bone golem. Its body was exceptionally tanky, combining a form that had no easy weaknesses and the annoying ability to regenerate. It could also mix up its attacks and even had a surprise factor.

  Quinn alternated between my descriptions and peeking around the room to try to find the golem. “I think it’s great!” she said enthusiastically. “It’ll be a real nasty surprise for those goblins!”

  Quinn crouched next to one of her sculptures, the latest one. She then poked some of the bones, whether that would tell her where the monster was.

  “But Gorge,” Quinn frowned. “I spent a lot of time making these sculptures. Like, a lot of time.”

  I zoomed my attention straight to the clock.

  Oh.

  This wasn’t good.