Dungeon Designs – The Living Vault Addendum

In the end, I really did want to go over some of the specific themes I had in mind while detailing The Living Dungeon, and so here we are! I wasn’t originally intending to do these so soon after the original articles, but I’m not in charge of the content pipeline. Come to think of it, I also don’t know who is in charge of it, or if it exists at all.

As a warning, this will include some body horror – the original iteration of this idea was intended to be an aberration-themed nightmare after all. But not all of them are. Just to make doubly sure, all the body horror options will actually have the words “body horror” in their section headers.

With all that out of the way, let’s dive right in!


A Living Dungeon Refresher

First, let’s run through the Living Dungeon concept real quick. The version found in my previous articles was called The Living Vault (part 1, part 2), and that’s what I’ll continue to refer to it as from here on.

The Living Vault was intended to be an adventure where the dungeon itself could play a role in combat and exploration. To that end, we statted the Vault out as a creature in order to define what it could and could not do. We then also defined what aspects of the design would affect exploration as well.

What we finished with was a creature who served primarily as an obstacle in combat – it could physically restrict players’ movement with area denial and grapples, but its actual attacks weren’t very strong. The Vault also had a caveat – deal enough damage to it in a single round and it would go dormant for a set time. Thus PCs could choose to either focus their damage on the other, more lethal enemies (at the cost of leaving the Vault’s ability to hinder them intact), or they could deal damage to the Vault to stop it from acting.

In exploration, the Vault was primarily able to preemptively set traps where it thought the party might go, and the creature’s bodily processes were the inspiration for why different rooms existed. Players could, through exploration, cripple the Vault by sabotaging areas vital to its various abilities.


All in all, I’m very happy with the Living Vault. I think it’s a fun concept. And it’s versatile. While writing the other articles, I came up with multiple other interpretations besides just the aberration-based idea I had begun with.

So let’s look at a few of those, shall we?


The Living Bounty

One of the alternate flavors I proposed originally was a “druid-blessed glen” with a more nature-based aesthetic. And I actually quite like that option. It’s new and interesting and has significant variations from the standard “inside the monstrosity” dungeon present in many other games.

The most important question here is whether the Vault is one single organism with other creatures living in it (such as a giant tree’s branches or roots) or if the Vault should be a collective (similar to a fungal growth). Both are valid options, though I prefer the former personally.

Mushrooms are just a bit overdone, and I can’t think of anything unique enough to make the Vault really stand out. A Fungal Vault would focus on mushroom zombies and a lot of poison. And there’s nothing wrong with that, it just isn’t super interesting to me.

The single organism idea, though, is much more intriguing.


Personally, I’d go for a forest glade with a single, gigantic tree as the dominating and controlling party – The Vault Tree. Going inside the tree isn’t the goal – instead, you have chambers made by walls of trees, where the roots of the Vault Tree run all along the ground and the other trees, allowing it to act.

This Vault would likely not be able to reconfigure rooms (a potential ability for some vaults), but that isn’t necessarily a problem. A static map is much easier for a variety of reasons. I’d also likely incorporate something to reflect the weakness of wood to fire. I don’t want the players to just burn the entire forest to the ground to avoid the dungeon, but it feels like any “plant monster” should have at least some weakness to fire.

There’s also the matter of lighting. Since the glade could be technically “outside” (though I’d hesitate to make the actual area anything more than dim light through the tree canopy), that means the day-night cycle could affect the dungeon as well. Perhaps certain monsters normally only found in specific chambers could instead patrol the Vault during the night – the wolves can guard locations, but they don’t appear as random encounters except under the moonlight.


When it comes to monsters, however, we have an issue. Even assuming magic, it feels a bit cheap to have an endless number of random animals to throw at the party. To solve this, I’d split the Vault’s “roster” into two groups – beasts and swarms.

The beasts would come from various dens throughout the Vault, each one likely connected to a specific species of animal. Thus you’d have the Wolf Den, the Bear Den, and so on. The idea here would be that the druidic magic of the Vault allows a much higher population of animals to live in a smaller space, and that the Vault’s will prevents infighting. But these animals would not be infinite. The Vault can’t create animals from nothing (if they’ve all been killed), nor can it massively quicken the aging of young animals.

For the more numerous encounters, I’d use insect swarms. Mosquitos, spiders, beetles, etc. These creatures would still be limited in number and take time to replenish, but the process would be much faster. If you wipe out all of the swarms in the dungeon, the Vault can still have a few swarms grown by the time your party finishes a long rest.


Alongside the beast dens, other important rooms in the Vault would be based on environmental needs. A spring that provides water to all the plants and animals could serve as an “Energy Source” – if players could either poison or block off the flow from this spring, they could drastically reduce the strength of monsters within the dungeon.

Then there’s food sources. You could have various chambers with different types of food for all the beasts in the dungeon. One room could have numerous berry bushes (picking them all takes time, but now you have food and the monsters don’t), another could have bright sunlight for plant monsters (if it comes in through a hole in the canopy, maybe you could block it?), and finally a prison for holding meat for carnivores. Freeing the prisoners, which could include both humanoids and beasts, could both weaken the Vault’s carnivorous inhabitants and gain the party help or treasure from the grateful ex-prisoners.

I’d also like to incorporate the druids themselves. There could be druidic campsites where wild shaped druid guardians still remain, or magical rune stones that act to empower the Vault. You’d want to avoid having too many of these, of course – it kills the atmosphere of “natural savagery” if the party keeps stumbling across cooking pots on old burned-out firepits.


The final issue with this concept is the fact that some characters might balk at ruining a pristine natural environment. Even if it’s trying to kill you, what if the Vault Tree was just trying to live? Poisoning the waterhole, ripping up berry bushes… it all feels a little mean.

There are two ways to fix this. The easiest is to simply say that the grove is “unnatural” and twisted by evil magic. That way, destroying these foul chambers is doing the natural world a favor. The other option is to allow players a non-destructive way to hinder these processes.

Beast dens could be used to remove monsters by setting some sort of lure to bring the creatures in before sealing the exit to the den, trapping them inside. The spring could be dammed up, rather than poisoned, and in a way that’s easily fixable once the party is ready.


One critique of this concept is that it lacks a lot of the “identity” that can be found in other versions. The Vault Tree doesn’t have much of a personality, and even if it can actively make decisions it still isn’t a true, thinking entity. There are fixes to this, of course, I just don’t think there are as many narrative options here as there are with some of the other concepts.

But at the same time, it really makes sense for a “natural savagery” Vault to lack a “humanoid-adjacent” personality. A tree wants to grow. It was made to protect something, or an area, and it desires to grow. That’s it. Perhaps it has enough sentience to understand that too much growth is bad and that culling is occasionally necessary. Or maybe it doesn’t and that’s why the party is here in the first place!


The Intelligent Manufactory

For more magitech or even outright sci-fi settings, a sentient machine factory could make for an excellent Living Vault. A central intelligence controls a variety of mechanical apparatus to manipulate the environment. If you get Portal 2 levels of conveyor belts and cranes involved, you can even let this Vault rearrange rooms.

It would be remiss of me to not mention one of the big inspirations for this concept, which is “The Clockwork Mansion” from Dishonored 2. That game actually has two exceedingly brilliant examples of level design (the other being the seventh mission, “A Crack in the Slab”). Regardless, the Clockwork Mansion from Dishonored 2 is a perfect example of this concept in action.

For those who haven’t played the game, the Clockwork Mansion is a typical “high-class mansion” design with the added wrinkle of having every room be highly customizable using various levers and devices. Dishonored 2, as a stealth game, primarily uses this for the existence of the “inner workings” – it’s actually possible to find your way into the walls at the start of the mansion and then complete your mission without the target ever realizing there was someone in the house. You can also use those levers to configure different rooms to whatever you need at the time.

Now, I’m not sure that same level of activity is really possible in D&D, unfortunately. But that doesn’t mean you can’t still use parts of it.


In the case of the Mechanical Vault, you’d essentially have a scenario where the Vault, as overseer, has control over the various rooms and can move and modify them as needed. Giving players that ability as well could be a fun mechanic, with the caveat being that the party can only control certain things from certain places (IE where the lever is). I envision the big room interaction type here being based around levers and physical obstructions.

The Vault could produce mechanical creatures to attack the party, until the PCs reach the Creature Assembly room and lodge a thick bar of metal into the gears, preventing the Vault from replenishing its forces. It could also have other manufacturing plants, making things like poison or acid, which the party could similarly sabotage.

Levers, meanwhile, could provide access to new areas or seal off other areas. Another option would be to imagine the Vault’s designer as having (correctly) worried about the Vault breaking free from control – and so a few choice levers hidden in various corners could completely shut down the Vault’s control over a certain area.


One idea I mentioned in the initial articles was having a mechanical Vault function differently when it comes to the party damaging it to remove it from combat. The “default” idea was for a certain amount of damage in a single round to cause the Vault to be unable to act in that room for a given amount of time. Hit it enough in a single round, and the Vault is basically out of that specific fight.

For a mechanical vault, though, I suggested this limit could be higher – but without a requirement for all the damage to be done during a single round. And then after the requisite amount of damage was done, the Vault’s control over that room would be permanently disabled.

I ultimately don’t like this idea as much as when I first wrote it. I worry that it would encourage players to manipulate combats to happen only in disabled areas whenever possible. And while that is a tactical decision, it’s one that doesn’t seem like it would be much fun to actually play.

And if you did want to do something a bit unique with the Mechanical Vault’s combat presence, you could always just make things a bit more modular. Rather than needing to deal a certain amount of damage in a single round to shut the Vault down for a number of rounds after that, you could instead have levers that take an action to pull and disable the Vault for only a single round.  It changes the value calculations quite a bit.


One thing that you would need to keep in mind, however, is the tendency for an automated facility to seem less like a singular entity and more like a set of if-then statements. A Mechanical Vault needs personality more than many of the others in order to ensure players view it as a “living entity” and not just an algorithm of pain.

Avoiding the typical “perfectly logical machine” theme would be a good way to do this. Give the Vault irrational thoughts, fears, and desires. And more importantly, let the Vault make mistakes. A perfectly played Vault that’s efficient in every way is both a boring enemy and an overpowered one.

Perhaps the Vault is lonely. Even if the party is obviously here to say, maybe it could prioritize preventing them from leaving even when that means it doesn’t do anything to hinder their progress. You could also allow the Vault to “break itself” to harm the party – a mechanical arm clearly not intended to bend backwards could literally snap itself to get an extra hit off on a PC. If you have a low level or inexperienced party, you could even have the Vault convinced that the PCs are quality testers from its creator – it never does lethal blows, since it figures you’re just testers on the payroll.


A more esoteric option would be to steal the personality cores idea from Portal. Have different important rooms also include logic cores or personality drives, which the players can then freely change or disable. In response, the Vault’s decision making could change. Remove the “Paranoia Drive” and the Vault suddenly stops actively searching for the party. The “Martial Drive” could be corrupted into an “Aggression Drive” that has differing groups of random encounters actively fight each other and not just the players. Progress comes from adversity, after all.

This should never get to the point that the party can simply “turn off” the Vault’s desire to engage in combat or prevent the party’s forward progress. Having drives that can only be corrupted instead of destroyed helps with this. An “Aggression Drive” will fight a lot harder, but there’ll be less monsters over all because they keep dismantling each other. As long as the party can change its style of action, they can tailor it to their own strategies.

This is definitely the most complicated of the Vault options. A good Mechanical Vault really needs a lot of pre-planning and a deep understanding of how the party might react.


The Body Horror Vault

It’s time to get gross. This was the original incarnation of the Living Vault concept before I decided to write a more general article on it. The first idea was tentatively named “The Chamber of Vaulted Flesh” or something along those lines.

There really wasn’t that much more to it. I never got around to fully fleshing out (ha) the idea, but it remained something I felt was too interesting to just abandon entirely. That being said, working on the articles did change my mind about a few things.

So let’s take a quick look at the Flesh Vault.


One rather core change concerns the Vault’s personality. My original concept had no recognizable humanoid consciousness – it was intelligent and conscious, but in an alien, inhuman way. And while this can still be good, I realized fairly quickly that you really need a relatable personality.

There are several ways to establish this. As I mentioned in the original articles, one good option is to include a “Pet Project” room within the Vault. Something that exists solely for the Vault itself, perhaps even being something not originally intended by the Vault’s creator.

For a Flesh Vault, I believe the best “Pet Project” would be an effort to create a new body for the Vault to inhabit. This can interact with the party quite neatly as well, since their day-to-day life (interacting with one another and freely traveling the world) is exactly what a lone, bound-in-place Vault would want.

(This could also serve as a handy way to mitigate a potential total party wipe, just as a note. If the Vault wants to explore the world like the party, having its monsters knock them out and tie them up somewhere for research would be an excellent way to turn a lethal encounter into a fun narrative twist. Or, if you want to go hardcore, just have it so the Vault was secretly cloning them all – the first to die is the first to have their clone perfected. Always check with the PC in question to make sure that kind of plot twist is okay – if it isn’t, just have it be a “healing vat” instead of a cloning one.)


In addition to personality, I also realized my initial thinking on how to “bring the dungeon to life” was flawed. I had primarily been considering the “rooms as organs” concept – a Vault with poison has a poison gland room, a Vault with fleshbeasts has a spawning pit, etc. – when in reality the best options are all in the immediate actions the Vault can take.

Think about it – any dungeon can have critical rooms which serve a purpose to the dungeon as a whole. There are always barracks or sleeping rooms where monsters stay when not roaming the halls. What makes the Living Vault different is its ability to act during combat as a monster.

So, what would make a Flesh Vault feel… well, fleshy?


Well, the visual descriptions you give are an obvious start. Having specifically “unique” and “disgusting” actions can enhance the atmosphere of dread. I said before that the Vault should be able to create areas of difficult terrain, but you can’t just leave it there.

When the Flesh Vault creates difficult terrain, it chooses one of two options: Saliva or Bone. Difficult terrain made from saliva is slippery, and requires a Dexterity save to move across without slipping – a creature who slips then has disadvantage against the Vault’s grapple for X number of rounds. It’s sticky. And gross.

Bone, meanwhile, would get to have the disconcerting visual of bone shards ripping their way out of the ground (which, in this case, is recognizably flesh). This could even add a health cost to the action, with the Vault taking X damage any time it uses this difficult terrain ability. In return, the difficult terrain might remain permanently instead of fading away after time. Those bones are just there now.

More so than any other type of Vault, the Flesh Vault needs its abilities described in detail and thought out carefully. Perhaps its grapple becomes harder to escape every turn as the wall slowly swallows the victim alive. Its tendril attacks could be more than just a simple 1d6 damage, and instead have a minor poison effect to reflect the fact that the tendrils are tipped with stingers. Or they could be tongues coming from gaping maws on the walls.

Just make really sure that your players are okay with this. It can get a little freaky.


In addition to this marriage of description and mechanical effect, it would also be good to make the monsters within a Flesh Vault interact with the Vault on a much deeper level than with other concepts. Players should be able to see that these creatures aren’t just inhabitants of the Vault, they’re a part of it.

This is partially handled by the way in which the monsters within the Vault serve as its “eyes” when targeting the players. But you can also take it further. Perhaps the Vault gains temporary hit points or an extra action whenever a friendly creature dies – it reabsorbs the corpse to fuel itself, or reshapes the fallen bones and muscle into an extra tendril to attack with. Or maybe the Vault can use an action to augment a creature, or forcibly meld two different creatures into a single horrid abomination.

You could even connect monsters to the Vault’s various in-combat abilities. By spending an action and losing access to its “Bone Difficult Terrain” ability, the Vault can create a “Bone Growth” creature in the room. While alive, the Vault cannot use its difficult terrain action. The Vault’s attacks are normally coated in a mild poison, but it can sacrifice that to summon a much more lethally poisonous creature to attack instead. I’d limit things so the Vault can’t just delegate all of its abilities this way (I’m thinking only one can be used at a time), but this would help to really solidify the creatures as “part of the bigger whole” and not just roaming inhabitants.

And, of course, these monsters should emerge from the walls in the most grotesque way possible. Again, make sure your group is alright with this. Because it’s a distinct possibility that they won’t be.

Hell, I’m starting to gross myself out.


Finally, in addition to all the above, keep in mind that this Vault in particular has a wealth of other examples to steal from. Ocarina of Time has Lord Jabu Jabu, of course, and there are around a dozen knock-offs of that dungeon in various JRPGs, RPGs, and non-RPGs. Or if you’re running Ghosts of Saltmarsh and/or a nautical campaign, you can take a page out of Pinocchio and use a giant whale.

Adventurers get eaten alive a disturbingly large number of times.


A Farewell to Vaults

With that, I think I’ve gotten out all of my more specific ideas surrounding the Living Vault idea. There are more alternate interpretations out there of course, but these are the three I wanted to cover the most. Other examples include a trippy Inception-style dream-dungeon, a haunted house possessed by an overwhelmingly powerful poltergeist, a githzerai monastery controlled by a mad monk, nearly anything involving the fey, and many more.

Now, am I going to do anything with this concept any time in the near future? Maybe! It’s always a possibility. But for now I think I’m done. On to other things, which should be coming soon.


Oh, and like always – let me know what you think!


Leave a Reply