Dungeon Designs – The Living Vault, part 1

Recently I’ve been brainstorming a new concept for a dungeon. Not for any particular reason – I just think it sounds cool. But I wanted to discuss it here to show off my method for making dungeons, which should be informative, entertaining, and/or horrifying.

That last option is primarily for people who have played in my games, who will presumably come away from this thinking “what the hell man, that’s how you came up with this nonsense?” To which I say… sometimes.

Other times I’m just sort of winging it.


What’s a “Living Vault” Anyway?

To put it simply, the Living Vault is a concept in which the main encounter of the dungeon (the final boss, essentially) is the dungeon itself. The dungeon is statted out like a creature and plays an interactive role in all of the rooms the players pass through.

At first this sounds like it would be best tagged with “body horror” and/or “stole this from Ocarina of Time,” but those are only two options. A Living Vault could be anything, really. While it could easily be a horrific aberration, it could also be a semi-sentient enchanted building, the insides of a giant monster, or even an awakened grove charged to guard something by the druids.

In order to make this discussion as “theme-agnostic” as possible, I’ll be referring to things rather plainly. Just always remember that the “Dungeon Core” could be a giant beating heart or a vast magical/mechanical clockwork device. The vault’s “arms” could be tendrils of flesh, vines, or pure arcane force. It’s all up to you… and your players’ stomach for body horror.


Another question of note is “what does this add to the dungeon delving experience?” I’m not a fan of meaningless gimmicks, but all the same it can sometimes be fun to have a “themed” adventure. However the primary use of this concept is in imagining a different sort of dungeon that can react to the players as they move forward. It’s a living space, rather literally in fact.

This is as opposed to many classic dungeons, which are mostly set pieces. The best dungeons usually have suggestions on how to change things to adapt to the party’s actions (such as noting which rooms are havens for retreating creatures, or sources of reinforcement) but rarely do these recommendations cover the whole dungeon in its entirety. There are always rooms which don’t have any particular use – this is fine, since most structures would logically have such rooms, but it does mean that a large part of the dungeon is inactive.

So that’s what this does! By reimagining the dungeon as a creature and giving it options similar to other creatures, we allow even rooms with no “set purpose” to interact with characters more deeply. But now we’re getting more into the mechanics, so we should move on to…


Biology of a Dungeon

Okay, so how does this dungeon work?

My ideal iteration is a typical dungeon crawl complete with traps, puzzles, and monster encounters, but with the added wrinkle that the dungeon itself is an entity involved both in combat and outside of it.

But this brings up an issue – if the dungeon is intended to be able to act in combat, what’s to stop it from just constantly harassing players as they try to solve puzzles or evade traps? No one wants to do a puzzle while getting slapped every six seconds by a tendril or vine or something.


My preferred fix for this is to make the dungeon-creature (which I will be referring to as “the Vault” from here on out) completely blind. No tremorsense, no blindsight, nothing. Instead, it can simply see anything that the resident monsters see. If there are no monsters with a view of the party, then the Vault can’t target them with attacks or combat actions. But, based on where it last saw them, it can guess at their path through the halls and try to set traps or complicate obstacles in their way.

This sets up an interesting combat/noncombat split. During combat the Vault acts as an extra creature. It can actively and immediately interact with the party. But while out of combat (IE with no monsters nearby) the Vault can only actively plan ahead – it can change traps, move halls, and send creatures to try to intercept the party. But it doesn’t know for a fact where they are, so all of these actions, while still interactive, are delayed.


With that basic premise set, let’s move on to actually statting out this creature. From there, we’ll have a good baseline to begin discussing specific concepts, mechanics, and activities later.

So to begin, let’s look at the Vault’s defenses.


Defenses and Deterrents

My proposed defensive traits would include a high HP, a low AC, a damage threshold, and resistances to both damage types and certain forms of attack. These traits should combine to make the Vault into an active force that simultaneously can’t be defeated at the entrance, thus requiring players to brave the depths.

Now, let’s look at those traits one-by-one, shall we?


The high HP is to make sure the Vault can survive long enough for the players to get through the dungeon. It would be easy to simply say that you can only attack the Vault in specific weak-points, but that makes the entire thing feel less like one interconnected living entity… and more like a simple “house of tricks” with a hidden mastermind in an inner room. Being able to hit this thing at any time and deal damage to it (even if it isn’t a lot of damage) helps reinforce that this entire structure is alive – it is always a creature in opposition to you.

The low AC, meanwhile, reflects the fact that the dungeon is literally as big as the broad side of a barn. It shouldn’t be that difficult to hit. This also affects how it should treat Dexterity saves, but we’ll come back to that in a moment. The key point here is that the Vault is big – how hard is it to just smack a wall, really? Not very.


Now on to the damage threshold. For those who haven’t read Secrets of Saltmarsh (which I believe remains the only place where these rules are printed, unless they also included it in Spelljammer), a damage threshold works to prevent small, repeated instances of damage from destroying a massive entity. For sailing ships, this is what keeps the party monk from repeatedly punching a wall in the brig until the entire ship disintegrates.

Basically, a damage threshold allows the creature to ignore any damage below a certain number. With a damage threshold of 5, for instance, the Vault would automatically take 0 damage from any attack that deals 5 or less damage. This does not mean the Vault reduces all damage by that number, so if an attack deals 10 damage, the Vault will still take the full 10 damage.

The number chosen for the threshold should reflect the intended party level. It needs to be something high enough to ensure that “incidental” damage (splash damage from area of effect attacks, Extra Attack hits made because there aren’t any other targets nearby, etc.) don’t end up killing the Vault. But it also needs to be low enough that concerted effort can still damage the Vault, which is a concept we’ll come back to later.


Lastly, the Vault should also have certain resistances. These can vary somewhat depending on what theme you’re going with (a non-organic Vault might be immune to psychic damage, for example), but largely I think the ideal is to give the Vault resistance to all damage types, and immunity to damage from nonmagical weapons.

This is to represent the differences in scale. A fire bolt to a normal-sized creature hurts, but with a being of this size it’s more comparable to a single cinder. We don’t want the players to be completely ineffectual against it, but we should reflect that scale somehow.

Another benefit of this is to allow us to make the Vault automatically fail all Dexterity saves while automatically succeeding on all Strength saves. After all, a wall can’t just dodge out of the way if the wizard uses fireball. Simultaneously, it doesn’t matter how much Strength the barbarian has – she isn’t going to be able to push the entire room around. Damage resistances, meanwhile, ensure that the damage from fireball or any Strength-based assaults don’t kill the Vault too early.

Of course the Vault should also be immune to certain conditions too. I feel like these are fairly self-explanatory, so I won’t spend too much time going over them. You need immunity to Blinded and Deafened due to the Vault’s lack of primary senses. It’s immune to Grappled and Prone due to its size, and should likely be immune to Restrained as well (though DM interpretation could affect that). And, finally, it should also be immune to Invisibility and Petrified because both of those would be weird.


Now, you might be looking at this and thinking “well how do the players beat it?” The answer is one final trait: Vitals.

Vitals. Certain rooms of this creature play an integral role in its operation and are more easily damaged. These areas do not benefit from the creature's damage threshold or damage resistances.

This is how players destroy the dungeon. I personally like the idea of there being multiple of these Vital areas (add “if a vital area takes X or more damage, it is destroyed – damage to it no longer lowers the creature’s hit points and the vital area no longer provides control of nearby regions“). But you could just as easily have a single vital area, the heart, where players will battle with the dungeon for their final confrontation.


Now let’s move on to the Vault’s offensive capabilities. And just to reassure you – there is a reason I still want players to be able to attack the Vault even when I don’t want them to kill it without engaging the Vital areas. It’s all part of making the dungeon feel more alive.


Offensive Options

What should the dungeon be capable of doing during combat? We want it to be an active participant, after all, so we need it to be able to take actions. Let’s give it an attack, a grapple, and an area denial option. The overall goal is to make the Vault into a weak combatant that nevertheless has immense effect on the battlefield.


The attack is the simplest – it’s a slam attack made with the Vault’s tendrils, vines, arms, or other appendages. It should do a noticeable amount of damage, but not too much. Remember, if we give the attack a normal melee range of 5ft, that means it can hit literally anyone in the dungeon so long as they’re within 5 feet of a wall, floor, or ceiling. This can’t be a big-hitter if it’s also functionally unavoidable.

One other caveat I’d add is to make the attack have a relatively low bonus to its attack rolls. The Vault operates on a different scale than the players. In addition to the sheer size difference, I also picture the Vault as not being terribly fast. So its strikes hurt and can reach you anywhere within the dungeon, but they’re also a little slow and not too precise.


The grapple, meanwhile, should be a bit more reliable. The dungeon can only grapple creatures that are close to its walls – trying to grapple to the ceiling adds gravity to the forces that have to be counteracted to keep the victim in place, and grappling from the floor could have players just remove their boots.

Other than that, though, I’d make the grapple very potent. Difficult to dodge (high bonus to attack rolls) and difficult to escape on your own (high DC to break or slip free). Your allies can help you free, of course, and I’d make sure the Vault can’t grapple more than maybe two people at a time (per room). But if you’re caught on your own and the party can’t help you, then your character should be in serious trouble… if not from the Vault itself, then from its lackeys.


An area denial option essentially reflects the Vault’s ability to create its own traps. Outside of combat, this is how you get tripwires, trapdoors, and all the other hazards (which we’ll discuss later), but in combat you would easily be able to see these hazards forming. So, instead, they make it difficult to move in certain ways.

Being able to make difficult terrain anywhere within the dungeon is one simple ability that would fit this idea. I also like the idea of the Vault “closing off” a doorway by placing a web or net across the opening. You can still get through, and you won’t be surprised by the net, but you’ll still have to spend a bit more time getting the netting off before you can move freely again.

Each of these options is, essentially, an “unavoidable” trap which is simultaneously less disastrous to trip. While exploring, hitting a tripwire might set loose a giant boulder that could crush you. In combat, the tripwire doesn’t do much more than send you onto your face… but it’s also directly in the path you need to take to get to your dying companion with a life-saving potion of healing. And, most importantly, the Vault knows this.


And now we come back to our question from the defenses section – why let players attack the Vault at all? The Vault’s defenses mean that attacking it outside of Vital areas is effectively worthless when it comes to killing it, and yet its offensive abilities represent a not insignificant threat to players that they can’t mitigate. Well, here’s the answer: Shock.

Shock. When a specific room takes X or more damage in a single round, the creature loses the ability to use any actions within that area for [one round, two rounds, etc.]. Currently grappled targets can immediately make a roll to escape with advantage, breaking free and falling prone on a success. Other existing manifestations (traps or areas of difficult terrain) remain unchanged, but during this time the creature cannot make new ones or use any other actions.

The goal here is to make each combat revolve around a simple calculation – how much interference from the Vault can the party deal with while still fighting off incoming creatures? After all, every attack against a wall or door is damage not being done to the monsters in the room (who are more likely to actually kill characters compared to the Vault’s own weak attacks). But, at the same time, dealing enough damage to the Vault can completely stop it from acting… thus buying the party time to finish off the other, more dangerous creatures.

Different parties will handle this arithmetic differently. A party which doesn’t rely much on positioning might feel comfortable completely ignoring the Vault during combat – its traps, area denial, and grapples aren’t a big deal. Simultaneously, a party with access to large damage spikes (paladin comes to mind) might decide that it’s best to start combat by smiting the hell out of the wall just to prevent the dungeon from setting up any environmental threats.


This is what I like about this concept. It forces players to make decisions and to remember the dungeon itself when doing so. Environmental effects are always important, but they rarely change all that much during combat. But with the Vault, you can have interesting situations arise.

Perhaps a player goes down somewhere that the Vault can easily isolate from the rest of the party. If the Vault gets to take its turn, it will seal off this downed PC, making it almost certain that they’ll die. But if the party can do enough damage to the Vault before its turn comes around… they have now saved their companion’s life!


You could also expand this to reflect unique weaknesses of the dungeon’s theme. Perhaps a plant-based Vault only uses this trait if it specifically takes fire damage equal to the given amount, while a constructed dungeon could have a higher breakpoint but then permanently lose function if exceeded, creating a safe space for the party.

The central idea, however, is to make the Vault into a noticeable factor in battle… and then give the players the ability to turn that variable off if they want to. Unlike most other scenarios, this is a combat effect completely controlled by the players themselves.


Dungeon Delving

Finally, we have to ask how the Vault will interact during exploration. It isn’t enough to have the dungeon simply interfere as an additional combatant – if we want to get this theming down right, we need to make sure the Vault is always active.

The easiest way to do this is just to say that the Vault creates all of the traps as the party moves. Thus if a party entered completely invisible, there’d be no traps of any kind… because the Vault didn’t see them coming. Yet despite how easy this would be, I don’t really like it as an option. It just feels too passive.


Instead, I’d prefer a situation where the Vault has active control over both the monsters and the traps within the dungeon. Some traps are pre-existing (permanent defenses the Vault keeps active), but others are direct responses to the players’ actions. Some examples might include…

  • A lock which secretes acid onto anything inserted into it, potentially ruining the rogue’s lockpicks. For a less severe option it might instead be adhesive, sticking the lockpicks inside until the door is opened.
  • A pitfall trap being inverted so the doors swing up and painfully hit players in response to a character’s easy access to levitate. Alternatively it could change all existing pitfall traps to instead open a section of ceiling, dropping rocks or acid.
  • A reactivated trap behind the party as they approach what the Vault knows to be a dead-end, though they have yet to discover that. This trap could trigger the same what that it did originally, or it might be tied to a new trigger.

I’d also want to set limits to various things, of course. You don’t want to metagame too hard, since that would feel unfair, but it seems very possible to make a fun, interactive experience. So instead you just set a few simple limits on how the Vault can change things.

If the Vault wants to make a door harder to pick, it should only be able to increase the DC by up to X, no more. If the Vault wants to make a door unable to be lockpicked, then it has to allow for the door to be forced open in a different way – the lock may secrete glue that immobilizes lockpicks, but that requires the locking mechanism to have tactile sensation, meaning holding fire or some other damaging contact to the lock will make it wince and release the door.


The other way I’d involve the Vault in the exploration side of things is to plan rooms out in a logical way for whatever the Vault truly is. If the Vault is an aberration that can secrete poison or acid, I want there to be a section of dungeon that is clearly the poison/acid production organ. If players can disable it, the damage on all future acid/poison traps is halved.

This can apply regardless of theme. An awakened grove might have dens full of beasts, explaining where common enemies come from. A construct Vault could have a manufacturing plant that produces the constructs that stalk its halls – perhaps players could repurpose parts into usable (and potentially magical) gear and equipment.

I’ll leave things there on that front, since it gets a little more into layout than I want to detail right now. Layout is for the next post. The point is that I want players to have reasonable expectations of what lies around the corner which they can then use to modify their plans. Smart, tactical exploration is a favorite of mine.


Next Up: Anatomy of a Dungeon

And there are the Vault’s basic characteristics. Next time I plan on diving deeper into the layout questions and options this dungeon concept has. Part of what we’ll focus on is how to avoid standard video game locales (the stomach, the brain, etc.) while another thing we’ll look at is how to make those standard settings work.

I mean… come on, Lord Jabu Jabu from Ocarina was a fun dungeon. I think. Maybe. It’s been a while since I played it, so I’m not sure. It definitely wasn’t the worst.


I also plan on doing more of these in the future, so look out for those. But in the meantime…

Let me know what you think!

Leave a Reply