When I posted the draft of the Nature of the Qilin for review, someone commented about the range of casting features it had attached. So I thought that would make for a good design notes: a roundup of some of the ways you can write features that interact with spellcasting, which ones I've used (or avoided), and what some of the pros and cons are.

First up: I started to draft something more comprehensive and quickly saw my word count was going to explode. So we're gonna do this as a multi-parter. Part 1 will cover the basics and talk about some feature designs. Part 2 will hone in on full casters and talk about what features work best for them. Part 3 will explore how these features relate to part-casters. Part 4 will specifically focus on gishes and features designed for gishes. And I'll speed this up rather than make you read about it for four whole months; probably it'll take the place of this month's and next month's notes.

Spellcaster Types

First up, let's define some terms. You have to know what sort of spellcaster you're writing before you can pick features that make sense. There are three axes to think about here.

First, spell list. 5E has unique spell lists for every spellcasting class, so wizards get a list, sorcerers get a list, druids get a list, etc. The alternative is common lists, generally grouped by power source (arcane, divine, occult, etc.). PF2E does this, Tales of the Valiant does this, WotC experimented with it in one of the playtest docs for 2024E. If you're pulling in homebrew classes, then you've got those spell lists to think about too. For some subclasses, you also get limitations around spell school selection.

Second, spell progression. Full casters usually get an additional spell slot every level, and a new spell level at every odd level. Half-casters, like paladins and rangers, get half that: literally the same progression occurring at every other level and capping out at 5th-level spells. One-third casters get, well, a third of that, for pretty anemic slot progression that caps out at 4th-level spells. Artificers are a special case, with standard half-caster progression but tweaked to give them spellcasting at 1st level. (In 2024E, all half-casters get this). My necromancer class adds a three-quarters progression, too, which caps out at 8th-level spells. I'll often refer to spell progression as a "spellcasting chassis", because like a car chassis, it's the basic machine you bolt everything else on top of.

Third, spellcasting variety. The big, obvious difference is between pact casters and everybody else; pact casters only get a few spell slots but always cast them at a high level. But you also have known-vs-prepared casters, who differ in how they learn spells, and innate casters who may get very little choice about the spells they learn (KibblesTasty's Warden stands out here, but you can argue for something like a Way of the Four Elements Monk as being in this category). And you can put points-based casters in this bucket too, like KibblesTasty's Psion.

When you survey those three axes, you basically get the "what", "how much", and "how" of a class or subclass's spellcasting.

A Quick Tour of Spellcasting Features

First off, consider that we don't always have to write spellcasting features. Every WotC wizard subclass gets an interaction: it's pretty necessary for wizards, since spellcasting is the ~only thing wizards do. But bard, cleric, and druid subs often don't have an interaction at all. Sorcerers often touch spell slots in some way, but more often than not it's just "spend a spell slot to get more uses out of your cool feature". (Or, more obliquely, get features powered by sorcery points, which you can make out of spell slots). So know that a spellcasting feature isn't mandatory (unless you're writing a wizard sub). In that vein, most of my examples will come from wizard subs, just because that's the first place to look for examples; but all the advice here is relevant to any spellcasting class.

Stronger Spells

Numbers Go Up

Straightforward numeric buffs are the simplest type of feature. Most often it's something like "add your ability modifier to the spell damage" or "one target of the spell gets additional healing equal to X". There's a little nuance in how you generate the number: it's a little more interesting if it's sourced from an ability or stat, but be wary with proficiency, because that's dip-friendly if it comes online at a low level. You can also put longer duration, longer range, or more targets in this bucket; depending on the spell, those can be very good or not terribly relevant. (Necromancy wizard's animate dead target boost is probably the hallmark for a non-damage "numbers go up" buff.)

The advantage of "numbers go up" buffs is they're very clear and very easy for players to reason about. They can go very well with an already-complicated subclass, but can be a little dry on a simpler one. They're also a case that can add quite a bit of numeric power without necessarily adding a ton of flavor, which is rarely ideal.

More Reliable Spells

I've got this as a subsection because "more reliable" is, in D&D terms, a type of "stronger". Mathematically, it increases your chance of success, which increases your expected damage. Gameplay-wise, it also increases the range of scenarios where you'd want to use a spell. For a counterexample, imagine a feature that added your character level to spell damage if the enemy failed their save, but subtracted it if they succeeded. On paper this would add damage (because PCs generally succeed about 2/3 of the time), but in practice it would probably be a net-negative; you'd be more likely to overkill enemies and more likely to get dud results, with a relatively small set of cases where you'd see a benefit.

Changing a damage type or a saving throw is usually a reliability boost. Eldritch Knights get a feature to force disadvantage against their spell saves; this is purely a reliability tool, but also numerically very strong. But also think about an Evocation wizard's Spell Sculpting, which makes it safe to fireball your friends. Or a Divination wizard's Portent, which can force a failed save (it's got more uses than just spells, but forcing enemies to fail saves is often the best use). Features that reroll damage or healing would also count.

More Upcasting

This is a feature design WotC has very rarely used (and I probably overuse). WotC's Order of Scribes has a feature that, 1/day, lets you cast a spell "counting as one level higher than normal". (It also gives you an extra spellcast, see below). I've written a bunch such features though: Antiquity, Geomancy, Mastery, and Spellshaper wizards all get some sort of bonus upcasting. One of the nice things about bonus upcasting is that it's universally applicable; it has pretty clean and predictable power with damage spells, but also works for control-oriented wizards (in fact, works better). It's also a really nice sort of feature to give to partial spellcasters, but we'll discuss that more later.

Note, of course, that this is a special case of "numbers go up" buffs. Generally more involved, because it asks the player to know more about their spells and be more careful with their selection.

More Spells

More spells! Everyone loves more spells. This can be a straightforward "you learn X spell". or it can be access to a new spell list, the way Divine Soul sorcerers get access to cleric spells, or KibblesTasty's Fiendish Bloodline can learn warlock spells. You can also put bonus spell lists (like the ranger/paladin standard, or the sorcerer wasn't-standard-originally-but-is-these-days origin spells) in this category. A warlock's expanded spell list counts, but also sucks as a feature. Lots of features in this vein also combine with the next category, which is...

More Spell Slots

For full casters, this can be a dangerous type of feature, because they already get to cast a lot of spells. The most basic sort of feature here is to grant a few extra spell casts, often something like "you learn X and can cast it once per day for free".

More aggressive is a feature that lets you regenerate spent spell slots. Consider one of the game's most annoying features: Divination wizards get Expert Divination at 6th, which lets them regain a lower-level spell slot anytime they cast a divination spell. The perfectly reasonable design goal is to make sure they're doing some divining when they're not busy Portent-ing. Divination spells are roleplay- and story-powerful but combat-weak. But the existence of this feature forces that trend into a hard rule: if you write a combat-powerful divination, then this feature immediately turns that into a degenerate strategy for div wizards to use. (The only combat divination WotC's ever written, Mind Spike, is basically trash for this reason.)

Spells as Triggers

This one can overlap with any of the others: "when you cast a spell, X happens". Maybe you can spend a resource to regain a slot, maybe you can force one enemy to make an extra save, etc. WotC's Abjuration and Necromancy wizards both create hit points (as a ward or as temps, respectively) when they cast leveled spells. More interesting is when the trigger is for something that's not directly related to the spell. WotC's Storm sorcerer gets a short flight on every spellcast. My Exploration wizard at high levels gets to tack an ocean-themed effect on top of any spell (lightning strike, prone-knocking wave, or gust of wind). In the same book, Ferromancy wizards get to wrap themselves in metal anytime they cast a spell for an armor boost.

One nice thing about this design is you do have the option of lightly interacting with the spell. That Ferromancy wizard includes the spell level as pat of the AC calculation. Knocking an enemy prone is especially effective if you combo it on top of a spell that creates difficult terrain.

Spells as Builders

This one is less common, probably because it's complex. You can use spellcasting to build up some other resource or condition. WotC's War Magic sub has a form of this with its Power Surge mechanic. My Stormborne sorcerer has a feature named Gathering Storm, where every spellcast increases your storm by 1 rank. At low rank the bonuses are relatively mild. At high ranks, you get tons of free upcasts. (But of course, it's very hard to get to a high rank and you won't be there for long.)

Casting Time

By far the most powerful lever, and one WotC has used cavalierly, is casting time. The common form here is "You can cast a spell with a casting time of 1 action as a bonus action instead". At a minimum, this enables someone to get a free cantrip in; so expect 4-20 bonus damage to come out that round. There are other forms, like replacing one attack, explicitly bonus action cantripping, etc. And you can adjust these; my Empire cleric gets a feature that lets you BA cast after an attack, but effectively "downcasts" it by requiring you spend a higher-level slot.

All that's to say that you don't need to shy away from this one entirely; just know how powerful it can be before you write it.

Conditions

This one can be a little more particular, but you can write features that interact with conditions, on the assumption that your spellcaster will be inflicting those conditions often. The most straightforward example is the charmed condition; WotC's Enchantment wizard at high levels can cause enemies not to know they've been charmed. I've written a few features before that deal bonus damage if the target is prone, or charmed, or frightened. These interactions are tricky because it's easy to get too strong, but they're definitely in the toolkit.

Concentration

There aren't a ton of ways to interact with concentration, but there are a few. You can grant advantage or even auto-success on concentration saves. You can replace concentration with fixed durations (usually one round; you're sacrificing a little duration overall in order to guarantee it). You can allow concentration on two spells at once. (That one is very strong; I think it's still very good on the Psion, where it's only available at high level and has a bunch of downsides. I'm not on board with creators who make it available earlier and easier). I've written features once or twice that allow other creatures to concentrate on your spells, which can be fun, and is a form of double-concentration that's not wild. See next!

Loaner Spells

This is a clunky term for a clear concept: letting another creature cast a spell on your behalf. WotC artificers get this with the Spell-Storing Item feature. I've written versions of it, most notably for my Artifice cleric and Antiquity wizard. In general, it's a really good approach for inventor- or alchemist-themed subclasses, where the idea of "imbuing a spell into an object" meshes very nicely with the concept. You can attach a feature like this to a different theme, but you'll want a good justification for it.

Vancian Spells

Closely related to loaner spells are Vancian spells; these are spells that follow the spellcasting rules of older D&D editions, modeled on the books of Jack Vance. That rule is "prepare a spell X times at the start of the day and you can cast it X times that day". Applied to spell memorization, it's clunky as hell and wildly unintuitive. But applied to an inventor or an alchemist, it's super clear; your "spells" are gadgets or potions, of course you have to prepare each one individually.

You can use this to write features by granting a subclass a limited number of Vancian preparations, either as extra spellcasts or with some other kind of boost. My Canvas bard gets to do this, painting a spell ahead of time which gets a free cast with a double-concentration benefit.

Next Time

Yeah, it was the right call to split this one up. I'm already at almost 3000 words! That's all the features I thought of off the top of my head plus some quick consults. Let me know what I missed in the comments and we can go over them next time, when I'll talk about how these various features interact with a full-caster's kit.