A muffin and a cookie want opposite things.

The muffin base is engineered to build structure — protein, moisture, leavening, a wet batter. That's why it's fluffy and springy.

A cookie is good precisely because structure fails to form. In a normal cookie, fat does that job: it coats the flour so the network can't build. We're not using much fat. So we sabotage the structure four other ways.

### ⚠️ The golden rule: if it pours, it's a muffin.
This must be a stiff dough you can roll into a ball. Play-doh, not batter.

And be honest about what this is: not shortbread. "Short" — crumbly, sandy, melts on the tongue — is fat, and shortbread is roughly half butter. We can't get there and shouldn't pretend to. This cookie lives at the chewy-crisp end: chewy center, crisp edge, hearty. Think Anzac biscuit (#101), oatmeal cookie, gingersnap. That's a great cookie. It's just a different one.

The full science, with citations, lives in RESEARCH.md — 22 peer-reviewed sources, 99 claims extracted, 25 adversarially fact-checked, 3 killed.

Makes 12 cookies (~35g dough each).

Batch: 1,336 cal · 40.7g protein · 22.8g fiber → 111 cal · 3.4g protein · 1.9g fiber per cookie.

Method

1. Whisk the dry. Whisk the wet separately. Combine.
2. The squeeze test. Grab a fistful. It should hold together and take a shape. Crumbles apart? Add water 1 tsp at a time. Sticky, or slumping? You've gone too far — add a spoon of oat flour.
3. Do NOT chill. Chilling is for solidifying fat. With no fat it just hydrates the flour and costs you spread.
4. Roll 12 balls (~35g each). Flatten to ~1cm. They won't spread much on their own — shape them as you want to eat them.
5. 375°F, 10–12 min. Pull when the edges are golden and the centers still look slightly underdone.
6. Cool ON the pan, 5+ minutes. ← Don't skip. They come out soft and set as they cool.

The failure mode to watch for is leathery, not bready. Low fat + oats + too much water + too long in the oven = rubbery. There's a famous review of a fat-free oat cookie that just reads "tastes like an inner tube." Under-bake rather than over-bake — with no fat, there's nothing to keep it tender once you dry it out.

220g of flour, split however you like. Same dial as the muffin board — but here it moves texture, not just macros.

Per 10g shifted oat → chickpea: −0.03 cal, +0.06g protein, +0.01g fiber per cookie.

Important — this dial is backwards from the muffins'. There, more legume was free protein and you took as much as flavor allowed. Here, chickpea buys protein but spends tenderness. Peer-reviewed sensory panels fall off past ~50% chickpea, and 100%-chickpea shortbread measured 61% harder. So each flavor's "best at" is the most chickpea it can carry before it turns hard or beany — not the most it can hide.

Whatever isn't chickpea is the grain slot (150g at the house ratio). It can be oat, milled hard red wheat, or a split.

Here's the part the board got wrong at first: wheat WINS on every macro.

Swapping the whole grain slot to wheat: −4 cal, +0.3g protein, +0.4g fiber per cookie. Lighter AND more protein AND more fiber. There is no nutritional argument for oat.

So it's purely a texture trade:

🔑 Wheat's "downsides" are upsides for snappy cookies

Everywhere on this board that wheat's bran is a liability — hardens, kills spread — is a place we wanted chew. But some cookies want the exact opposite:

#121 Thin Mint · #19 Gingersnap · #102 Biscotti · #92 Anise pizzelle · #20 Vanilla crisp · #73 Gingerbread · #61 Black cocoa · #104 Danish vanilla · #126 Trefoil · #25 Cornmeal · #31 Espresso crisp · #79 Speculoos · #103 Speculaas · #21 Chocolate crinkle · #22 Lemon crinkle

For these, hardness and low spread are the goal. Wheat gets you there, gives you better macros, and — because real cookies are made of wheat flour — tastes more like the thing you're duping. Oat's toasty note is the #1 tell that a dupe is homemade. Wheat removes it.

On the site these are flagged 🌾 wants wheat, and every recipe has its own grain toggle.

🔧 Can't grind coarse? Don't sweat it — here's the fix.

The effect is asymmetric. The research says FINE flour "remarkably increased hardness and decreased spread", while COARSE flour only "marginally reduced both." So fine is a real penalty; coarse is a modest bonus. You lose very little by not grinding extra-coarse.

And grind is the LAST of the six levers. In order of actual impact:

1. Water — dough, not batter. The whole ballgame. Completely independent of your mill.
2. No baking powder.
3. Honey + crystalline sugar.
4. Flatten thin, don't overbake.
5. Chickpea ratio.
6. Grind. ← the smallest, and the easiest to work around.

The hack: you don't need a coarse mill setting. You need coarse PARTICLES.

Pulse the rolled oats in a food processor instead of milling them. That gives an uneven, flaky, coarse meal — a better particle profile than most grain mills produce. Or just leave some oats whole. That's literally what an oatmeal cookie is.

The ideal build — needs no mill setting at all:

• Chickpea (70g): mill it however your mill wants to. Fine is fine — it's only 32% of the flour, and it's your binder. (The research is firm that you need a fine minority for cohesion: all-coarse dough was "impossible to laminate.")
• Oat (150g): get your coarse fraction here. Pulsed, or partly left whole.

That's coarse-dominant with a fine minority — exactly the optimum the papers describe, achieved with a food processor.

(Wheat is different — you have to mill the berries — but whatever your mill gives you is fine. Wheat's job here is SNAP, and a finer wheat grind actually helps it.)

Why the crystal slot is load-bearing: sugar isn't just sweetness — it's the #2 anti-structure agent after fat. Dissolved sugar competes with the flour for water, starving the network that would otherwise make bread. Then it re-sets on cooling and gives you snap. Pull it out and the flour gets all the water to itself.
For scale: this base is 120g sweetener : 220g flour ≈ 55%. A Tollhouse cookie runs 90–100%. We're already lean.

• Canned / cooked chickpeas — ~60% water. This is exactly what made the first batch spongy. Mill dry chickpeas instead.
• Baking powder — a cake ingredient. Out.
• Greek yogurt — ❌❌ Never. It's the muffin's protein engine because it's water, which is the one thing we're fighting. Adding yogurt to "healthy up" the cookie converts it straight back into a muffin. This is the single most important rule on the board.
• Purées as a "fat replacer" — the research claiming bean/pea purée and oat beta-glucan replace fat was actively refuted. They don't replace fat. They just add water.
• Chilling the dough — pointless without fat, and it costs you spread.