Reasoning
Every recommendation has a path
The desk separates inputs, decisions, actions, and results so a cook can challenge the path instead of trusting a black box.
Decision Desk
Do not give me another list. Show me how to decide.
Visual flows, explicit tradeoffs, and reviewed teaching material for the moments when a recipe list is not enough. The reasoning behind a call, laid out so you can challenge it.
Reference data
Shared knowledge can live remotely
Reviewed public resources can be loaded from a shared catalog, so the library grows without an app update.
Privacy boundary
The household still stays local
Inventory, cook history, preferences, dates, and constraints stay on this device. The shared catalog is reference material only.
Visual flows
Pick the decision you are actually making
Each flow is intentionally small enough to use at the counter. The detailed tools stay on their existing surfaces; this is the map that tells you which tool to use and why.
Start with shelf truth, then spend urgency before novelty.
What should I cook tonight?
A five-step path from inventory to the most defensible dinner choice.
1 · input
Confirm the shelf
Use only what is actually present. Starter and inferred rows are not treated as household truth.
2 · decision
Find what is turning first
Expiry urgency changes rank; it does not create a safety verdict.
3 · decision
Look for a ready match
Prefer a complete method that uses the urgent ingredient without opening a new shopping trip.
4 · action
Compare the smallest unlock
If nothing is ready, ask which single missing ingredient unlocks the strongest useful set.
5 · result
Cook, skip, or smart-buy
The result is a reasoned choice with visible tradeoffs, not a recipe roulette wheel.
Decision matrices
When two answers both sound reasonable
These are not scoring systems pretending to be objective. They make the competing signals visible so the cook can see which side has the stronger case.
Cook now vs. buy one thing
| Signal | Cook from the shelf | Buy one unlock |
|---|---|---|
A complete method already exists A complete, useful dinner with no added dependency usually wins. | Strong | Weak |
One ingredient is materially urgent Urgency should move dinner forward, not become a notification that nothing acts on. | Strong | Only if the buy uses it |
One missing item unlocks several good meals The best purchase is the one with leverage across the shelf, not the fanciest recipe. | Maybe | Strong |
The missing item serves only one low-priority recipe A single-use errand is usually poor shelf economics. | Strong | Weak |
Substitute vs. buy the original
| Signal | Substitute | Buy original |
|---|---|---|
The missing ingredient is structural Binders, leaveners, gelling agents, and major liquid/fat ratios can change the entire method. | Cautious | Stronger |
The missing ingredient is primarily aromatic Aroma roles are frequently more flexible when another shelf ingredient points in the same direction. | Often strong | Optional |
The substitute changes water or fat substantially Functional similarity is not enough if hydration or richness shifts the system. | Adjust first | Safer |
The dish can be tasted and corrected incrementally Reversible changes support experimentation better than all-at-once structural changes. | Strong | Weak |
Shake vs. stir
| Signal | Shake | Stir |
|---|---|---|
Citrus or other cloudy juice is central Aeration and vigorous integration are usually part of the intended texture. | Strong | Weak |
The drink is spirit-forward and clear Stirring chills and dilutes while preserving a denser, clearer texture. | Weak | Strong |
Egg white, cream, or another foaming ingredient is present Texture development is part of the construction, not merely temperature control. | Strong | Weak |
The drink is already long with a carbonated mixer The carbonated component should not be aggressively shaken. | Before carbonation only | Build gently |
Pairing confidence
| Signal | Established | Experimental |
|---|---|---|
Strong culinary precedent + strong aroma overlap Two independent signals point in the same direction. | High | Low |
Strong precedent + weak measured overlap Culinary history can capture interactions a simplified aroma model misses. | High | Low |
Weak precedent + strong aroma overlap This is the classic wild-card zone: plausible enough to test, not enough to promise. | Low | High |
Weak precedent + weak aroma overlap There is little evidence either way; a bridge ingredient or different direction may be more productive. | Low | Very high |
Learning library
Teach the technique beside the decision
Reviewed video for the flow you have open, rather than a wall of everything. Picks are made for what they teach and the source they come from; a channel with reach is evidence that people found it useful, not evidence that it is right.
Reviewed copy that ships with the app. No household inventory, cook log, preferences, or personal constraints are sent to this catalog.
For "What should I cook tonight?"
Third-party video. Opens in a new tab — nothing is embedded and nothing is fetched from a video platform until you click.
Three picks, chosen for this decision. The rest of the library is on the Watch page.
Open the full watch library — 90 reviewed entries across 21 techniques, searchable.