Kitchen Prep

Kitchen Prep is where you start. The PIE Cookbook calls this the environmental scan: before you cook, you look at what’s already in the kitchen. The single most common failure mode for first-time builders is reaching for a tool before naming the problem — so this stage exists to slow that down.

From the source

This adapts the Kitchen Prep → Environmental scan section of the PIE Cookbook.

Three questions, in order

1. What do you already have? Skills, relationships, existing tools, data, trust within the community. Most builders undercount this.

2. What’s the gap? State the problem as a sentence a community member would recognize: “organizers can’t document evictions without exposing the tenant.” Not “we need an app.”

3. What’s your working theory? The smallest intervention you believe would move the problem. This is a hypothesis, not a spec — you’ll revise it.

Why this protects you later

If you skip Kitchen Prep, an AI agent will happily build something — but you’ll have no way to tell whether it’s the right something, and no frame for saying no to a dependency that doesn’t fit. Naming the problem first is what lets the exclusion policy and the recipes do their job: they can only protect a decision you’ve actually made.

This is exactly where Socratic Intent Engineering sharpens Kitchen Prep — the Ten Questions turn “name the problem” into a rigorous practice: what do you want, why, and what does success actually look like (not the convenient proxy)?

Don’t categorize too early

The catalog’s largest bucket is deliberately the vaguest — Misc & Everything Else — because premature categorization hides options. During Kitchen Prep, browse broadly; you’re scanning the environment, not committing to a stack.

When your working theory is stable, move to Cooking →, where the catalog’s vetted ingredients are organized by what they’re for.