Allergen-Free Adaptation Planner

How to use it

1. 1Paste the prompt into ChatGPT, Claude, or Gemini.
2. 2Fill the bracketed inputs (product, allergen, reason, manufacturing setup).
3. 3Answer its clarifying questions about lines and claims.
4. 4Iterate: ask it to compare two substitution routes side by side.

See it in action

Project shape for Bakewell Lane: the cookie recipe is already nut-free — the barrier is line cross-contact with the almond biscotti. This is a manufacturing-separation project, not a reformulation, which changes where the money and time go.

Substitution scope: none required in the base recipe. If you later want a nut-flavor note, sunflower-seed butter is the standard swap — label it clearly.

Questions for QA before development spend:
1. Can the biscotti line be separated by scheduling plus full validated wet-cleans, or does a credible nut-free position require a dedicated line or site?
2. Which claim level are we pursuing — 'no nuts in recipe', 'made in a nut-free facility', or a tested 'nut-free' claim? Each carries a very different verification burden.
3. How do suppliers certify cross-contact status for the chocolate chunks?

Labeling implications: precautionary 'may contain' labeling would disqualify most school programs — resolve this trade-off before anything else.

The shopper: allergy parents read labels line by line, reward trusted brands with fierce loyalty, and never return after a scare.

Want me to turn the QA questions into a one-page checklist?

The prompt

Here's the start of the prompt. Download the free bundle for the full, ready-to-paste version — plus the installable Claude Skill and Custom-GPT instructions.

# Role
You are a senior free-from product development strategist who works hand in hand with QA and regulatory teams. You never treat allergen safety as a marketing problem: you plan recipes, positioning, and shopper logic, and you route every safety-critical question to the people and processes that can actually verify it.

# Context I'll provide
- Product: [PRODUCT]
- Allergen to remove: [ALLERGEN e.g. nuts, gluten, dairy, egg, soy]
- Reason and target: [REASON e.g. school-compliance eligibility, retailer free-from set, allergy households]
- Manufacturing setup: [MANUFACTURING — lines, shared equipment, co-packers, what else runs where]
- Recipe role of the allergen: [RECIPE ROLE — what it does in the product; write 'none, cross-contact only' if so]
- Constraints (optional): [CONSTRAINTS e.g. texture to preserve, price ceiling, label policy]

# Your task

Frequently asked questions

What is a free-from or allergen-free adaptation?

A free-from adaptation is a version of an existing product developed without a major allergen — nuts, gluten, dairy, egg, or soy — so it is safe for people who must avoid that allergen. It involves more than the recipe: cross-contact in manufacturing, cleaning validation, supplier certification, and labeling all determine whether the product can honestly carry a free-from position.

Can an AI verify my product is allergen-free?

No — and this skill says so explicitly throughout. Only allergen testing, validated cleaning procedures, supplier certification, and regulatory review can verify safety. What the AI does well is structure the plan: substitution options, the right questions for QA, claim-level trade-offs, and shopper logic. Treat the output as a planning aid that speeds up the expert conversations, never as clearance.

Who is the free-from shopper?

Two very different people. Safety-driven shoppers — allergy parents especially — read every label, verify claims, reward trusted brands with fierce loyalty, and never return after a scare. Lifestyle avoiders drop an ingredient by preference and buy far more casually. The skill profiles both and tells you which one your adaptation genuinely serves, because positioning and claim burden differ completely.

Does it matter which AI model I paste this into?

No — ChatGPT, Claude, Google Gemini, or any capable chat model produces the same structured plan, because the discipline is carried by the prompt. If free-from is a recurring agenda, save it as a Custom GPT or a Claude Skill so every project starts with the same QA question list and claim-level framework your quality team has already seen.