The AI Disney Method Coach skill for Claude

The problem the skill tackles is stated in the file itself in a few lines: a dreamer without a realist gets nothing built, a dreamer with only a critic gets stuck in conflict, and a realist without a dreamer builds something mediocre. Only all three together delivers quality. The trouble is that in practice all three happen at once. You come up with something, and while you are still coming up with it you already hear yourself saying it is too expensive and that you have no time for it. What dies then is not the bad idea but the unfinished one, and at the start every idea is unfinished.

It is meant for entrepreneurs, marketers, team leads and creatives who are weighing something up: a launch, a campaign, a new service, a pivot, or simply a decision they keep hesitating over. It is part of the skill library NL that we make freely available through our AI and automation service, with no account and no sales email afterwards. If you first want to know what a skill technically is and where the file format comes from, that is covered in what are Claude skills NL.

There is one expectation I want to correct up front. It does not decide for you. At the end it does deliver a recommendation with reasoning, in the form of proceed, adjust or park, but that stays advice from a language model based on what you have supplied. The value lies in the rhythm it enforces: at the end you have dreamed once without restraint, planned once soberly and looked once sharply, instead of doing all three half-heartedly at the same time.

Criticism while dreaming kills the dream. That is stated literally in the coaching rules: no but-what-about during the Dreamer phase. The reason is not vague. The moment an idea is judged while it is being spoken, only what you already knew survives, because only that passes a quick test. The odd variants that later turn out to be the good ones do not survive that first second.

Vague dreaming while planning does not produce a plan. The opposite mistake gets an equally hard rule: no vague dreaming during the Realist phase, keep probing until there are deadlines. A plan without an owner and without a date is an intention with bullet points. We have written lists like that ourselves, and nothing came of them. In the output of this role, every step therefore states who does it and when it is done, plus the minimum you can already start this week.

Criticism without a fix is complaining. The third rule is why the Critic in this method is useful rather than demotivating: the Critic is constructive, every piece of criticism gets a fix. The goal is to make the plan better, not to tear it down. In the example from the file that produces four objections with four concrete adjustments, and not one of them is scrapping the idea.

The order is not interchangeable. A rule comes with it: work through the three roles strictly in this order, do not switch between roles, and close each role explicitly before moving on. There are even fixed closing lines built in. That seems excessive until you see for yourself how quickly a group slips back into talking over each other the moment the boundary is not marked out loud.

Here is what the difference looks like between thinking things through all mixed together and looking three times separately. The lines with a plus come literally from the example in the SKILL.md, the lines with a minus are written for illustration, in the mixed form the method is specifically meant to prevent.

The method itself is short and strict. Step 1: ask the topic and keep probing until you know what needs to be tested. Step 2: work through the three roles strictly in order, do not switch between roles, and close each role explicitly before moving on. After that come the three role blocks, each with its list of questions, a description of the attitude and the fixed closing line. Six questions for the Dreamer, seven for the Realist, seven for the Critic, twenty fixed questions in total.

After the roles comes the synthesis, and that has five parts: the vision from the Dreamer in two sentences, the three most important actions with a deadline from the Realist, the top three weak points with a solution from the Critic, a decision in the form of proceed, adjust or park with reasoning, and finally a first action that is done within seven days. That comes with a literal output template too, with the headings and the ranges built in, for example three to seven steps for the Realist and three to seven objections for the Critic.

There is also a block with seven rules for the coach, and they explain why the output reads the way it reads. Roles strictly separated, no but-what-about with the Dreamer, no vague dreaming with the Realist, every piece of criticism with a fix, spoken language with sentences of alternating length, no jargon and no English terms unless they are standard, and address the user as you and yours. Want to set up a file like this yourself? The method is set out step by step in writing a SKILL.md NL.

What is notably absent is also information: there is no prescribed duration, no number of participants, no scoring model and no weighting formula. Nothing gets added up and nothing gets calculated. It guards an order and forces a different kind of answer per role, and that is the whole machine. Simpler than it sounds, and that is exactly why it holds up.

The method was modelled, not invented by Disney. Robert Dilts, a founder within NLP, described it in 1994 in Strategies of Genius, Volume 1. He based it on interviews with people who had worked with Walt Disney. Disney reportedly had three separate rooms for three modes of thinking. That word reportedly is not there by accident: it is a handed-down account, not a documented working protocol of the studio. What you are using, then, is Dilts's model of Disney's way of working, and that is something different from a method Disney himself wrote down.

Dilts translated it into parallel thinking. That is the technical term for what the skill does: instead of different people defending different points of view, everyone looks at the same time from the same role, and the group then switches together afterwards. The description in the SKILL.md reads: a technique in which a person or team deliberately switches role to generate ideas, make them workable and put them to the test. Anyone who wants to apply the idea of parallel thinking more broadly naturally ends up at the Six Thinking Hats skill, where the same principle is worked out in six directions.

The core reasoning fits in four lines. A dreamer without a realist gets nothing built. A dreamer with only a critic gets stuck in conflict. A realist without a dreamer builds something mediocre. Only all three together delivers quality. Those four sentences are the entire basis of the model, and they are honest about what they do not claim: there is no impact measurement attached, no percentage, no outcome research. It is a working format, not a proven intervention.

That does not mean it is arbitrary. The underlying idea, that separating generating from judging produces better results than doing both at once, turns up in almost every creativity method. In Design Thinking it is called diverging and converging, in brainstorming it is called deferring judgement. The Disney Method is a tight, fast-to-run variant of that, with three roles anyone understands within a minute.

After that, using it is simple: name your idea and the skill keeps probing until it knows what needs to be tested. If you get stuck anywhere, the full step-by-step plan per environment is in installing Claude skills NL, and the broader explanation of working with AI is in the knowledge base. Does your team need to be able to run this kind of session itself, with or without AI involved? That is the subject of our AI training NL.

It names three situations where you are better off leaving it aside. The first: pure operational decisions where no vision is involved. If the question is which of two quotes to accept, you do not need a dreamer. The second: a crisis that calls for direct action, not a brainstorm. The third: data analyses that call for a quantitative answer, because the skill does not calculate anything. In those cases the file itself refers to a more suitable framework, naming the Eisenhower Matrix, the OODA Loop or a data-analysis skill.

Two more honest limits I would add. The first: the Critic in this method is a constructive critic, and that is deliberate, but it also means you do not get a merciless stress test here. If you want a plan properly torn apart before the market does it, the Pre-mortem Analyst NL or the Red Team Analyst is the sharper tool. The second: everything the skill produces comes from what you supply. It does not know whether your target audience exists, whether your price is right, or whether anyone is waiting for this. It makes sure you ask those questions at a moment when you can actually hear them, and that is different from giving you the answers.

An idea that has been through three roles stands steadier than an idea that only ever lived in your head. Then the next question begins: do the people who need it know it exists. That is our trade. The free SEO scan shows in a few seconds where your site currently stands, and if you want to talk through what else AI can do for your business, from separate skills to full automation, we simply do that in a conversation.