Low-poly 3D scene of a person at a night-time crossroads holding a glowing one-page plan while a small robot guide points the way

Quick answer

To write a one-page plan for an AI advisor, keep a short document you write yourself with five sections: your top goals this year, what you are not doing, what gives you energy, what drains you, and your personal non-negotiables, plus a line on how you want advice delivered. Then save a reusable “advisor” skill that reads the doc before answering, so every decision is checked against your plan rather than generic advice. Keep it to one page, and never let the AI edit it.

Key takeaways

  • Less context gives better advice. One focused page beats dumping all your files on the AI.
  • Include what you are not doing. That section does most of the work when saying no.
  • Write it yourself and mark it as human-edited only.
  • Ask for the reasoning: which part of your plan the answer is based on.
  • Use it monthly for direction, and when a real decision comes up.

📘 Main article: This guide expands on the advisor in our article Personal AI System: How Peter Yang Runs a One-Person Business With AI, based on Matt Wolfe’s episode with Peter Yang.

An AI chatbot will happily give you advice on anything. The trouble is that it does not know what you are trying to achieve, so the advice is generic. In our breakdown of Peter Yang’s personal AI system, his “advisor” works differently because it measures every answer against a single page he wrote about his own goals.

This guide shows you how to write that one-page plan for an AI advisor, turn it into a skill, and test it with a real decision. It works with Claude, ChatGPT or any assistant that can read a linked document or saved file.

Why does an AI advisor need a one-page plan?

Peter’s point in the video is that AI models are good at picking up context, but if you tell one to look at all your memory and documents, it gets “too much”. Important priorities are buried under old notes, half-finished ideas and other people’s plans.

A single page forces you to decide what matters. It also gives the AI a clear standard to measure against, so it can say “that fits” or “you said you weren’t doing that this year”.

What should the one-page plan include?

SectionWhat to writeWhy the AI needs it
Top goalsThree goals at most, ideally measurableTo judge whether an opportunity moves you forward
Not doing this yearThings you have decided to avoid, plus hard limits (for example, one trip a month)To spot distractions and say no for you
Gives me energyWork you enjoy and do wellTo favour options you will actually follow through on
Drains meTasks that wear you outTo flag hidden costs of an opportunity
Non-negotiablesFamily, health and personal commitmentsTo protect what work should not override
How I want adviceTone and format, for example “be direct”To stop vague, overly positive answers

Matt Wolfe’s version in the video used almost exactly these sections, plus a rule at the top that only he edits the document.

What does a filled-in template look like?

Here is an example for a small online business. Replace everything with your own details; the shape is what matters.

MY PLAN FOR THIS YEAR (I edit this doc. AI may suggest changes, never make them.)

Top goals this year (3 max)
1. Grow my email list to a number I choose and can measure
2. Launch one digital product
3. Keep Fridays free for planning and learning

Not doing this year
- No new social platforms
- No client work under my minimum fee
- No more than one work trip a month

Gives me energy
- Writing guides, recording short videos, talking to customers

Drains me
- Long calls with no agenda, chasing invoices, editing long videos

Family and personal non-negotiables
- School run on weekdays, no work on Sunday

How I want advice
- Be direct. Tell me what I need to hear, not what I want to hear.
- Say which part of this plan your answer is based on.
- If you're missing information, say so instead of guessing.

💡 Tip: Keep each line short and specific. “Grow the business” gives the AI nothing to measure; “no client work under my minimum fee” does.

Why should you write it yourself?

Peter calls his plan a “slop-free zone”. The AI can read it and suggest changes, but he writes and edits every word.

There are two good reasons. First, the plan is only useful if it reflects what you want; if the AI rewrites it, it slowly drifts towards generic goals. Second, it keeps you honest. If the advisor keeps telling you something clashes with your plan, either the decision or the plan needs a deliberate change from you.

⚠️ Warning: Do not put passwords, bank details or private information about other people in the document. Keep it to goals and preferences you would be comfortable sharing with the AI tool you use.

How do you turn the plan into an advisor skill?

  1. Save the plan in Google Docs or a simple text file, and copy its link or location.
  2. Open your AI tool and make sure it can read that document (in Claude, the Google Drive connector does this).
  3. Ask it to create a skill (or a custom instruction, project or custom GPT, depending on your tool) using the instructions below.
  4. Save it with a clear name like “advisor”.
Name: advisor
Purpose: Help me make decisions that fit my plan.
Before answering, read my one-page plan: [link to your doc]
For any decision or opportunity I share:
1. Give a short answer first: yes, no, or "only if".
2. Explain which goals, limits or non-negotiables it supports or breaks.
3. Check my calendar if relevant and flag clashes.
4. List what you don't know and would need to be sure.
5. If I still want to go ahead, give the conditions that would make it worthwhile.
Never edit my plan. You may suggest changes at the end, clearly labelled.

Matt says creating his version took about 30 seconds once the doc existed. The work is in writing the plan, not the setup.

How do you test the advisor?

Test it with a real decision you are facing. Matt asked his advisor whether to accept a two-day conference trip in March. It said no, because the dates clashed with an existing trip and broke his own one-trip-a-month rule, and it listed the conditions under which a yes would make sense.

A good answer from your advisor should:

  • Name the specific section of your plan it relied on.
  • Mention clashes with your calendar or limits.
  • Admit what it does not know.
  • Avoid generic encouragement.

If it gets something wrong, do not rewrite the skill immediately. Note the problem and fix it in a batch later, as explained in our guide How to Improve an AI Skill Without Filling It With One-Off Rules.

How often should you use it?

Peter uses his advisor for direction about once a month, and says more often than that is too much for long-term planning. It is different from a daily “chief of staff” briefing, which handles what needs doing today.

  • Monthly: “Here is what I did this month. Am I on track with my plan?”
  • When an opportunity arrives: a collaboration, a trip, a new product idea.
  • Every quarter: reread the plan yourself and update it by hand.

Peter also admits he uses it as a sounding board when he is frustrated, and that he is not sure how healthy that is. It is a tool for checking decisions, not a substitute for talking to people you trust.

Frequently asked questions

Can I use ChatGPT instead of Claude for an AI advisor?

Yes. Any assistant that can read a document you link or upload and follow saved instructions will work. Custom GPTs, Claude projects and skills all suit this.

How long should the plan be?

One page. If it grows longer, cut it back. Short, specific lines produce clearer advice.

Should the AI update my plan for me?

No. Let it suggest changes at the end of an answer, then decide yourself. That keeps the plan genuinely yours.

Is an AI advisor the same as a chief of staff agent?

No. A chief of staff handles today’s tasks from your inbox and calendar. An advisor checks bigger decisions against your long-term plan.

The bottom line

An AI advisor is only as good as the plan it reads. Spend half an hour writing one honest page about what you want, what you are avoiding and how you like advice delivered. Then let the AI measure decisions against it, while you stay in charge of the plan itself.

author avatar
Garry Knight
I'm Garry Knight, the person behind Prodify Digital. I write about email list building, email marketing, SEO, AI search and the tools that connect them. My aim is to make online marketing easier to understand, so creators and small business owners can make informed decisions about building an audience and keeping people engaged. Here you'll find straightforward guides and product reviews that explain what something does, where it fits and which limitations matter. The focus is on clear explanations and useful next steps—not hype, shortcuts or promises of easy earnings.

Leave a Reply