The five-note setup: copyable templates for AI context

Ask an assistant about a decision at work and you get advice you’ve read before. Considering a price increase? Raise them gradually, communicate clearly, watch for pushback. None of it is wrong. None of it is about you.

The cause isn’t the model. It’s that nothing you know about the situation ever reached it.

So what do you hand over? This is where most people stall. You can’t hand over everything, and putting what’s in your head into words every time you open a chat is tedious enough that you stop bothering. You ask the bare question and get the generic answer back.

The fix is to write it down once, in advance. Five notes are enough.

Existing notes often hide the real problem

You may already have the information. It is simply compressed into fragments that make sense only to you:

new clients only keep the two don’t repeat March

You can reconstruct the missing meaning from memory. An assistant cannot know what “new” refers to or what happened in March.

Before adding more material, decompress the fragments into decision-ready statements:

Goal: Raise rates for new projects and move toward fewer, higher-value engagements.
Constraint: Keep the two existing clients on their current rates this quarter.
Past decision: An across-the-board increase cost us a valued client. Do not repeat it.

Organizing notes is not mainly about making tidy categories. It is about turning private shorthand into material another capable reader can use to reason.

Why five

When an answer comes back generic, what’s missing is almost always one of these:

What you’re trying to accomplish (Goals)
What you can’t change (Constraints)
What you already tried and ruled out (Decisions)
Who the decision is for (People and audience)
How you want the answer shaped (Standards)

Keep those five as short notes and you can paste in whichever parts matter, one decision at a time. No finished vault, no system to build first. A few minutes of writing each.

Below is each one filled in. The running example is a two-person design studio deciding whether to raise prices for new clients. Copy the structure, swap in your own situation.

The five notes

Goals — what you’re trying to accomplish

A sentence or two per goal. Treat it as current rather than permanent; it will change as your priorities do.

Without this, an assistant assumes you want more of everything — more clients, more revenue — and optimizes for that.

## Goals
- This quarter: fewer, higher-value clients rather than more of them.
- Long-term: build a retainer base so revenue doesn't reset to zero each month.

Constraints — what you can’t change

Budget, deadline, team size, anything already decided that limits your options.

Without this, you get options you couldn’t act on even if you wanted to.

## Constraints
- Two long-standing clients can't absorb a sudden price jump.
- Team of two — no capacity to take on more clients even at higher prices.
- Studio brand still leans on "affordable," so a sharp jump needs a story.

Decisions — what you already ruled out

A line added at the moment you decide, with the reason attached. Written in one batch afterward, it never gets written at all.

Without this, you get advice to try the thing that already cost you something.

## Decisions
- 2026-03: Tried an across-the-board price increase. Cost us a client we wanted to keep. Won't repeat a blanket increase.
- 2026-05: Started quoting new work 20% higher than existing rates. No pushback so far.

People and audience — who the decision is for

Briefly, who you’re actually doing this work for. Advice changes depending on whether it’s for a client, a team, or a specific reader.

Without this, you get textbook advice that ignores the relationships you’re protecting.

## People and audience
- New inquiries: small businesses referred by past clients — price-sensitive, but value speed.
- Existing clients: two retainer accounts we want to keep at current rates.
- Advice should assume we're protecting relationships, not just maximizing revenue.

Standards — how you want the answer shaped

Where your quality bar sits, and how you like things written or argued.

Without this, answers come back abstract and you spend the next round making them concrete yourself.

## Standards
- Prefer specific numbers and dates over vague language ("raise prices" → "raise new-client rates by 15%").
- Flag any suggestion that would affect the two existing retainer clients — don't bundle them into general advice.

Don’t aim for complete

None of these need to be long. A few lines each, like the examples above, is a real start, and a blank field is fine — better than a guess you’ll have to unwind later.

The point of these notes is not to finish them. It’s to have somewhere for the next relevant thought to land.

Do not share all five every time

Creating five notes does not mean pasting all five into every conversation.

Use one test to select a line:

Could a capable colleague give the same advice without knowing this?

If yes, leave it out. If the answer could change, keep it.

A pricing question may not need your writing standards or design preferences. It probably does need the relationships you are protecting and the failed blanket increase.

The final handoff can be this short:

Context
- Goal: Move toward fewer, higher-value projects this quarter.
- Constraint: Keep the two existing clients on their current rates.
- Past decision: A blanket increase cost us a valued client. Do not repeat it.

Question
If we raise rates for new clients, give us three options for the first step.

A practical first week

Do not try to complete the system in one sitting.

Day 1: Create the five files and write one line in each.
Days 2–3: Take one piece of shorthand from your workday and return it to Goals, Constraints, or Decisions.
Day 4: Choose one real question and share only the lines that could change the answer.
Day 5: Record the option you chose, the option you rejected and why, and the condition that would make you revisit it.
End of week: Spend ten minutes removing stale, unclear, or duplicated lines.

Success is not a vault that looks organized. It is a conversation that needs less re-explanation and begins where the last decision ended.

What changes once you have them

The explaining stops. Before you ask, you pick the few lines that bear on this particular question and paste them in. Not the whole vault — just the context that changes the answer.

These five notes feed the Capture and Organize stages of the loop. Set up the five files in Obsidian in ten minutes, then use them on one real question.

Join the Starter Kit waitlist if you want one notification when the kit is ready.