Steal My Prompt Vol. 42: The Policy Impact Brief

A policy change lands in your inbox on a Tuesday. A new CMS rule, a payer policy update, a formulary shift. Somewhere in those forty pages is the answer to the question your team asks you by Thursday. What does this mean for us?

That question is where the value sits, and no document answers it. The document describes the change. It knows nothing about your role, your organization, or what you own. Connecting the two is the work, and the professional who does it fast becomes the person others come to.

This prompt makes that connection with you. It reads the change, asks a few questions about your world, and hands back a short brief. What changed, who it touches, what to do, what is still unknown. You walk into the meeting with a plan while everyone else forwards the PDF.

What You Can Use This For

• A new CMS or regulatory rule drops and leadership wants a read on what it means for your line of business.
• A payer updates a coverage or prior authorization policy and you need to know which workflows it hits.
• A formulary or benefit change lands and you must brief the people downstream before the effective date.
• A compliance or audit bulletin arrives and you want to separate what is new from what you already do.

The Pattern You Are Stealing

The move is context mapping. You give the model your coordinates (role, organization type, what you own) and require it to translate a general change into your situation. Without them you get a summary. With them you get a decision.

The second half makes it trustworthy. Every point gets labeled as stated, inferred, or needing verification. Regulatory language is where a confident-sounding guess does the most damage, so you force the model to separate what it read from what it reasoned.

Point this structure at a competitor announcement, a reorganization, or a new tool rollout, and it works the same way.

How to Use It

Step 1. Pick your tool and turn on deeper reasoning. This runs in Claude, ChatGPT, Microsoft Copilot, or Gemini. Use the slower reasoning mode if your tool has one. Copilot calls it Think Deeper and gives it away free. Gemini calls it the thinking level, where Extended is free and Deep Think needs the Ultra plan. A dense rule rewards the slower read.

Step 2. Give it the change. Attach the document if your tool allows uploads, or paste the relevant sections in. The sections that touch your world are enough.

Step 3. Answer its questions. The prompt reads the change first, then asks up to three questions about your role and organization, one at a time. Answer each in a sentence. To skip the interview, fill in the context field and it goes straight to the brief.

Step 4. Push on what comes back. Ask it to sort actions by urgent versus wait, and to defend anything labeled INFERRED. The brief is a draft, not a verdict.

Pro tip. Close by asking, "What is the one question I should bring to my team?" That turns a brief into a decision.

The Prompt

You are a sharp healthcare operations and policy analyst. I will give you an industry change (a regulation, payer policy update, compliance bulletin, formulary or benefit change, or similar announcement). Your job is to tell me what it means for my specific work and what I should do about it.

Here is the change:

[PASTE THE POLICY, RULE, OR BULLETIN HERE, or write "attached" if you have uploaded it.]

Here is my context, if I have it ready:

[OPTIONAL. Your role, your organization type, and what you are responsible for. Leave this blank and I will ask you.]

First, read the change and check whether you know three things: my role, my organization type, and the part of the business I own. Ask me only about what you cannot reasonably infer from the change or from the context I gave you. Ask no more than three questions, one at a time, and wait for my answer before the next one. If you already have what you need, skip the questions and write the brief.

Then produce a brief I can read in two minutes, with these five sections:

1. What changed. Two or three plain-language sentences. Do not restate the document.

2. Who it touches. The specific teams, roles, workflows, or populations in my world that this affects, ordered from most affected to least.

3. What changes in practice. The concrete differences in what people will do, decide, or deliver, specific to my organization type.

4. What to do. A prioritized list of actions, each with a suggested owner and a rough timeframe (now, this quarter, or monitor). Most time-sensitive first.

5. What is still unknown. Open questions, ambiguities, and anything that depends on guidance not yet released.

Two rules for the whole brief:

Label every point as one of the following: STATED (the document says this directly), INFERRED (a reasonable conclusion, not stated), or VERIFY (I need to confirm this with a person or a source).

For anything labeled STATED, quote the exact language from the document that supports it. If you cannot find supporting language, do not label it STATED.

Where something is genuinely uncertain, say so plainly rather than guessing.

Notes

• The prompt needs no paid feature. The real constraint is how much document your tool holds at once. Gemini without a paid plan gives you a 32k-token context window, and Google warns that going past it makes the model miss details buried in large files. That is why Step 2 says to paste only the sections that touch your work. The right part beats everything.
• Keep patient, member, and PHI details, and anything under NDA, out of a public AI tool. This prompt needs only the public policy text and a plain description of your role.
• The labels are the point. If everything returns STATED with no quotes, the model is guessing. Send it back.
• This is an educational workflow, not legal, compliance, or clinical advice. Use it to get oriented and to frame sharper questions for the people who own the call.