Lesson 21 of 32 · Part One, the principles

The Master's Review

Two reviews make the difference between a gamble and a plan. The first is yours: before you press send, read your own prompt as if you were the machine receiving it, and look for the word that could be read two ways or the detail you forgot. The second you ask for: have the model inspect your request for gaps before it answers, and inspect its own draft afterwards - through the eyes of the person who will read it.

Reach for it whenBefore anything that goes out of your hands - a message to an upset customer, notes you will revise from, a chapter you are about to send to a reader - and whenever your first instinct is to type "is this good?"

How to use The Master's Review in Gemini

The prompts came from here and need only tightening; what this platform adds is a second opinion that is not the model's own. On a factual answer, "double-check response" has Search go through the statements one by one and mark what it can and cannot support - which is a genuinely different kind of review from asking the model to grade itself, and the right tool for revision notes full of section numbers. Then there is the quiet last step the chapter is really about: export the final version to Google Docs and read it yourself, on a page, away from the chat. Reviews done in the scroll of a conversation are not reviews.

  • Run double-check on factual answers, then follow the links.
  • Say "do not search" when reviewing your own writing.
  • Export to Docs and read it on a page, away from the chat.

From GenAI for Everyone: Prompt Engineering Handbook, First Edition 2025, by Raja Rajeshwari Seshadri and Shreeram Iyer.

The prompts came from here and need only tightening; what this platform adds is a second opinion that is not the model's own. On a factual answer, "double-check response" has Search go through the statements one by one and mark what it can and cannot support - which is a genuinely different kind of review from asking the model to grade itself, and the right tool for revision notes full of section numbers. Then there is the quiet last step the chapter is really about: export the final version to Google Docs and read it yourself, on a page, away from the chat. Reviews done in the scroll of a conversation are not reviews.

From the book After Gemini answers, use "double-check response" to have Google Search highlight statements it can or cannot support - useful when reviewing facts like tax rules. Export the final version to Google Docs for your own last read.
Worth knowingDo not mistake double-check for verification. It shows whether Search can find support for a statement, which is not the same as the statement being right: an unsupported line may be perfectly correct, and a supported one may be sourced from a coaching blog repeating the same error. For tax matters, follow the links and land on the CBIC or the Act itself. It is also no use at all on the two creative examples - whether a sentence shows a feeling instead of telling it, or whether an apology sounds cold, is not a claim the web can confirm. And grounding a review of your own memoir will drag in generic writing advice. Say plainly: do not search, review only what I have pasted.
  • Run double-check on factual answers, then follow the links.
  • Say "do not search" when reviewing your own writing.
  • Export to Docs and read it on a page, away from the chat.
  • Upload the official PDF and tell it to work only from that.
  • Ask it to question your request before it writes a word.
The Unread Order vs. The Reviewed Request

Instead of

Explain GST input tax credit for my exam.

Gemini version

[If you have the bare Act or the official CBIC extract as a PDF, attach it and tell Gemini to work from that alone.]

I am a CA Intermediate student in Surat, exam in May.

First, question my request. Here it is: one page of revision notes on GST input tax credit - conditions for claiming, blocked credits under Section 17(5), time limits, short points, one worked example with round numbers. Tell me what is unclear or missing and ask me up to three questions before writing anything.

Once I have answered and you have written the notes, I will run double-check response on them. So keep each factual claim as its own short line - a section number, a time limit, a rate on a line of its own - because a claim buried inside a long sentence cannot be checked cleanly.

Then export the notes to Google Docs, where I will read them against my study material line by line.
Open Gemini 870 characters

Then loop it

  1. Good questions. My exam is in May, I know the basics already, so give most space to blocked credits.
  2. Now quiz me with five tricky MCQs from these notes. Put the answers at the end.

Why it worksIt makes the AI inspect the order for gaps before acting, and review its own answer for weak points afterwards.

"Is It Good?" vs. A Clear Editor's Brief

Instead of

Is my chapter good? Please check it.

Gemini version

[Attach the chapter as a document rather than pasting it, so the formatting survives.]

Do not search for anything. I want a reading of my pages, not general advice about memoir writing.

Be a kind but honest memoir editor. The attachment is the first chapter of my memoir, about 1,200 words, about my first teaching job in a village school near Kottayam in 1978.

Look for three things and nothing else: where the reader may lose track of time or of people; sentences that run too long; and moments where I announce a feeling rather than letting it show.

Up to ten numbered suggestions, each quoting the line it is about. Do not rewrite my chapter - I write plainly and I intend to go on writing plainly.

Afterwards, export your suggestions to Google Docs so I can sit with them beside my own printout.
Open Gemini 805 characters

Then loop it

  1. Thank you. I disagree with suggestion 4 - that headmaster really was that strict. Skip it, and show me how you would fix suggestion 2 only.
  2. Read your own suggestions again. Which three matter most for a first-time reader?

Why it worksA named set of review criteria turns a vague "check it" into a focused critique she can act on line by line.

The First Draft vs. The Self-Critiqued Draft

Instead of

Write a sorry message to an angry client.

Gemini version

Do not search. This is my own apology in my own words, and anything found on the web will read like a company notice.

I am a wedding photographer in Jaipur. A couple is upset because their album is three weeks late - the printing lab's machine broke down. Write a WhatsApp reply under 100 words that apologises without pushing the blame onto the lab, gives the new delivery date of 20 November, and offers ten extra prints free.

Then review your own draft as though you were the bride, reading it on her phone while still angry. Quote any line that sounds defensive, cold or copied from a template, and say how she would take it. Give me one improved version afterwards.

I will send whichever version I choose from my phone, so keep it as plain text with no formatting that will not survive WhatsApp.
Open Gemini 803 characters

Then loop it

  1. Better. The second version still says "unfortunately" twice. Remove that and make it warmer.
  2. Now give me a Hindi version I can send to the bride's mother.

Why it worksAsking the AI to critique its draft through the reader's eyes catches tone problems before the message is sent.

Questions

What is The Master's Review in prompt engineering?

Two reviews make the difference between a gamble and a plan. The first is yours: before you press send, read your own prompt as if you were the machine receiving it, and look for the word that could be read two ways or the detail you forgot. The second you ask for: have the model inspect your request for gaps before it answers, and inspect its own draft afterwards - through the eyes of the person who will read it.

When should I use this technique?

Before anything that goes out of your hands - a message to an upset customer, notes you will revise from, a chapter you are about to send to a reader - and whenever your first instinct is to type "is this good?"

How do I apply it in Gemini?

The prompts came from here and need only tightening; what this platform adds is a second opinion that is not the model's own. On a factual answer, "double-check response" has Search go through the statements one by one and mark what it can and cannot support - which is a genuinely different kind of review from asking the model to grade itself, and the right tool for revision notes full of section numbers. Then there is the quiet last step the chapter is really about: export the final version to Google Docs and read it yourself, on a page, away from the chat. Reviews done in the scroll of a conversation are not reviews.

What are the limits of this technique in Gemini?

Do not mistake double-check for verification. It shows whether Search can find support for a statement, which is not the same as the statement being right: an unsupported line may be perfectly correct, and a supported one may be sourced from a coaching blog repeating the same error. For tax matters, follow the links and land on the CBIC or the Act itself. It is also no use at all on the two creative examples - whether a sentence shows a feeling instead of telling it, or whether an apology sounds cold, is not a claim the web can confirm. And grounding a review of your own memoir will drag in generic writing advice. Say plainly: do not search, review only what I have pasted.

How do I improve the answer after the first try?

With this principle, the loop begins before you even press Send. Read your prompt once as if you were the AI: is anything ambiguous or missing? Then, when the first answer arrives, ask the AI to review it - "What is weak or uncertain in your answer?" or "Check this against my original request." Read its critique, accept what is useful, overrule what is not, and ask for a revised version. A single review round usually lifts the quality more than any other follow-up.

Loop it

With this principle, the loop begins before you even press Send. Read your prompt once as if you were the AI: is anything ambiguous or missing? Then, when the first answer arrives, ask the AI to review it - "What is weak or uncertain in your answer?" or "Check this against my original request." Read its critique, accept what is useful, overrule what is not, and ask for a revised version. A single review round usually lifts the quality more than any other follow-up.

This lesson on the other tools

The technique is the same; the way you write it is not. ChatGPT, Claude, Perplexity.