# CO-STAR Method Prompting for Testers | Write Perfect AI Prompts (Step-by-Step) ## Метаданные - **Канал:** Automation Step by Step - **YouTube:** https://www.youtube.com/watch?v=LE578lYq2iw - **Дата:** 06.05.2026 - **Длительность:** 4:53 - **Просмотры:** 1,166 - **Источник:** https://ekstraktznaniy.ru/video/50056 ## Описание Still struggling to get useful results from AI? In this video, you’ll learn the CO-STAR Prompting Framework — a simple and powerful way to write clear, structured prompts that actually work in real testing scenarios. Instead of guessing what to ask AI, you’ll learn how to guide it step-by-step using: CO-STAR 00:00 Introduction 00:55 Context 01:19 Objective 01:47 Style 01:58 Tone 02:12 Audience 02:41 Response 03:12 Final Prompt using Co-Star method 03:31 Demo of Co-Star with Persona We walk through a real example (Login with OTP) and improve a bad prompt into a professional, job-ready one—live. 👉 You’ll see how real SDETs use ChatGPT to: Generate structured test cases Improve output quality Save time in daily tasks This is a hands-on, practical session designed for beginners and working testers. 👉 By the end, you’ll stop writing random prompts and start writing smart prompts that deliver results 🚀 CO-STAR prompting, prompt engineering testing, AI testing prompts, ChatGPT for ## Транскрипт ### Introduction [] Let's say you want to take help of AI and you want to generate some test cases for login with OTP. So, you will write a prompt and you will tell AI, let's say a very basic prompt, generate test cases for login. So, let's say I go to some AI chat, chat GPT, Gemini, etc. And I will say here generate test cases for login with OTP and I will hit enter and you will see it will give us a response and the response will be very generic, very vague, not exactly what maybe I may be needing for my exact project or work because I have not given all the details. So, this may not work very well. So, let's build this prompt step by step using the Co-Star framework. So, what's the Co-Star framework? So, here every letter has a meaning. So, let's see. Let's start with C. C stands for ### Context [0:55] context. So, what exactly you need, what exactly you are testing. So, this will give a very, very good starting point to AI. So, here our context is we want to test login with OTP. Okay? So, let's keep it very simple. I will write the final prompt in a moment. Let's first discuss and see what exactly this means. Next is O ### Objective [1:19] objective. So, what do you exactly want from AI? So, let's be very clear and very specific. It will help AI a lot. Here we want to generate five test cases. Now, if you have seen the earlier session, it is something uh close to the constraints. Here we are telling that we don't want any number of test cases, but only five test cases. And it can be objective can differ for different scenarios. Here our objective is to generate five test cases. ### Style [1:47] Next is S for style. Here you can say how should the output look, what format do you want. So, we want here a BDD format. Then T stands for tone. Now, here you ### Tone [1:58] can tell AI do you want detailed, concise, or any kind of tone that you want the AI to take care. Here I will say I want very concise output. ### Audience [2:12] output. Then A is for audience. This is very important. If you tell AI who is the audience for this response, it will create the response as per that audience and it will be very, very useful. So, here we are getting it created or generated for testers. Now, in place of testers, if you say QA or maybe developers or some business people, the output will differ and therefore it is very important to tell the who will be using the output or who is the audience. Then last, R is for ### Response [2:41] response. Now, here we can say how we want the response. Are there any more constraints or maybe format or structure? So, here we can say we want a table format and no explanation. So, these things, the Co-Star framework, if we apply and create our prompts using all these things, context, objective, style, tone, audience, response, it will make our prompts so much better and will give us very useful and efficient output and responses. So, this is how our final ### Final Prompt using Co-Star method [3:12] prompt will look like. We have added all these elements of the Co-Star framework. Now, in addition to this, we can also add persona or role prompting that we have learned earlier where we can say that act as a senior estate or senior QA engineer. So, let me just go here and say ### Demo of Co-Star with Persona [3:31] I will say you are a senior QA engineer and then I will just paste what we have created our prompt using the Co-Star framework. Now, this thing, when we give a give our persona or role, it will help AI a lot because now it will have to search within some limited data sets and not all the data sets. So, let us just run and check this. And you will see the difference in the responses. So, you can see now it is giving us a table format and very specific very accurate, exactly to what we need, five test cases with all the details we need. So, it will be very useful. We can directly use it. So, here you can see this is how this Co-Star framework can help us a lot. And whenever you are chatting with AI chats or you are trying to get these things for you with the help of AI, you can apply this Co-Star technique. You can see you can exactly say this is my context, this is my objective, this is the style, tone, or whatever you need along with the persona or role and you will see you will get so much better output and useful outputs. I hope this was useful. I will see you soon. Thank you for watching and never stop learning.