(Make sure you get version 3.0 when its out. Leave your email address here: https://prove.guru/swipefile) 👈🏻 Also please share this link to others too! 🙏🏻

<aside> ⚠️ “AI won’t be taking our jobs in the near future, BUT people who know how to use AI will.” -Scott Galloway, Nordic Business Forum

</aside>

⚙️ Introduction

Welcome to Version 2.0 of "85 AI Prompts For Software QA Professionals." This resource is designed to provide software QA professionals with a wide range of prompts to enhance their skills and productivity. Whether you're a professional software tester, an agile coach, a senior software developer, or any other role within the software testing field, these tools are here to inspire and challenge you.

In this document, you'll find a framework to create AI prompts that help you cover various aspects of software testing, including analyzing screenshots, reviewing test automation scripts, improving resumes, creating content calendars, managing calendars effectively, and much more. To get started, explore the "13 Ideas To Get You Going" section for ready-to-use prompt ideas.

<aside> ⚡ (Powered by PROVE - www.testguruacademy.com)

</aside>

The TestGuru AI Framework: PromptPro3000

To create effective AI prompts to aid your QA process, it is highly important to tailor the role, task or outcomes, format, tone, and thinking framework to suit your specific needs and objectives. We have created the ultimate framework to maximize your productivity: The Testguru AI Framework: PromptPro3000.

"Act as a [role], [task or outcomes], in a [format] using [tone, thinking framework or language]. The main objective is to [objective], don't [restrictions], here is an example [example]."

Screenshot 2024-02-01 at 13.51.03.png

Examples below!

▶️ 13 Ideas To Get You Going

<aside> 💡 These examples can be used to spark your creativity when creating prompts for your own specific context.

</aside>

👀 “You are a professional software tester. Next, I’m gonna show you a screenshot of an app I’m testing. Can you spot any bugs or potential problems there?”

🕵🏻‍♀️*“You are an agile coach. Answer my questions like a coach would using the Six Thinking Hats”*

🕷️*“Act as a TestguruGPT, Given a list of reported bugs in software login function, rank them based on severity, user impact, and complexity of the fix”*

🤖*“Play the role of a senior software developer. Review the following test automation script and suggest optimizations for efficiency and coverage”*

🔍*“Review this resume like you were a HR consultant. Make corrections to spelling and suggest improvements that will get me an interview”*

📌“Read this logfile of a series of Rest API transactions and hilight anything that seems out of the ordinary”

↔️*"What are the advantages and disadvantages of [option A] compared to [option B]? I need insights to make an informed decision.”*

📈*"I work in software testing. I need a list of advanced prompts that will enable someone in my occupation to be more productive and work faster”*

📲*"Help me create a 7 day content calendar for linkedin. My niche is software testing and I want to be noticed by potential new employers”*

📆*“I want to learn how to manage my calendar better. Can you guide me on the best approach and resources, and produce a 30 day plan to take me from beginner to advanced.”*

🤯*“I'm looking for fresh ideas on how to convince my bosses to do more exploratory testing. Can you help me brainstorm 7 different approaches to do this?”*

💀“I need an agenda for a post-mortem meeting where we discuss a show stopper that got into production, keep it a maximum of 500 words, and give an emoji for each title on the agenda”

📝*“Write a funny invitation letter for our team building day. Write it like Mark Manson does.”*

🕷️ A Practical Example

<aside> 💡 In this experiment, ChatGPT was used to analyze a screenshot for bugs.

</aside>

The Prompt:

“You are a professional software tester. Next, I’m gonna show you a screenshot of an app I’m testing. Can you spot any bugs or potential problems there?”

The Screenshot:

Triangle App Screenshot.PNG

Example result in 3/2024:

ChaGPT demo 02-2024.mov

ChatGPT understands the concept of bugs (inconsistent behaviour between two reference points) and sums up ideas and results in highly usable way.

What next?!

Consider exploring GPT4 API. Use that to automatically upload every screenshot you take for analysis and collecting reports as txt files for each screenshot. Now you either check them yourself or have GPT4 sum up the content of all text files after a test run.

Action item: Today, there are simple no-code tools that allow you to automate big chunks of the workflow

Here is an example:

https://www.loom.com/share/0678bb9fd2004d1cbbeb7eee5df6a5a9

For more information check the tools: https://www.make.com/ and https://notion.so

🔮Future: OpenAI Sora Examples

In FEB’24 OpenAI announced it’s latest model with the ability to create videos out of text based prompts.

This means that very soon AI systems have a high level capability in making sense of videos and also live video feeds. Hence, the screenshot example will soon extend to real time application usage.

The prompt: A stylish woman walks down a Tokyo street filled with warm glowing neon and animated city signage. She wears a black leather jacket, a long red dress, and black boots, and carries a black purse. She wears sunglasses and red lipstick. She walks confidently and casually. The street is damp and reflective, creating a mirror effect of the colorful lights. Many pedestrians walk about.

tokyo-walk.mp4

More OpenAI examples about Sora in practice: https://openai.com/sora

When MML can create videos that make sense, the next evolution-step is the ability to make sense of videos. Short clips at first, then live video feed.

With the ability to make sense of a video feed, soon comes to ability to interact with the environment.

Figure 01 Robot combined with ChatGPT demonstrates this.

https://www.youtube.com/watch?v=Sq1QZB5baNw

🏗️ 13 Building Blocks For Your Prompts

<aside> 💡 These elements will get you started. Experiment, research, imagine more as you start playing around with AI assistants.

</aside>

  1. Assigning a Role: Specify a role for the AI to adopt, such as a teacher, historian, scientist, or fictional character. This can shape the AI's tone and perspective.
  2. Requesting Specific Outcomes: Ask for particular types of outcomes, like explanations, stories, summaries, dialogues, lists, or plans or assign a task.
  3. Applying Thinking Frameworks: Request the AI to use certain thinking frameworks like critical thinking, creative thinking, problem-solving, or ethical considerations.
  4. Asking For Specific Format: Usually AI responds in text, but can also generate images or certain file types. Read on to find out more about the potential.
  5. Follow Examples: Give AI examples of what you mean, instruct to do more of the same or just follow the similar style, expand on your original examples.
  6. Setting Constraints or Challenges: Impose certain constraints or challenges, like answering in a limited word count, using only questions, or avoiding certain words.
  7. Mimicking Famous Personalities: Ask the AI to respond as if it were a famous person, historical figure, or renowned expert in a field. This can add a unique flavor to the response.
  8. Using Specific Genres or Styles: You can ask for responses in the style of a particular genre (e.g., noir, fantasy, scientific) or writing style (e.g., formal, conversational, poetic).
  9. Integrating Emotion or Tone: Specify a tone or emotional angle for the response, like humorous, serious, empathetic, or sarcastic.
  10. Incorporating Analogies or Metaphors: Ask the AI to use analogies or metaphors to explain concepts, making them more relatable or easier to understand.
  11. Using Real Life Examples: Ask the AI to back up the work with some real life examples of what you are working on.
  12. Addressing Hypothetical Situations: Pose hypothetical scenarios and ask the AI to explore or solve them.
  13. Combining Multiple Elements: Mix and match the above elements to create complex and interesting prompts (e.g., "Explain quantum physics like you are a poet" or "Give a humorous summary of today's news").

<aside> 💡 Remember, the effectiveness of these elements can depend on how they are combined and the complexity of the request. The AI's responses are generated based on its training and the specific input it receives, so clear and creative prompts can lead to more engaging and accurate outputs.

</aside>

🎯 24 Outcomes To Ask For: Begin With The End In Mind

<aside> 💡 To get the best results you need to start with the end result and ask yourself the question: What is the goal that you are trying to achieve? Thus, Begin With The End In Mind!

</aside>

  1. Analyzing Screenshots: Assisting in interpreting and diagnosing issues from screenshots of the application under test.
  2. Analyzing Test Results: Providing insights into test results, identifying patterns or anomalies, and suggesting areas for improvement.
  3. Analyzing Bug Reports: Helping to triage and prioritize bug reports based on severity, impact, or other criteria.
  4. Analyzing Parts Of The Source Code: AI can provide support in reviewing source code by offering insights on coding standards, identifying potential coding errors or bad practices, and suggesting improvements for code readability and maintainability. While it can't understand or interpret code as a human would, it can offer guidelines, best practices, and general advice on common programming patterns and potential pitfalls. This could be particularly useful for new developers or in educational contexts where learning about coding standards and clean code principles is essential.
  5. Log File Monitoring and Analysis: AI can assist in interpreting and analyzing log files generated during software testing. This includes identifying error patterns, highlighting critical issues, filtering noise from useful data, and summarizing key events. The AI can help in pinpointing anomalies or trends that might indicate underlying system problems or bugs. Additionally, it can assist in translating technical log details into more understandable terms for reporting and communication purposes. This application is particularly useful in continuous integration/continuous deployment (CI/CD) environments, where real-time monitoring of logs is crucial for maintaining system health and performance.
  6. Suggesting New Test Ideas: Generating creative and diverse test scenarios and cases that might not have been considered.
  7. Providing Examples/Cases/Stories of Similar Bugs in Other Organizations: AI can draw upon its vast knowledge base to provide historical examples, cases, or stories of similar bugs that have occurred in other organizations or software projects. This can include:
    1. Describing how similar bugs were identified and resolved.
    2. Outlining the consequences these bugs had on those organizations, such as security breaches, data loss, financial impact, or reputational damage.
    3. Detailing the lessons learned and best practices developed as a result of these incidents.
  8. Suggesting Root Causes for Issues: Offering hypotheses for the root causes of identified bugs or issues based on the data provided.
  9. Give Coaching: Offering guidance on best practices, methodologies, and strategies in software testing.
  10. Improving Resume/CV: Making improvements to your resume, offering guidance and suggesting additions.
  11. Preparing Reports: Assisting in the creation of detailed test reports, summaries, and documentation.
  12. Writing Compelling Emails: Drafting clear and effective communication for stakeholders, such as updates on testing progress or bug reports.
  13. Generating Test Data: Creating mock data sets for testing purposes, considering various scenarios and edge cases. Here is an example list of filetypes that ChatGPT can support as of early 2024:
    1. Text Files (.txt): Simple text files for basic data needs.
    2. CSV Files (.csv): Comma-separated values files for structured data, useful for database, spreadsheet, or data analysis tasks.
    3. JSON Files (.json): JavaScript Object Notation files for structured data, commonly used in web applications and for configuration files.
    4. XML Files (.xml): Extensible Markup Language files for structured data, often used in web services and configurations.
    5. Markdown Files (.md): Markdown language files for formatted text, useful for documentation or README files.
    6. HTML Files (.html): Hypertext Markup Language files for web page structures, useful for web testing.
    7. SQL Script Files (.sql): SQL code for database querying and manipulation, useful for preparing database test environments.
    8. YAML Files (.yaml or .yml): YAML Ain't Markup Language files for configuration data, often used in software deployment and development environments.
    9. Configuration Files (.conf, .config, etc.): General configuration files for various software applications.
  14. Reviewing Test Scripts and Plans: Providing feedback on test scripts and plans to ensure coverage and effectiveness.
  15. Writing Scripts For Test Execution Or Automation: AI can assist in drafting code snippets, test cases, or even entire scripts. Here is an example list of languages that ChatGPT supports Early 2024:
    1. Python: Widely used for test automation, especially with frameworks like Selenium, PyTest, or Robot Framework.
    2. JavaScript: Commonly used for web testing, particularly with tools like Jest, Mocha, Cypress, or Puppeteer.
    3. Java: A popular choice for test automation in enterprise environments, often used with Selenium, TestNG, or JUnit.
    4. C#: Used in test automation, particularly in .NET environments, with tools like NUnit or SpecFlow.
    5. Ruby: Known for its readability and ease of use, Ruby is used with tools like Capybara and RSpec.
    6. PHP: Less common, but can be used for testing with tools like PHPUnit.
    7. Shell Scripting: Useful for automating simple tasks in Unix/Linux environments.
    8. Groovy: Often used for scripting in Jenkins pipelines and for API testing with tools like SoapUI.
    9. Kotlin: Gaining popularity for Android app testing.
    10. Swift: Used for iOS app testing.
  16. Automating Routine Queries: Answering frequently asked questions about the testing process, tools, or methodologies.
  17. Training and Educational Support: Assisting in the training of new testers by providing educational content and answering queries.
  18. Collaboration Enhancement: Facilitating better communication and collaboration among testing team members.
  19. Integrating with Testing Tools: Offering insights or control commands for integration with various testing tools and environments.
  20. Performance Analysis: Assisting in the analysis of performance testing data to identify potential bottlenecks or optimization opportunities.
  21. Security Perspective: Providing a security-focused perspective in testing, identifying potential vulnerabilities or suggesting security test cases.
  22. Compliance and Standards Check: Helping ensure that testing processes and outputs align with industry standards and regulatory requirements.
  23. User Experience Feedback: Offering insights on user experience and usability aspects that can be tested or improved.
  24. Predictive Analysis: Using historical data to predict future testing needs, challenges, or outcomes.

<aside> 💡 By leveraging ChatGPT in these ways, software testing professionals can enhance their efficiency, creativity, and effectiveness in ensuring the quality and reliability of software products.

</aside>

🎭 18 Roles To Work With: Identity Calibrates Your Answers

<aside> 💡 To make the most of the powers of AI, you can calibrate the answers by making AI identify as a certain persona. The answers you get will be highly altered by the role you assign to the AI, so experiment away!

</aside>