AI Software Testing: ChatGPT & GitHub Copilot for QA

Discover how ai and its subset ml differ from generative ai and llms, with real-world examples from ChatGPT and Copilot, and see how ai generates test cases and bug descriptions.

Explore how large language models like ChatGPT work, focusing on transformers attention and fine tuning, to improve qa results such as generating test cases, automation code, and bug report summaries.

Master tokens power GPT-4’s language tasks, balancing prompts and the context window to avoid token caps. Learn to craft precise prompts for QA artifacts like test cases and bug reports.

Demonstrate how text is tokenized and counted with a tokenizer, examining words, punctuation, spaces, and contractions, then master token budgeting for efficient prompts and QA tasks.

Compress prompts by removing fluff and preserving meaning, aiming to minimize token and character counts. Share results in a competition by posting the new prompt and totals in the comments.

Understand AI hallucinations as predictions that sound plausible but are wrong, and learn to avoid them with clear prompts, examples, and thorough QA review of test scripts and logs.

Master prompting for qa by crafting precise, context-aware prompts and applying styles like instructional, step-by-step, role-based, and refinement. Generate tests, summaries, and structured outputs for qa tasks.

Iterative prompting refines clarity, expands scope, and debugs AI output in QA workflows. Apply small, layered steps with precise constraints and concrete examples to improve results.

Practice safe prompting and learn what not to share with AI tools, from PII and production data to internal code and credentials, ensuring security and compliance.

Explore how prompts gain a skeleton through two practical frameworks, Clear and Race, to ensure consistency, reuse, and faster validation by defining context, action, constraints, and expected output.

Master prompting techniques to turn vague requests into precise automation test instructions, using step-by-step prompts with Playwright JS login tests, data test IDs, and code-only outputs.

Tour the ChatGPT web layout, learn to start new chats, switch models (auto, fast, thinking, legacy GPT‑4), and access the library, GPT store, Codex, and Sora.

Configure custom instructions in ChatGPT to set your name, role, and preferences, then tailor personalization with a chosen personality and traits for QA workflows.

Create your own personalization settings in ChatGPT by filling in custom instructions with your desired personality and other details, then share and update based on peer feedback.

Leverage ai prompts to accelerate test case design, draft edge cases, and extract tests from long specs or Jira tickets, while you review, validate, and document clear results.

Apply equivalence partitioning and boundary value analysis to generate focused test cases, identifying valid/invalid partitions and boundary edges for passwords, ages, promo codes, and order totals with ChatGPT.

Explore negative testing and edge cases with prompts, generating invalid credentials, empty fields, SQL injection risks, and system behavior scenarios to build a solid testing baseline.

Discover how AI-assisted requirement analysis identifies ambiguities and risks in agile stories, then translates them into test coverage areas and mini test plans for robust QA.

Use AI to generate realistic but safe synthetic data for QA. Test edge cases and fuzzing with long strings, international characters, and malicious inputs.

See how ai powers qa automation with up-to-date environment setup guides tailored to your OS and tools, and how ChatGPT generates a full Playwright test environment plan.

Design better test cases for a web app's reset password flow using ai, covering happy path and negative scenarios, including 30-minute link expiry and errors for invalid or unregistered emails.

Explore how AI augments exploratory testing by sparking ideas, generating test charters, and turning notes into structured test cases, helping uncover edge cases and usability issues while you validate functionality.

Explore exploratory testing to reveal edge conditions and risks beyond requirements. Let AI brainstorm scenarios for a retail search with filters—category, price range, availability, brand—and turn ideas into test charters.

Transform messy exploratory testing notes into clear observations, possible issues, and structured bug reports with AI. Draft debriefs and follow-up test cases for regression, using login-page scenarios as examples.

Leverage AI as a writing assistant to craft clear bug reports, generate reproduction steps, assess severity, and speed triage, while reviewers ensure accuracy and avoid hallucinations.

Use AI to turn error logs and tester notes into professional bug reports for developers, featuring clear reproduction steps and automated severity and category triage.

Use ai as a partner to brainstorm exploratory test ideas for a product search with filters, structure them into charters with goals, risks, and data variations for 30-minute session.

Explore how GitHub Copilot serves as an AI-powered coding assistant for QA engineers, generating test scripts, locators, and assertions inside Visual Studio Code to accelerate testing.

Explore the Visual Studio Code workspace, including the activity bar, explorer, search, source control, run and debug, and extensions. Create folders and files, manage tabs, and use the integrated terminal.

Install and activate GitHub Copilot in Visual Studio Code, sign in, verify readiness, then use chat and inline chat with smart actions to generate code, commit messages, and PR descriptions.

Install node.js from nodejs.org to run JavaScript outside the browser, verify with node -v and npm -v, and prepare Copilot and playwright for JavaScript projects.

Install and verify git on Windows or Mac. Use VSCode terminal, git bash, or PowerShell, and confirm with git --version; Mac users may brew install git.

Explore GitHub Copilot inside Visual Studio Code, using code completions, context-based edits, and modes (ask, edit, agent) to generate, test, explain, and rename files.

Explore how session history preserves context across chats, automatically continuing test-case numbering and enabling rapid generation of login, password reset, remember me, and profile settings test suites.

Boost Copilot by adding a repository-level instructions file in a hidden .github folder, teaching our tech stack and QA rules so it follows our standards in chat and code review.

Discover how QA testers use Copilot with Visual Studio Code to generate Playwright tests, create test scenarios, improve test IDs, mock APIs, and write bug reports.

Turn plain English into JavaScript with Copilot in Visual Studio Code, create a folder, run npm init -y, paste the prompt, build a CLI app, fix errors, and report results.

Explore Copilot in agent mode and its tool sets, from reading and writing files to terminal commands, web search, and external services, through a practical requirements.md analysis workflow.

Navigate the gear menu to edit requests, choose thinking styles, and select the chat plan agent model and queue behavior for your QA conversations.

Explore vibe coding with Copilot as you build three qa tools: a cli, a node service with health and test data users endpoints, and boundary buddy for equivalence partitions.

Learn to build a test data generation cli in node that outputs synthetic users and fuzz strings, guided by copilot with inline comments, including setup, debugging, and usage.

Build a tiny lorem ipsum cli using Visual Studio Code, npm, and Copilot to generate code from a full comment block, then run commands and share results.

Design a vibe coding mini test tool service with an Express Node.js app, seeded RNG, deterministic fake users, and csv/json endpoints, plus nodemon auto reload.

Build a tiny lorem ipsum service on GitHub with Copilot to create two endpoints, health and lorem text, then run and verify locally.

build a no-backend web tool that generates equivalence partitions and boundary values for numeric ranges and string lengths, using Copilot to scaffold HTML, CSS, and JS.

Build a single-page web app with HTML, CSS, and JavaScript to generate deterministic lorem ipsum text, with units, counts, copy or download options, and local storage to remember settings.

Boost test automation with GitHub Copilot through prompts to speed tests, expand UI, API, and edge-case coverage, and pair automated scripts with manual exploratory testing.

Explore Chrome devtools essentials: inspect the dom, read the console, and monitor network activity. Learn stable selectors using id or class, and test them with document.querySelector to empower Copilot-driven tests.

Write and run your first playwright ui test with Copilot by configuring base URL, creating a monster spec, and validating title, heading, and button with a full page screenshot.

Learn to automate playwright tests with Copilot by adding run scripts for headless, headed, and UI modes, and generate and view reports with combined commands.

Create and integrate page objects with Copilot to replace inline selectors, building a login page object with locators and methods like fill username, fill password, and login.

Generate a fixtures file with login data and selectors using copilot, and refactor tests to import fixtures and move selectors into the pageobjects. Retest to confirm all tests pass.

Use Copilot to generate and run Playwright API tests in Visual Studio Code, configure the base URL for the Poké API, and validate endpoints with 200 and 404 responses.

Learn to generate Playwright tests with Copilot by crafting prompts, using API docs to pick resources, and creating UI tests focused on the drop down and inputs.

Discover what an AI agent is, how it plans steps and uses tools to complete multi-step tasks. Contrast agents with LLMs, noting memory, verification, and bounded autonomy for repeatable workflows.

See how an AI agent extends beyond chatbots to plan and use tools. The lecture demonstrates ChatGPT triaging error logs, summarizing patterns, drafting bug reports, and proposing tests.

Demonstrates using ChatGPT as an AI QA agent to explore a login page, generate a page object model, selectors, manual test cases, and Playwright code for automated testing.

Use the ai agent inside Copilot to build a Node.js test results summarizer that reads Xml reports, aggregates passes, failures, and skips, and outputs summary md, json, and interactive dashboard.

Practice turning ChatGPT into a small QA agent that plans work, uses tools, iterates, and delivers real artifacts, with track A and track B options.

Learn how to create custom agents that encode context once for all sessions, using VS Code, GitHub Copilot, and ChatGPT, with QA test writer instructions and Jest tests.

Demonstrates a multi-step QA agent workflow that starts with chat-based prompting in ChatGPT to generate test cases for an API, then transfers prompts into the IDE to implement tests.

Master GitHub Copilot CLI, a terminal tool that converts plain English into shell commands. Install with npm, set up, and log in via for-login for QA workflows.

Navigate Copilot's interactive, plan, and autopilot modes to streamline playwright automation, using slash commands, question mark and help prompts, and exclamation-run standard commands for efficient qa testing.