Free Exam GH-300: GitHub Copilot Practice Questions

Sample Practice Questions

Try these 10 sample questions from the Exam GH-300: GitHub Copilot practice bank. Each includes the answer and explanation. Continue with the interactive practice above to review the full question set.

1. 1Which responsible AI principle addresses avoiding different impacts on similarly situated groups?

   1. A. Transparency
   2. B. Fairness
   3. C. Accountability
   4. D. Inclusiveness

   Answer: B

   Explanation: The module states, "AI systems should treat everyone fairly, avoiding differential impacts on similarly situated groups." This is the fairness principle; inclusiveness focuses on access and empowerment.

   Reference: Microsoft and GitHub's six principles of responsible AI — Fairness: AI systems should treat all people fairly

2. 2A developer has a GitHub account with an existing Copilot subscription. How should they begin setting up Copilot in VS Code?

   1. A. Open a repository owned by the subscribed account while using another sign-in
   2. B. Sign in to a separate personal account and expect the existing license to transfer
   3. C. Set the Git commit email to the subscribed account without changing the sign-in
   4. D. Select Use AI Features and sign in with the account holding the subscription

   Answer: D

   Explanation: The guide says, "Hover over the Copilot icon in the Status Bar and select Use AI Features." It also states, "If you already have a Copilot subscription for your account, VS Code will use that subscription." Copilot access follows the authenticated account, not repository ownership or commit metadata.

   Reference: Set up GitHub Copilot in VS Code — Set up GitHub Copilot

3. 3Before a Copilot Chat request reaches the language model, how is the user's prompt handled?

   1. A. It is sent alone, with repository and chat context discarded
   2. B. It is pre-processed and combined with available contextual information
   3. C. It is compiled into executable code regardless of the input format
   4. D. It is added to the model's training weights before inference

   Answer: B

   Explanation: The card states, "The user's prompt is pre-processed by the system, combined with contextual information (for example, the current repository, open files, or chat history), and sent to a large language model." This is runtime input construction, not retraining.

   Reference: Application card: GitHub Copilot Chat — 6. Performance / Input processing

4. 4Which prompt organization follows GitHub's guidance for stating a goal and its requirements?

   1. A. Give the broad goal, then specific requirements with important ones first
   2. B. List minor formatting preferences before mentioning the task's purpose
   3. C. Supply a broad goal and leave required behavior for Copilot to infer
   4. D. Place lower-priority details first and state the main objective at the end

   Answer: A

   Explanation: The guide says, "first give Copilot a broad description of the goal or scenario. Then list any specific requirements. Put your most important requirements first." This makes the goal and priorities explicit.

   Reference: Prompt engineering for GitHub Copilot — Start general, then get specific

5. 5Which request asks Copilot to refactor an existing implementation?

   1. A. Change the business rules so previously rejected inputs are accepted
   2. B. Improve the code's structure while retaining its existing behavior
   3. C. Replace the feature with a new workflow that returns different results
   4. D. Remove an existing validation requirement to simplify execution

   Answer: B

   Explanation: The module describes refactoring assistance as "Handles sophisticated code transformations while preserving functionality". Changing requirements or accepted results is a functional change, rather than the stated refactoring goal.

   Reference: Boost developer productivity with AI — Orchestrated AI workflows / Advanced reasoning capabilities / Complex refactoring assistance

6. 6A repository collaborator has the Maintain role. What can they do with that repository's Copilot content exclusion settings?

   1. A. Edit repository exclusions but not view inherited ones
   2. B. Edit inherited organization exclusions for that repository
   3. C. View the settings but not edit them
   4. D. Edit exclusions after changing their local editor settings

   Answer: C

   Explanation: The guide says Maintain-role collaborators "can view, but not edit, content exclusion settings for that repository." Local editor settings do not grant administrative permissions.

   Reference: Excluding content from GitHub Copilot — Who can use this feature?

7. 7An AI-assisted development process must handle unexpected conditions without causing unintended harm. Which responsible AI principle most directly addresses this requirement?

   1. A. Reliability and safety
   2. B. Transparency
   3. C. Fairness
   4. D. Accountability

   Answer: A

   Explanation: Under reliability and safety, the module says, "These systems need to function as designed, respond safely to unexpected conditions, and resist harmful manipulation." The requirement concerns dependable and safe behavior.

   Reference: Microsoft and GitHub's six principles of responsible AI — Reliability and safety: AI systems should perform reliably and safely

8. 8With the default VS Code keybinding, how do you accept a displayed Copilot ghost text suggestion?

   1. A. Press Escape
   2. B. Press Tab
   3. C. Press Ctrl+F
   4. D. Press Shift+Tab

   Answer: B

   Explanation: The inline suggestions guide states, "Accept a suggestion by pressing the Tab key." The other keys are not the documented acceptance binding.

   Reference: Inline suggestions from GitHub Copilot in VS Code — Ghost text suggestions / Getting your first suggestions

9. 9When enabled Copilot Chat web search is used, what information flow does GitHub describe?

   1. A. The browser sends only its locale, without prompt-derived content
   2. B. The search engine receives only the finished answer after generation
   3. C. The prompt and context generate a query sent to Bing
   4. D. The model searches solely within its training weights offline

   Answer: C

   Explanation: The card says, "the content of your prompt, along with additional context, is used to generate a Bing search query on your behalf." Search therefore involves prompt-derived information leaving the conversation for the search service.

   Reference: Application card: GitHub Copilot Chat — 7. Limitations / Leveraging web search

10. 10You give Copilot several input-output examples before requesting a new result. Which prompting approach are you using?

    1. A. Zero-shot prompting, because examples replace instructions
    2. B. Model fine-tuning, because examples permanently update weights
    3. C. Few-shot prompting, because examples demonstrate the desired behavior
    4. D. Output validation, because examples prove the generated result is correct

    Answer: C

    Explanation: Microsoft explains that few-shot learning involves "including one or more examples of the desired behavior of the model, typically by including input and output pairs." These examples condition the current inference; prompts without examples are zero-shot.

    Reference: Prompt engineering techniques — Basics / Examples