GH-300 Official Certification Question Bank (125 Questions)
125 curated questions sourced directly from official Microsoft Learn study domains, GH-300 exam prep guides, and real test sets covering all core competency areas of the GH-300 exam.
Q121 [Scenario]: A developer has personal custom instructions in VS Code user settings specifying tabs for indentation, but the project's .github/copilot-instructions.md specifies spaces. Which rule takes precedence during Copilot Chat generation?
Repository-level instructions (.github/copilot-instructions.md) override personal user settings to ensure project consistency across all contributing team members.
- Repository instructions take precedence over individual user settings.
- Ensures consistent output styling across team members working in a shared repo.
When a conflict exists, which instruction source takes precedence in Copilot Chat?
Q122 [Scenario]: An enterprise trains a Fine-Tuned Custom Model on private proprietary repositories. A developer worries that contractors with Copilot access in other divisions might see proprietary code completions. How is access isolated?
Custom Models are scoped strictly at the Organization / Enterprise policy level. Admins control exactly which teams or user groups have access to query the fine-tuned custom model via seat settings.
- Custom model access is controlled via granular organization settings.
- Prevents unauthorized organizational groups from querying specialized models.
How do enterprise administrators restrict who can query fine-tuned custom models?
Q123 [Scenario]: A developer wants Copilot Chat to parse a messy text file into a JSON array. Asking 'Convert this to JSON' returns inconsistent schemas. How can Few-Shot prompting fix this?
The developer should supply 1 or 2 concrete examples inside the prompt showing exact input text samples alongside the desired target JSON structure before providing the full text file.
- Few-Shot prompting uses concrete input/output pairs to enforce output formatting.
- Drastically reduces structural formatting errors.
What prompting technique resolves inconsistent JSON formatting by providing sample input/output pairs?
Q124 [Scenario]: An engineering manager is traveling and receives an urgent notification to review a critical hotfix PR. They use GitHub Mobile. How can Copilot assist directly on their phone?
The manager can open the PR in the GitHub Mobile App, invoke Copilot Chat to summarize the changes, ask questions regarding potential side effects of the diff, and review the code before approving.
- Copilot Chat in GitHub Mobile enables code review workflows on mobile devices.
- Allows reviewing PR summaries and querying changesets on the go.
What mobile tool allows managers to query Copilot Chat for pull request reviews while traveling?
Q125 [Scenario]: A developer types a standard 5-line algorithmic function. Copilot suddenly stops generating inline ghost text. What filter likely triggered and why?
The Duplication Detection Filter was triggered because the candidate completion matched a contiguous sequence (~150 characters) of existing public code on GitHub, causing Copilot to suppress the suggestion.
- Duplication detection suppresses suggestions matching contiguous public code blocks (~150 chars).
- Protects against unintentional verbatim public code insertion.
Why did Copilot suppress the inline completion during the typing of a common algorithm?