Book companion · Edition 1

GitHub Copilot GH-300 Exam Study Guide

An Unofficial Guide to All Six Exam Domains

An unofficial study guide by Nex Arc Learning. Not affiliated with, endorsed by, or sponsored by GitHub or any exam provider. GH-300, GitHub and related names are trademarks of their respective owners.

About the book

This is the reader companion for GitHub Copilot GH-300 Exam Study Guide, edition 1, published by Nex Arc Learning as a Kindle ebook. The book prepares you for the GH-300 GitHub Copilot exam and follows the GH-300 exam guide, August 2026. It has one chapter for each of the six exam domains, in the order of the online study guide.

Buying the book does not unlock the premium practice exams or the online study guide.

What the book covers

  1. Chapter 1 Use GitHub Copilot features Use GitHub Copilot in the IDE · Use GitHub Copilot CLI · Use GitHub Copilot Features and Capabilities · Manage Organization-Wide Settings and Policies
  2. Chapter 2 Understand GitHub Copilot data and architecture Describe Copilot Data Handling and Flow · Assess the Copilot Lifecycle and Limits
  3. Chapter 3 Improve developer productivity with GitHub Copilot Enhance Productivity and Code Quality · Support Testing and Security
  4. Chapter 4 Configure privacy, content exclusions, and safeguards Manage privacy settings and exclusions · Apply safeguards and troubleshoot
  5. Chapter 5 Apply prompt engineering and context crafting Craft Effective Prompts · Engineer Prompts for Performance
  6. Chapter 6 Use GitHub Copilot responsibly Understand Responsible AI Principles · Validate and Operate AI Tools

Each chapter title links to the free overview of that domain in our online study guide.

Free practice questions

Check yourself with the free GH-300 sample questions. Pick an answer and select Check answer to see why each option is right or wrong. No account needed, nothing is stored.

Loading the practice questions…

Prefer exam conditions? Take the same questions in the free GH-300 practice exam.

Colour figures

On black-and-white Kindle e-readers the book's decision trees appear in greyscale. Here are all 20 overview decision trees in full colour, numbered as in the book. Follow a tree from the top question to the answer that fits your scenario.

Chapter 1: Use GitHub Copilot features

Figure 1.1 Decision tree: Use GitHub Copilot features
Is it about organization policies, seats, or the audit log, or a feature missing on GitHub.com? Yes Your organization's settings Manage Organization-Wide Settings and Policies No Does the work run through the copilot command in a terminal? Yes Your terminal Use GitHub Copilot CLI No Is it about enabling Copilot or triggering a completion or chat surface in the editor? Yes Your editor Use GitHub Copilot in the IDE No Is it about which capability does a job: change code, delegate work, extend and ground, review changes, converse and reuse? Yes Across surfaces, by job Use GitHub Copilot Features and Capabilities No Another domain covers it
Read this topic in the online study guide
Figure 1.2 Decision tree: Use GitHub Copilot in the IDE
Should Copilot choosefiles and edit them?YesAgent modeNoDoes it start from arequest in a chat box?NoYesCompletion at the cursor,or a predicted next edit?Is the request aboutselected code?Completionat the cursorPredicted next editGhost textNext editsuggestionsYesNoInline ChatAsk mode
Read this topic in the online study guide
Figure 1.7 Decision tree: Use GitHub Copilot CLI
Must work continue withyour machine off?Yes/delegate(Copilot cloud agent)NoIs a person present tosteer and approve?NoProgrammatic interface(copilot -p)YesMust the approach be agreedbefore code changes?YesPlan mode(review the plan first)NoTrusted task in an isolatedenvironment, no per-step approval?YesAutopilot mode(full permissions)NoDefault mode(prompts for actions not already approved)
Read this topic in the online study guide
Figure 1.13 Decision tree: Use GitHub Copilot Features and Capabilities
Does the taskchange code?YesNoRun in the backgroundon GitHub?YesCopilot cloud agentNoRequirementsstill open?YesPlan modeNoWant to choosethe files yourself?YesEdit mode(JetBrains IDEs)VS Code: edit modedeprecated (1.126);agent mode instead,Copilot picks filesNoAgent modeFind issues in apull request?YesCopilot code reviewNoOrient reviewers onwhat changed?YesPull request summaryNoContext must persistand be shared?YesCopilot SpaceNoNeed a liveexternal system?YesMCP server(in IDE agent mode)NoCopilot Chat withfocused context
Read this topic in the online study guide
Figure 1.19 Decision tree: Manage Organization-Wide Settings and Policies
Changing which features or Chat models members can use? Is the organization setting read-only? Enterprise owner changes the enterprise policy Copilot policies: Policies page or Models page Finding who changed a setting, policy or license, and when? Audit log: action:copilot (last 180 days) Adding or removing seats for users or teams? Organization settings or the Copilot user management REST API Missing feature: check plan, seat, policy, surface in order yes yes no no yes no yes no
Read this topic in the online study guide

Chapter 2: Understand GitHub Copilot data and architecture

Figure 2.1 Decision tree: Understand GitHub Copilot data and architecture
Is it about what the request carries,who receives it, or whether it iskept or used for training?YesDescribe Copilot DataHandling and Flowcontext, model host, trainingNoIs it about what the proxy ora content filter does?YesDescribe Copilot DataHandling and Flowproxy and filtersNoIs it about stage order oraccepting the output?YesAssess the CopilotLifecycle and Limitsstage order, your decisionNoIs it about why plausibleoutput can be wrong?YesAssess the CopilotLifecycle and Limitsthree sources of limitationNoOutside this domainManage privacy settings and exclusionsor Craft Effective PromptsEvery path: a filtered suggestion still needs your review
Read this topic in the online study guide
Figure 2.2 Decision tree: Describe Copilot Data Handling and Flow
Problem found beforethe model generates?Attempt to manipulatethe system?Harmful, offensive orinsecure output?YesNoProxyblocks itInput contentfilteringYesNoOutput safetyfiltersMatches public codewith Block set?YesNoWithheld bypublic code filterFormattedand shownYesNo
Read this topic in the online study guide
Figure 2.10 Decision tree: Assess the Copilot Lifecycle and Limits
Is it a terminal command?YesReview itbefore you run itNoIs it code for a security-sensitive application?YesSecure coding reviewplus code scanningand testsNoIs the output code?YesHuman reviewand testingplus lintingNo (prose)Human reviewcheck claims forhallucinationAlways: focused context reduces errors at the source
Read this topic in the online study guide

Chapter 3: Improve developer productivity with GitHub Copilot

Figure 3.1 Decision tree: Improve developer productivity with GitHub Copilot
Checking whether existing code is correct,secure or fast enough (a known slow pathincluded), not fixing a known error?No: improving the codeYes: checking the codeImproving the codegenerate, refactor, document,explain, debug a known error,sample data, modernizebaseline: set per taskBehavior checkedagainst a contract?YesTestsunit, integration, end-to-endbaseline: the contractcheck: run the testsNoExploitable weakness ininput, access or data?YesSecurity reviewbaseline: no exploitable weaknesscheck: code scanning, secretscanning, human reviewNo, slowPerformance reviewbaseline: the originaloutput and timingcheck: run it, compareoutput and timingEvery leaf: Copilot proposes, you check the proposal against the baseline
Read this topic in the online study guide
Figure 3.2 Decision tree: Enhance Productivity and Code Quality
Does the code give an erroror unexpected output?YesDebuggingNoNeed test or mock data thatfits a table schema?YesSample dataNoMoving code to another languageor a newer framework version?YesModernizationNoRestructure existing codewithout changing what it does?YesRefactoringNoNeed to understandunfamiliar code?YesExplanationNoNeed comments, docstringsor a README for the code?YesDocumentationNoWriting new code from a commentor a chat request?YesCode generationNoTests, security or performancelisted under Support testing and securityCheck Copilot's work: always validate the code it suggests
Read this topic in the online study guide
Figure 3.8 Decision tree: Support Testing and Security
Testing what thecode should do?YesWhich scope doesthe test exercise?one functionor classUnit testdependencies mockednamed componentsinteractingIntegration testname real vs mockeda user-visibleflowEnd-to-end testassert what the user seesNoExploitable weakness ininput, access or data?YesSecurity reviewthen code scanningNo: slow codePerformance reviewselect the hotspotThen: run the result, measure it when speed is the concern, and ask Copilot what is missing
Read this topic in the online study guide

Chapter 4: Configure privacy, content exclusions, and safeguards

Figure 4.1 Decision tree: Configure privacy, content exclusions, and safeguards
Is Copilot, or a control you set,not working as expected?YesTroubleshoot by how far the failure reachesService; client and sign-in (the Copilot extension,its editor settings, the signed-in account);network; or policy (an administrator settingsuch as content exclusion)Apply safeguards and troubleshootNoKeep chosen files out ofCopilot's context?YesContent exclusionCopilot Business or Copilot Enterprise planManage privacy settings and exclusionsNoChange where Copilot offersinline suggestions to you only?YesEditor settingsAffect only the developer who sets themManage privacy settings and exclusionsNoOwnership, training useor claim defense?YesCopilot termsDecided by the purchase channelManage privacy settings and exclusionsNoBlock a suggestion matching public code,or show it with a code reference?YesSuggestions matching public codeOutput control: Block or AllowApply safeguards and troubleshootNoBlock harmful, offensive or insecurecontent before it is shown?YesContent filtersOutput control: scans prompts and responsesApply safeguards and troubleshootNoNone of these five controls
Read this topic in the online study guide
Figure 4.2 Decision tree: Manage privacy settings and exclusions
Administrator rule: make Copilot ignorespecified files for other users too?YesNoCopilot Business orCopilot Enterprise plan?YesNoContent exclusionNot included inindividual plansChange where Copilot offersinline suggestions to you only?YesNoEditor settingsOwnership, training use,or claim defense?YesNoCopilot termsSee page: Applysafeguards and troubleshoot
Read this topic in the online study guide
Figure 4.7 Decision tree: Apply safeguards and troubleshoot
Copilot failing for many users at once? Yes Check GitHub's Status page No Does the message name a rate limit? Yes Wait for the reset, retry, or contact GitHub Support No Do only some files lack suggestions? Yes Content exclusion: check the status icon's diagonal line No Do the Copilot logs show proxy, certificate or timeout errors? Yes Fix proxy auth, allowlist or custom certificates No Update the extension, then sign in again and check the plan
Read this topic in the online study guide

Chapter 5: Apply prompt engineering and context crafting

Figure 5.1 Decision tree: Apply prompt engineering and context crafting
About what one prompt says, shows,attaches, or how much it asks?YesCraft EffectivePromptsrequest text, surrounding context,examples, task scopeNoDepends on earlier turns ora full context window?YesEngineer Promptsfor Performancehistory moves, context windowNoChoosing the next attempt after a miss,or steering a running agent?YesEngineer Promptsfor Performanceredirect, refine, restartNoEngineer Promptsfor Performancecontext layers,prompt process flowEvery path: the model sees only the assembled request
Read this topic in the online study guide
Figure 5.2 Decision tree: Craft Effective Prompts
Is the approachalready clear?Plan firstexplore, plan, review the planNoYesOne small task with anobservable result?Split and sequencethen take each step through this treeNoYesOutput must follow your ownspecific pattern or convention?Examples in the first promptone-shot or few-shot, representativeYesNo or unsureZero-shot baselinemeets the requirement?Keep zero-shotbaseline was enoughYesNoAdd examplesone-shot or few-shot:representative examples or testsEvery prompt: goal first, then requirements; attach essential context explicitly
Read this topic in the online study guide
Figure 5.9 Decision tree: Engineer Prompts for Performance
Is the next request onthis session's task?NoNew sessionold history stays behindYesDid a turn go stale orset a failed direction?YesDelete a turnremoved from chat historyNoHas history grown largefor this same task?YesCompactionsummary replaces older turnsNoFollow-up promptearlier turns resolve references
Read this topic in the online study guide

Chapter 6: Use GitHub Copilot responsibly

Figure 6.1 Decision tree: Use GitHub Copilot responsibly
Asks why output fails, orwhich principle is at stake?YesPrinciplesname the source or principleNoIs the request stillbeing written?YesOperation: before Copilot actsscope: context, a clear prompt;for an agent task, acceptance criteriaNoIs an agent working, or askingyou to approve an action?YesOperation: while an agent actssupervise: read any actionit asks you to approveNoOperation: after Copilot actsvalidate: each check proves only its own questionAlways: a person answers for the result
Read this topic in the online study guide
Figure 6.2 Decision tree: Understand Responsible AI Principles
Similarly situated people orgroups get different outcomes?YesFairnesscompare segments, override controlsNoPersonal data or secrets collectedwithout consent, over-kept or exposed?YesPrivacy and securityconsent, minimize, protectNoSome users cannot use it or benefit(ability, human language, connectivity)?YesInclusivenessalternative modes, adaptationNoUsers cannot tell AI is involved,what it can do, or why it acted?YesTransparencydisclose, explain, logNoNo person answers for the outcome,or the model has the final say?YesAccountabilityhuman authority, recourseNoFails on edge cases or unexpectedinput, or causes unintended harm?YesReliability and safetyedge cases, red teaming, reviewNoNo match? Consider a Copilot limit or the govern, map, measure, manage processillustrative, not an exhaustive test
Read this topic in the online study guide
Figure 6.8 Decision tree: Validate and Operate AI Tools
Waiting for your approvalto run a command?Approve only after readingthe exact action and targetYesNoA code review comment?from Copilot code reviewEvaluate each commentit may be a false positiveYesNoExplanation, pull requestsummary or commit message?Check against code and diffbefore relying or publishingYesNo, it is codeSecurity-sensitive orcritical code?Validate intent, edge cases, fitplus deeper security review, scannersYesNoValidate intent, edge cases, fitplus security review and scannersAlways: an accountable person owns the result
Read this topic in the online study guide

Errata for edition 1

No known errata for this edition.

Found a mistake in the book? Report it and quote the chapter and figure or section number.

Get the book

Keep studying

How this book was made

The book's text was generated using AI and reviewed through Nex Arc Learning's editorial and fact-checking process. Nex Arc Learning assumes editorial responsibility for the published content. Exam objectives change over time; check the official exam page for the current objectives. More on our methods: About Nex Arc Learning.