Top 30 CI CD Interview Questions and Answers

30 questions. Pick an answer, or open “Show the answer” to read it.

Answers are graded in your browser. Nothing is saved, and no XP or streak is earned here. The app keeps score.

1. Question 1 of 30

   What is the main engineering benefit of Android CLI 1.0 reaching stable for CI/CD workflows?

   • aIt ships with Stability Doctor to diagnose Compose recomposition issues in CI logs.
   • bIt enables project scaffolding, device management, and builds to run without Android Studio on headless agents.
   • cIt introduces a framework for exposing in-app actions to AI assistants like Gemini.
   • dIt replaces Gradle as the build system for Android projects, offering faster compilation speeds.
   Show the answer

   Answer: b · It enables project scaffolding, device management, and builds to run without Android Studio on headless agents.

   Android CLI 1.0 stable exposes first-party terminal commands for scaffolding, device management, and builds, eliminating the need to install Android Studio on headless CI agents. Option C describes Android skills, a separate I/O feature for AI assistant integration.

   Read the full bite: Android CLI 1.0 Stable Unlocks Headless CI

2. Question 2 of 30

   Which scenario best describes a genuine shift-left practice rather than a common misconception?

   • aHiring additional QA staff to write more regression tests executed the day before release
   • bIntegrating static analysis into pull request builds so vulnerabilities are caught before merge
   • cRunning the existing end-to-end suite on faster hardware to cut execution time from hours to minutes
   • dRequiring developers to manually perform user acceptance testing before they commit code
   Show the answer

   Answer: b · Integrating static analysis into pull request builds so vulnerabilities are caught before merge

   Integrating static analysis into pull request builds moves security feedback to the coding phase, which is the essence of shift left. Hiring more QA staff to test before release only increases test volume at the same late stage, confusing more testing with earlier feedback.

   Read the full bite: What does shift left mean in CI/CD, and give two concrete examples?

3. Question 3 of 30

   Which promotion flow best reflects a robust automated testing strategy for a weekly retrained production model?

   • aOffline per-slice thresholds and bias checks, data validation for training-serving skew and drift, shadow deployment comparing latency and prediction distributions, then canary gated on business metrics with automatic rollback
   • bOffline accuracy check, then an A-B test on 50% of users, then manual review of predictions before full rollout
   • cData validation on the training set, then shadow deployment on 100% of live traffic, then promotion if prediction distributions match the incumbent model
   • dCanary deployment on 5% of traffic with predictions logged but not served, followed by offline accuracy validation, then rollout based on offline accuracy exceeding the previous version
   Show the answer

   Answer: a · Offline per-slice thresholds and bias checks, data validation for training-serving skew and drift, shadow deployment comparing latency and prediction distributions, then canary gated on business metrics with automatic rollback

   This option captures the four-layer strategy from the card: offline statistical validation, data validation, shadow deployment, and canary gated on business metrics with automatic rollback. Option D is the most tempting distractor because it uses correct terminology but reverses the order and incorrectly uses offline accuracy as the final promotion gate rather than live business metrics.

   Read the full bite: Design a robust automated testing strategy for ML models before production

4. Question 4 of 30

   Which best describes how TDD and CI map to the Jidoka principle in software development?

   • aCI is mainly about frequent code merging and TDD guarantees bug-free software without needing additional verification.
   • bTDD and CI are independent agile rituals that improve code quality but share no underlying philosophy with Lean manufacturing.
   • cJidoka is primarily about manual QA gates that catch defects at the end of the development process.
   • dTDD provides an early unit-level stop mechanism while CI provides an integration-level stop mechanism, embedding quality directly into the workflow.
   Show the answer

   Answer: d · TDD provides an early unit-level stop mechanism while CI provides an integration-level stop mechanism, embedding quality directly into the workflow.

   This answer correctly identifies that TDD stops defects at the unit level and CI stops them at integration, together instantiating Jidoka's stop-the-line philosophy. The most tempting distractor is the first option because candidates often treat these practices as separate buzzwords rather than structurally equivalent feedback loops.

   Read the full bite: Describe the relationship between Jidoka and TDD/CI

5. Question 5 of 30

   Which statement best explains how Test-Driven Development (TDD) and Continuous Integration (CI) embody the Lean principle of 'Build Quality In' (Jidoka)?

   • aThey primarily aim to increase code coverage and reduce the number of bugs found in production by shifting testing left.
   • bThey facilitate continuous improvement (Kaizen) by providing metrics on code quality and build stability over time.
   • cThey implement automated mechanisms that detect defects and immediately stop the process, preventing faulty work from moving downstream.
   • dThey streamline the development workflow by automating testing, allowing developers to focus on new feature implementation.
   Show the answer

   Answer: c · They implement automated mechanisms that detect defects and immediately stop the process, preventing faulty work from moving downstream.

   The core of Jidoka is the automated detection of a defect and the immediate stoppage of the process to prevent that defect from moving further. Both TDD (failing test stops development) and CI (failing build stops integration) directly implement this 'stop the line' mechanism. While increasing code coverage and reducing bugs (option A) are outcomes, they do not describe the specific 'stoppage' mechanism central to Jidoka.

   Read the full bite: Relate 'Build Quality In' (Jidoka) to TDD and CI

6. Question 6 of 30

   When is it most appropriate to use Expo Application Services (EAS)?

   • aDuring the initial phase of development to quickly prototype UI components in Expo Go.
   • bFor managing project dependencies and installing native modules directly on your local machine.
   • cTo start a local development server and debug JavaScript code on a simulator or device.
   • dWhen preparing a production-ready build for app store submission or pushing over-the-air updates.
   Show the answer

   Answer: d · When preparing a production-ready build for app store submission or pushing over-the-air updates.

   EAS is designed for production workflows, handling cloud builds, app store submissions, and over-the-air updates for React Native apps. It is explicitly stated as distinct from local development tools like the `expo` CLI, which are used for prototyping and running local development servers.

   Read the full bite: Expo Application Services (EAS): The Cloud Toolchain for React Native

7. Question 7 of 30

   Why should a team not rely solely on a pre-push hook to guarantee that all tests pass before merging?

   • aIt is local-only, not cloned with the repository, can be skipped with --no-verify, and does not run for web or API commits.
   • bThe --no-verify flag only skips commit hooks, meaning pre-push checks cannot be bypassed by developers.
   • cIt must be committed to version control before Git will automatically execute it for everyone who clones the repository.
   • dIt runs on the remote server after the push completes, so it cannot prevent broken code from entering the repository.
   Show the answer

   Answer: a · It is local-only, not cloned with the repository, can be skipped with --no-verify, and does not run for web or API commits.

   Pre-push hooks reside in .git/hooks and are not copied on clone, can be bypassed with --no-verify, and do not run for web or API commits, so they cannot replace server-side enforcement. Option B is a tempting misconception: --no-verify skips pre-push hooks as well as commit hooks.

   Read the full bite: Describe using a pre-push Git hook for checks and its CI limitations.

8. Question 8 of 30

   In scaled Trunk-Based Development, what is the intended purpose of short-lived branches?

   • aTo isolate unfinished work for multiple days before integration
   • bTo replace trunk as the main collaboration branch for large teams
   • cTo publish versioned artifacts for downstream testing
   • dTo perform code review and validation before merging to trunk within hours
   Show the answer

   Answer: d · To perform code review and validation before merging to trunk within hours

   Scaled Trunk-Based Development uses short-lived branches solely for review and CI validation before same-day integration to trunk, not for artifact publication or long-term isolation. Keeping branches for multiple days violates the core principle of integrating at least every 24 hours and reintroduces merge risk.

   Read the full bite: Describe Trunk-Based Development principles and CI/CD benefits

9. Question 9 of 30

   When is a build tool like Maven or Gradle necessary instead of invoking a compiler directly?

   • aWhen you need an orchestration layer that resolves dependencies, runs tests, and packages deployable artifacts
   • bWhen you need to replace the compiler with a faster code-translation engine
   • cWhen you want a tool to download libraries but not to orchestrate tests or packaging
   • dWhen you simply need a script that calls the compiler with the correct flags
   Show the answer

   Answer: a · When you need an orchestration layer that resolves dependencies, runs tests, and packages deployable artifacts

   A compiler only translates source into runnable code, whereas a build tool is an orchestration layer that handles dependency resolution, testing, and artifact packaging. Option D reflects the common misconception that a build tool is merely a wrapper around the compiler, while Option C wrongly reduces it to a package manager.

   Read the full bite: How do build tools differ from compilers or interpreters?

10. Question 10 of 30

    After a build tool automatically resolves a diamond conflict, what is the most critical next step before considering the issue fixed?

    • aExclude the older transitive dependency from every direct dependency that pulled it in
    • bPin the resolved JARs directly into version control to prevent future changes
    • cUpgrade every dependency in the graph to the latest available version
    • dRun the full test suite and integration tests against the resolved version to detect binary incompatibility
    Show the answer

    Answer: d · Run the full test suite and integration tests against the resolved version to detect binary incompatibility

    The card emphasizes that semantic versioning does not guarantee binary compatibility, so the resolved version must be verified with full tests. Simply excluding the older transitive dependency or upgrading everything to latest skips runtime verification and risks hidden breaking changes.

    Read the full bite: Explain dependency management and diamond conflicts in automated builds

11. Question 11 of 30

    What should an automated pipeline validate after a model is tagged Staging but before it receives production traffic?

    • aThat the registry label is updated to Production, which immediately routes 100% of traffic to the model
    • bThat training holdout metrics meet thresholds and the model card has been signed off
    • cData drift, performance regression, bias thresholds, schema compatibility, and security scanning
    • dThat the model owner approved the promotion via email and moved the registry tag to Production
    Show the answer

    Answer: c · Data drift, performance regression, bias thresholds, schema compatibility, and security scanning

    Before any traffic exposure, the pipeline must automatically validate data drift, performance regression, bias, schema compatibility, and security against production baselines. Option B is tempting because it mentions holdout metrics and model cards, but a signed card is a trigger rather than a validation gate, and holdout metrics alone omit critical production-specific checks like drift and bias.

    Read the full bite: Design a CI/CD pipeline that automates model promotion from Staging to Production

12. Question 12 of 30

    Which strategy best guarantees that CI and developers use bitwise-identical Docker dev environments without host-specific drift?

    • aPin the Dockerfile base image by digest in Git and have CI and developers rebuild the image locally before each run
    • bBuild the image once, push it to a registry with an immutable tag or digest, and have both CI and developers pull that exact image
    • cVersion control the Dockerfile and require every developer to build the image locally before running tests
    • dUse a private registry and standardize on the 'latest' tag so all machines automatically receive the most current environment
    Show the answer

    Answer: b · Build the image once, push it to a registry with an immutable tag or digest, and have both CI and developers pull that exact image

    The card treats the built image—not the Dockerfile—as the immutable artifact to distribute, and pulling an exact tagged or digested image eliminates host dependency and cache variability. Option A is tempting because pinning a base image by digest is correct, but local rebuilds still reintroduce 'works on my machine' discrepancies from host caches and build contexts.

    Read the full bite: How do you version and distribute Docker dev environments consistently?

13. Question 13 of 30

    What is a required step before successfully uploading a Docker image to a specific remote registry?

    • aExporting the image to a .tar file using docker save.
    • bTagging the image with the full registry address and desired name.
    • cVerifying the image runs correctly in a local container.
    • dLogging into the Docker Hub account from the command line.
    Show the answer

    Answer: b · Tagging the image with the full registry address and desired name.

    The card states, "Before you can push, you must tag your image with the registry's address." This tag is essential for Docker to know which specific registry to send the image to. While logging in (option D) is often necessary for authentication, tagging (option B) is the fundamental step to specify the destination registry.

    Read the full bite: Docker Push and Pull: Moving Container Images

14. Question 14 of 30

    What is the most significant downside of dynamically generating CI/CD pipelines from repository contents?

    • aPipelines can never run in parallel
    • bIt always runs slower than a static pipeline
    • cIt is impossible to build Docker images this way
    • dThe effective pipeline is not visible in the repo, hurting transparency, debugging, and auditability
    Show the answer

    Answer: d · The effective pipeline is not visible in the repo, hurting transparency, debugging, and auditability

    Generated pipelines trade explicitness for convenience: developers cannot read their real pipeline from a checked-in file, which complicates debugging, reproduction, and security auditing. Parallelism, Docker builds, and speed are not inherently blocked.

    Read the full bite: Dynamically generate CI/CD pipelines

15. Question 15 of 30

    What is the most critical factor that distinguishes an effective automated performance and accessibility pipeline from one that merely generates reports?

    • aIt audits production URLs with PageSpeed Insights instead of preview environments.
    • bIt relies on Chrome UX Report field data to validate every pull request.
    • cIt stores historical Lighthouse scores in BigQuery to visualize long-term trends.
    • dIt enforces assertion budgets that block merges and routes regressions to named owners.
    Show the answer

    Answer: d · It enforces assertion budgets that block merges and routes regressions to named owners.

    The card emphasizes that operationalizing quality gates requires enforcing thresholds that block merges and assigning regressions to specific owners, not just collecting metrics. Option C describes a useful but insufficient data layer, since trends without enforcement are commonly ignored.

    Read the full bite: How do you architect an automated performance and accessibility testing pipeline?

16. Question 16 of 30

    Which combination of automated CI tests best validates a classification model artifact before deployment?

    • aEnd-to-end tests confirming the training script finishes and the inference API returns predictions within 50 milliseconds
    • bHTTP endpoint tests, response latency checks, and load testing of the serving container
    • cData schema checks, performance regression against a baseline, bias audits, and artifact integrity tests
    • dUnit tests for the training code, code coverage thresholds, and pipeline linting
    Show the answer

    Answer: c · Data schema checks, performance regression against a baseline, bias audits, and artifact integrity tests

    Data schema checks, performance regression, bias audits, and artifact integrity tests validate the model artifact itself rather than just the surrounding code. Option A is tempting because it verifies the pipeline runs and responds quickly, but it never asserts whether the model's accuracy, fairness, or data assumptions have degraded.

    Read the full bite: What automated tests belong in CI before deploying a classification model?

17. Question 17 of 30

    In a CI/CD4ML pipeline, why is a model registry considered an active control plane rather than passive storage?

    • aIt provides a temporary staging area that holds the latest trained model between the CI training job and the CD deployment step.
    • bIt serves as a versioned artifact store where CI publishes validated models and CD consumes approved versions, enabling promotion workflows and rollback by switching version pointers.
    • cIt replaces experiment tracking tools by storing training parameters and metrics for every research run.
    • dIt acts as a cloud storage bucket for data scientists to manually upload model binaries, which engineers later copy into production.
    Show the answer

    Answer: b · It serves as a versioned artifact store where CI publishes validated models and CD consumes approved versions, enabling promotion workflows and rollback by switching version pointers.

    The registry is an active control plane because CI publishes validated, versioned artifacts with metadata and CD consumes approved versions, supporting promotion and rollback. Treating it as a manual cloud storage bucket misses the versioning, governance, and automated lifecycle contracts between data science and engineering.

    Read the full bite: What is a model registry's purpose in CI/CD4ML and its CI/CD interaction?

18. Question 18 of 30

    When configuring Terraform for an RDS database, which approach best limits secret exposure across version history and state files?

    • aStore the master password in AWS Secrets Manager with rotation enabled, reference the ARN in Terraform, and use an encrypted remote state backend
    • bEncrypt the master password with git-crypt and commit it to the same repository as the Terraform code
    • cPass the master password as an environment variable to the Terraform apply step
    • dStore the master password in terraform.tfvars inside a private Git repository
    Show the answer

    Answer: a · Store the master password in AWS Secrets Manager with rotation enabled, reference the ARN in Terraform, and use an encrypted remote state backend

    AWS Secrets Manager keeps the actual value out of code and enables automatic rotation and fine-grained access control, while an encrypted remote state backend prevents plaintext storage in state files. Passing the password via an environment variable is a common shortcut, but environment variables leak in process listings, crash reports, and potentially logs.

    Read the full bite: How do you manage secrets within IaC configurations?

19. Question 19 of 30

    How should a CI/CD pipeline enforce a fairness constraint on a candidate model?

    • aLog subgroup metrics as warnings while always allowing the deploy to proceed
    • bCompute sliced subgroup metrics and exit nonzero to block promotion when a threshold is breached
    • cRun fairness checks manually after deployment
    • dGate solely on aggregate accuracy across the full test set
    Show the answer

    Answer: b · Compute sliced subgroup metrics and exit nonzero to block promotion when a threshold is breached

    Enforcement requires sliced metrics compared to thresholds that fail the build and block deployment. Aggregate accuracy hides subgroup harm, and warning-only or post-deploy checks do not prevent shipping a violating model.

    Read the full bite: Fairness and robustness gates in CI/CD

20. Question 20 of 30

    Which mechanism best lets the core team stay the gatekeeper of quality while still accepting outside contributions?

    • aDisabling pull requests and accepting changes by email
    • bGranting all product engineers direct merge access to main
    • cRequiring CODEOWNERS approval plus automated CI gates on every PR
    • dHaving the core team rewrite every contribution from scratch
    Show the answer

    Answer: c · Requiring CODEOWNERS approval plus automated CI gates on every PR

    CODEOWNERS enforces core review while CI gates catch regressions automatically, scaling contribution safely. Direct merge access removes the quality gate, and rewriting everything recreates the bottleneck that openness was meant to solve.

    Read the full bite: Designing a contribution model with quality gates

21. Question 21 of 30

    Why is contract testing especially important in a microservice-based ML system?

    • aIt eliminates the need for API versioning
    • bIt replaces the need for any unit tests in each service
    • cIt allows all services to be deployed as a single unit
    • dServices evolve independently, so verifying interface and data-schema expectations catches breaking changes before integration
    Show the answer

    Answer: d · Services evolve independently, so verifying interface and data-schema expectations catches breaking changes before integration

    Independent release cadences make interface and schema drift the main risk; consumer-driven contract tests catch breaks in the producer's pipeline early. It complements, not replaces, unit tests and versioning.

    Read the full bite: CI/CD for microservice-based ML systems

22. Question 22 of 30

    When moving Figma Variables from REST API into Style Dictionary, which step resolves the impedance mismatch between the two systems?

    • aTrigger the sync via Figma webhook on every file change
    • bFlatten all mode-specific values into a single default theme
    • cResolve aliases, normalize types, and restructure the payload into W3C draft JSON
    • dPaginate API requests to retrieve all variable collections and mode values
    Show the answer

    Answer: c · Resolve aliases, normalize types, and restructure the payload into W3C draft JSON

    Figma returns flat IDs with unresolved references and raw value types, so the payload must be transformed into W3C draft JSON before Style Dictionary can consume it. Option B is a tempting anti-pattern because flattening feels like simplification, but it actually destroys multi-theme context and still leaves the format incompatible.

    Read the full bite: How would you extract Figma Variables via REST API for Style Dictionary?

23. Question 23 of 30

    Which approach best keeps database credentials out of Git while supporting local development?

    • aEncode the credentials with base64 and store them in a committed config file
    • bCommit a template file and manually replace values before each deployment
    • cStore credentials in a committed file and delete it in a later commit if exposed
    • dUse a .env file listed in .gitignore and read via environment variables
    Show the answer

    Answer: d · Use a .env file listed in .gitignore and read via environment variables

    A gitignored .env file keeps secrets out of the repository while the app reads them at runtime via environment variables. Deleting a committed file later is insufficient because Git history is immutable and distributed, so the secret remains in every clone and fork.

    Read the full bite: Why avoid committing secrets to Git, and secure local alternatives?

24. Question 24 of 30

    When leading a multi-sprint refactor of an inverted test pyramid, which sequencing best protects delivery while reducing end-to-end test reliance?

    • aImmediately delete overlapping end-to-end tests and rely on faster unit tests added afterward to catch any regressions
    • bPause feature development for two sprints to rewrite the entire suite from scratch, then switch over all at once
    • cAudit coverage, scaffold unit and integration tests for the targeted logic, then incrementally migrate redundant end-to-end tests per sprint while preserving a critical-path subset
    • dKeep all existing end-to-end tests, add unit tests for new code only, and reduce flakiness by increasing retry counts in the pipeline
    Show the answer

    Answer: c · Audit coverage, scaffold unit and integration tests for the targeted logic, then incrementally migrate redundant end-to-end tests per sprint while preserving a critical-path subset

    Scaffolding unit and integration tests before incrementally migrating redundant E2E tests preserves a safety net and avoids halting delivery. Deleting overlapping E2E tests before verifying lower-layer coverage catches the same failures is a common red flag that invites regressions.

    Read the full bite: Propose a multi-sprint strategy to fix an inverted test pyramid

25. Question 25 of 30

    What is the primary role of a container registry in a build-and-deploy pipeline for Cloud Run?

    • aIt compiles the source code into a Docker image during the deploy stage
    • bIt runs the application's unit tests before the image is built
    • cIt stores the built, tagged image so the deploy stage can pull that exact artifact
    • dIt generates the Dockerfile automatically from the source repository
    Show the answer

    Answer: c · It stores the built, tagged image so the deploy stage can pull that exact artifact

    The registry stores immutable tagged images and is the handoff between build (which pushes) and deploy (which pulls). It does not compile code, run tests, or generate Dockerfiles.

    Read the full bite: CI/CD pipeline for a container PaaS

26. Question 26 of 30

    Which break-glass workflow best satisfies zero-standing-access requirements when manual production secret access is unavoidable?

    • aLog all secret retrievals to a mutable SIEM dashboard and require the on-call engineer to file a post-incident ticket within 48 hours.
    • bRequire two senior staff approvals for a one-hour scoped session, record the session, and rotate the secret immediately after use.
    • cIssue a long-lived break-glass credential encrypted at rest that the on-call engineer downloads to their laptop for emergency use.
    • dSingle on-call engineer grants themselves a 24-hour admin role on the secrets manager with automatic rotation afterward.
    Show the answer

    Answer: b · Require two senior staff approvals for a one-hour scoped session, record the session, and rotate the secret immediately after use.

    Option B is correct because it enforces dual-control approval, time-bound scope, session recording, and immediate rotation as described in the card. Option D lacks dual-control and grants excessive duration, while Option C exemplifies the common mistake of delivering credentials to a laptop despite encryption at rest.

    Read the full bite: Prevent developer access to production secrets while preserving debuggability

27. Question 27 of 30

    In the proposed Next.js test pyramid, why are integration tests allocated to data fetching and route boundaries instead of expanding end-to-end coverage?

    • aThey run faster than unit tests and provide more precise failure signals than e2e tests
    • bThey validate framework wiring like middleware redirects without the speed and compute cost of a real browser
    • cServer Components cannot be unit tested, so integration tests serve as the base of the pyramid
    • dEnd-to-end tests lack sufficient confidence for business-critical flows like checkout and auth
    Show the answer

    Answer: b · They validate framework wiring like middleware redirects without the speed and compute cost of a real browser

    Integration tests target Next.js boundaries to catch framework wiring bugs without the cost of a real browser, while unit tests still dominate for speed. Option C is tempting but wrong because Server Components can be unit tested in a Node.js sandbox, so integration tests should not replace the pyramid base.

    Read the full bite: How do you balance unit, integration, and end-to-end tests in Next.js?

28. Question 28 of 30

    What is the best long-term resolution for repeatedly paging to clear a cache after deploys?

    • aAutomate the clear in the deploy pipeline and address why the deploy invalidates the cache
    • bIncrease the cache size so it never needs clearing
    • cRotate the on-call schedule so no one gets paged twice
    • dAdd the manual step to the on-call runbook so it is faster each time
    Show the answer

    Answer: a · Automate the clear in the deploy pipeline and address why the deploy invalidates the cache

    Eliminating toil means automating the action and, better still, removing the need via versioned cache keys. A runbook still requires a human, and resizing or rotating on-call does not address the recurring trigger.

    Read the full bite: Automating a recurring manual cache clear

29. Question 29 of 30

    Which scenario describes a valid automatic trigger for full model retraining in production?

    • aA cron job retrains the model daily solely because more frequent retraining feels safer.
    • bA new unlabeled dataset arrives in the raw data lake and exceeds a size threshold.
    • cProduction monitoring detects statistical drift on a held-out live slice below an SLO threshold.
    • dA nightly batch inference job finishes scoring the latest records in the data warehouse.
    Show the answer

    Answer: c · Production monitoring detects statistical drift on a held-out live slice below an SLO threshold.

    Production monitoring that detects statistical drift or accuracy decay below an SLO is a core automatic trigger for remediation retraining. The unlabeled data option is tempting because data volume is a valid trigger, but the card specifies that the incoming data must be labeled to kick off a training run.

    Read the full bite: What triggers automatic full retraining in an ML pipeline?

30. Question 30 of 30

    Why escalate a deprecation lint rule from warning to blocking error gradually rather than blocking all builds immediately?

    • aIt prevents new usages while giving existing consumers time and tooling to migrate
    • bESLint cannot emit errors until it has first emitted warnings
    • cBlocking errors are only supported in CI, not local development
    • dGradual escalation reduces the size of the compiled bundle
    Show the answer

    Answer: a · It prevents new usages while giving existing consumers time and tooling to migrate

    Staged escalation stops new debt immediately while letting teams migrate existing code via codemods, avoiding broken pipelines. The other options misstate how ESLint and bundling work.

    Read the full bite: Automated enforcement of deprecated component removal