Microsoft AB-100 Practice Test 79

Question 780

A knowledge pipeline adds `effective_to` and changes document IDs, but an old agent build expects only `effective_from`. What does coordinated release require?

Deploy the index before checking whether consumers can parse its fields.
Introduce a compatibility-tested schema migration and update dependent retrieval filters before activating the new index.
Ignore `effective_to` so superseded documents remain eligible.
Allow the model to guess document IDs when lookup fails.

Explanation

Correct Answer

Authoritative effective-interval filtering must remain consistent across data and agent versions.

Incorrect Answers

Uncoordinated breaking changes can cause runtime failures or missing authorization filters.

Loss of expiration semantics can produce obsolete policy advice.

Generated identifiers do not establish source provenance or valid retrieval links.

Question 781

Two departments merge customer-case evidence into one search corpus. One source uses account IDs and another relies on names. What data ALM precondition is needed?

Erase source provenance after merging to reduce payload size.
Allow agents to infer which department owns a record from writing style.
Validate canonical identities, source ownership and transformation lineage before publishing merged retrieval records.
Match customers solely by the closest normalized display name.

Explanation

Correct Answer

Without reliable crosswalks, combined context can attach one customer’s sensitive events to another.

Incorrect Answers

Without lineage the team cannot correct or retract a bad source record.

Generated guesswork cannot replace governed source ownership metadata.

Names can duplicate or change, so similarity is not authoritative identity.

Question 782

A production search index is lost, but policy PDFs and an agent prompt remain. Which artifact supports exact reconstruction?

Only a screenshot of the last correct chat answer.
Only the name of the embedding model.
An instruction to generate plausible policy snippets while rebuilding.
A versioned manifest of source document revisions, ingestion transforms, embedding settings and index filters.

Explanation

Correct Answer

Rebuilding requires more than raw files because chunking and permission metadata alter retrievable evidence.

Incorrect Answers

One screenshot cannot reconstruct the indexed corpus or retrieval controls.

Model selection is one component; source revisions and transforms remain missing.

Generated replacements would fabricate regulatory content rather than restore approved knowledge.

Question 783

A tax-policy revision is approved today but legally effective next month. An ingestion job indexes it immediately. How should the agent release handle both versions?

Index with effective intervals and select the binding revision based on the transaction date and user context.
Use the newly uploaded revision for all transactions immediately.
Keep the old revision forever without an expiration rule.
Average values from both tax policies.

Explanation

Correct Answer

Published availability and legal effectivity are distinct dimensions and need explicit retrieval logic.

Incorrect Answers

An approved future policy may not yet govern current accounting.

Stale guidance would persist after the new policy becomes effective.

Blended rates are not legally defined merely because two versions exist.

Question 784

A team develops a customer-service agent directly in production, changing prompts and actions while customers are online. Which ALM design is preferable?

Assume a new agent version automatically restores the previous one after failure.
Use development, test and production environments with controlled solution promotion and documented approvals.
Let all makers edit production at will to minimize deployment latency.
Copy individual screenshots of the agent settings into production.

Explanation

Correct Answer

Staged environments enable independent validation before customer-facing changes.

Incorrect Answers

Recovery requires explicit versioned artifacts and a tested rollback path.

Unreviewed live edits can alter behavior without reproducible testing or rollback.

Screenshots do not preserve executable solution dependencies or configuration state.

Question 785

A Copilot Studio agent works in test but sends production orders to the staging API URL after promotion. What should the ALM package use?

Allow the model to infer the production endpoint from its hostname.
Reuse staging credentials in production to preserve successful tests.
Environment-specific configuration bindings validated during solution deployment and post-deployment smoke tests.
Edit every prompt manually after release without recording the changed URL.

Explanation

Correct Answer

Endpoint values differ by environment and must not be baked into conversational logic.

Incorrect Answers

Guessed URLs are not an approved service contract.

Cross-environment credential reuse violates isolation and may target the wrong service.

Untracked prompt changes make release state hard to reproduce.

Question 786

A Copilot Studio agent is exported to a new environment but its custom action is missing. What should the deployment review check?

Add a sentence telling the model to use the missing action anyway.
Use the developer’s personal connection to make the production action work.
Count a successful chat greeting as proof action deployment succeeded.
The solution’s action, connector and connection-reference dependencies are included and resolved in the target environment.

Explanation

Correct Answer

Prompt text alone cannot recreate executable flows and permissioned external connections.

Incorrect Answers

A written instruction cannot invoke a connector that was not deployed.

Personal credentials are not a stable least-privilege production identity.

A greeting does not exercise the missing downstream business operation.

Question 787

A newer agent solution changes topic trigger phrases, and a cancellation request now starts a refund workflow. What ALM validation is needed?

Run version-specific intent-routing regression tests on cancellation and refund near-neighbor scenarios.
Compare only the number of topics in the two solutions.
Release because the new topic display names are clearer.
Test only a single generic greeting.

Explanation

Correct Answer

Topic changes can alter executable business decisions even when all connectors remain healthy.

Incorrect Answers

Topic count does not prove correct route selection for user language.

Naming aesthetics do not establish protected transaction behavior.

Greeting success cannot detect overlap between high-impact write intents.

Question 788

An imported solution references a connector identity with broader rights than in test. The agent can now edit all Finance entities. What should release governance require?

Allow the model to decide which extra permissions are safe.
Validate the production connection identity and permission scopes against approved least-privilege requirements.
Accept broader scope because all integration calls now succeed.
Hide the expanded features from the agent’s start screen.

Explanation

Correct Answer

Environment imports can bind to different credentials and must not silently expand authority.

Incorrect Answers

Permission grants need authenticated administrators and documented policy, not model reasoning.

Eliminating denials by overprivileging creates unauthorized data and action access.

UI concealment does not secure callable backend actions.

Question 789

A new agent solution introduces an invalid order-submit action schema. Existing conversations remain open. What rollback plan preserves continuity?

Leave the broken action enabled until users stop reporting errors.
Erase all pending conversation state to simplify deployment.
Promote the last known-good solution and reconcile pending order jobs using stable transaction identifiers.
Restart every customer order with fresh IDs after the solution revert.

Explanation

Correct Answer

Rollback must restore the tool contract while avoiding duplicated or abandoned in-flight orders.

Incorrect Answers

Known schema failure is a material release defect that should be contained.

Deleting identifiers prevents safe recovery of unresolved external transactions.

Uncoordinated replays can duplicate existing pending or completed purchases.

Question 790

A deployed agent logs sensitive support transcripts to a development workspace readable by all makers. Which deployment defect must be fixed?

Continue logging raw transcripts because only internal users see the workspace.
Base64-encode transcripts and call them confidential.
Disable all monitoring permanently rather than separating secure telemetry.
Bind production telemetry to a controlled destination with role-based access, minimization and retention policy.

Explanation

Correct Answer

Analytics and traces are part of the operational data surface and need environment-specific protection.

Incorrect Answers

Internal visibility does not authorize every maker to view customer information.

Encoding does not prevent anyone with access from decoding the text.

Operations still need protected diagnostics for failures and audit events.

Agentic AI Business Solutions Architect (AB-100) Practice Tests

Microsoft AB-100 Practice Test 77 - 10 Questions (Last Reviewed 25-Oct-2026)
Microsoft AB-100 Practice Test 78 - 10 Questions (Last Reviewed 24-Oct-2026)
Microsoft AB-100 Practice Test 79 - 10 Questions (Last Reviewed 23-Oct-2026)
Microsoft AB-100 Practice Test 80 - 10 Questions (Last Reviewed 22-Oct-2026)
Microsoft AB-100 Practice Test 81 - 10 Questions (Last Reviewed 21-Oct-2026)

Pass Your Microsoft AB-100 Exam Easy!

Question 780

Explanation

Correct Answer

Incorrect Answers

Question 781

Explanation

Correct Answer

Incorrect Answers

Question 782

Explanation

Correct Answer

Incorrect Answers

Question 783

Explanation

Correct Answer

Incorrect Answers

Question 784

Explanation

Correct Answer

Incorrect Answers

Question 785

Explanation

Correct Answer

Incorrect Answers

Question 786

Explanation

Correct Answer

Incorrect Answers

Question 787

Explanation

Correct Answer

Incorrect Answers

Question 788

Explanation

Correct Answer

Incorrect Answers

Question 789

Explanation

Correct Answer

Incorrect Answers

Question 790

Explanation

Correct Answer

Incorrect Answers

Agentic AI Business Solutions Architect (AB-100) Practice Tests