Microsoft refresh group 5 of 5 — October 28, 2026¶
This final group refreshes eight guides and maps 544 current detailed objectives. After delivery, zero groups and zero guides remain in the frozen 48-guide refresh. The operations ledger records exact objective/guide/teaching-span bindings, opened-source scopes, initial findings, corrections and review limits.
Two subset exam researchers read four complete guides each, all detailed current objective tasks, and complete final guides/diffs. Technical and learning researchers record their narrower scopes. This is bounded AI-assisted refresh work; a full independent rubric-2 semantic audit, human review and product/lab execution remain separate.
| Guide | Detailed objectives | Source gate | Links |
|---|---|---|---|
| AB-410 | 48 | Blocked: retained evidence gap | 57/58 reachable; 1 access-blocked; 0 missing/error |
| AZ-120 | 78 | Blocked: retained evidence gap | 56/58 reachable; 2 access-blocked; 0 missing/error |
| AZ-802 | 115 | Blocked: retained evidence gap | 86/91 reachable; 5 access-blocked; 0 missing/error |
| MO-110 | 70 | Passed; human review pending | 20/20 reachable; 0 access-blocked; 0 missing/error |
| MO-111 | 47 | Blocked: retained evidence gap | 31/32 reachable; 1 access-blocked; 0 missing/error |
| MO-210 | 65 | Passed; human review pending | 32/32 reachable; 0 access-blocked; 0 missing/error |
| MO-211 | 48 | Passed; human review pending | 32/32 reachable; 0 access-blocked; 0 missing/error |
| MO-310 | 73 | Passed; human review pending | 27/27 reachable; 0 access-blocked; 0 missing/error |
Objectives and history¶
All eight accepted objective files remain byte-identical. The five Office adapter snapshots contain coarse domain lists; they are preserved separately from detailed current official PDF tasks. Detailed wordings, PDF raw hashes and extraction provenance are recorded in the matrix/research evidence, rather than falsely claiming each task appears in the coarse snapshot. AB-410 and AZ-802 status metadata now reflect October 16 and October 19 page updates; neither publishes a new skills-effective date. Six other status files and all prior source/deep/AI-audit review values remain unchanged.
The initial role findings, exact repair packets and current semantic dispositions remain distinct. A retains 49 initial workflow flags: 48 substantive closures and one WAC workflow with its UI limitation, plus four separate initial provisional rows. B retains 61 initial partial flags: 60 substantive closures and the optional style-set investigation with its current-source gap. The independent conditional-rule teaching repair is additional; it does not inflate the initial counts. Fifty-two exam-research packets include workflow teaching, platform, lifecycle and assessment corrections. Learning routes add eight focused outputs plus targeted resource warnings. The duplicate Teams lifecycle repair is applied once with both role evidence records. Final rereads additionally clarify the Excel IF input/output cells, a past SAP connector deadline, four current-watch dates and two retired-course redirects. Independent review additionally requires the unconfirmed Word style-set investigation to be optional, restores two direct PDF links, adds actual built-in conditional-formatting teaching, and corrects several evidence-span selections while preserving earlier matrices. These are source-based proposed operations; no Office, Windows Server, Power Platform, SAP or Azure workflow ran.
AB-410¶
Add Dataverse, app/form/chart, canvas save, formula/component, flow, prompt/model and process workflows. Current official pages announce Navigator Practice Assessment access, but signed-in launch/questions remain untested. Preserve the SAP prompt-knowledge contradiction. Dataverse for Teams deprecation dates are product lifecycle limits, not new exam objectives.
AZ-120¶
Add policy, RISE customer-interface, deployment, placement/private-endpoint, storage, SQL listener/replication, isolated restore and LaMa workflows. RISE ownership and excluded monitoring integrations stay explicit. Preserve the native-HSR versus FAQ procedure conflict; no restore/failover or production action is authorized by the examples.
AZ-802¶
Add a bounded WAC observation-to-AMA/DCR/alert path while rejecting the source article’s retired onboarding tools. Keep SSH Direct support and conflicting assessment availability unconfirmed. Official Lab 0 is under development and must not be used to provision. Old AZ-800/AZ-801 course URLs redirect to the general browser; earlier course duration/coverage is historical.
MO-110¶
Add a document-production workflow with properties, distinct editable/PDF/print/sharing outcomes and accessibility checks, plus explicit table construction. Current official PDF has 70 tasks. All file/print/cloud outputs are prospective checks, not Word execution.
MO-111¶
Add document dependency/protection, pagination, reusable design, captions/fields/forms, merge and review-artifact workflows. Exact current Save as New Style Set UI remains unconfirmed; historical first-hand discussion does not prove current build behavior. The style-set investigation is explicitly optional and cannot be a required lab success. Keep Windows/Mac/web and exam-command boundaries beside the task.
MO-210¶
Add workbook navigation/formatting, names, structured table/formula checks and chart revision. IF boundary tests keep D2 input and E2 output separate on a test sheet. An independently requested built-in conditional rule now has a threshold, explicit B3 input change, expected matches and targeted removal that preserves manual formatting. Totals and comparisons are expected arithmetic, not observed Excel results.
MO-211¶
Add criteria/logical/legacy lookup, consolidation/what-if, formula diagnosis, PivotTable/PivotChart and specialized chart artifacts. Preserve classic versus Data Model calculation differences, source extent, refresh/slicer and platform limits.
MO-310¶
Add text/shape/screenshot/ink and table-structure operations. Reading order and SmartArt publication/build observations retain their scope; the inaccessible announcement body was not newly verified. No presentation engine or screen-reader run occurred.
Evidence and validation¶
Initial health: 255 records, 247 reachable, eight access-blocked and zero broken. Additional health: 86 associations, 85 reachable, one access-blocked and zero broken. One additional independently identified built-in conditional-formatting workflow adds a primary source checked HTTP200. The catalog registers 86 new exact evaluated URLs, appends scoped observations to 36 existing sources, and records two course-availability updates and one association-only update separately (39 existing catalog objects changed in total). Transport success is not technical proof; course metadata is not paid-content inspection.
Both fresh-context reviewers completed bounded final checks with zero actionable guide/mapping findings, including separate verification of the final five Office disclosure sentences. All 544 objective wordings and 717 teaching spans match their current bindings. All 329 earlier source reviews, 206 deep receipts and 42 AI-audit batches retain their original values; eight objective files and six status files are byte-identical, with two metadata-only status updates. All 199 unit tests, repository validation and reviewed-resource drift checks passed. A strict site build and generated-site validation are the final required local gate; actual outcomes are recorded in the ledger before delivery. Retained source blockers are not cleared merely because a URL works.