ServiceNow CSA deep review — September 29, 2026

The CSA guide now maps 30 official subtopics, answers 40 original readiness prompts and includes three worked scenarios, eight proposed instance activities and 34 executed local checks. ServiceNow instance execution and independent human review remain pending.

Canonical scope and preserved history

The canonical blueprint still carries January 2026. Full browser main text supports six domains weighted 7/10/20/20/30/13 and subtopic counts 4/3/9/4/6/4. The saved snapshot had 26 merged/paraphrased entries. This review restores all 30 separately; it does not announce a new exam version. September 7 objective/status bytes were archived from HEAD, two historical audit references redirected, and the September 7 historical review preserved. The original September 2 review was also recovered from commit433983b with unchanged objective bytes, so the September 6 audit binds to real historical evidence instead of falling back to the new review. New snapshot hashes and the manual comparison are recorded with the operation.

The blueprint confirms 60 multiple-choice/multiple-select questions, 90 minutes, Pearson test-center/OnVUE delivery, and a 90-day registration-to-completion window. Cut score is undisclosed; section percentages and practice scores do not calculate a pass guarantee. Training recommendations are not formal prerequisites. Annual maintenance remains distinct from mainline study. The automated client still returns an empty body: manual browser verification is not a successful generated extraction. The official sample tail was incidentally returned/read but excluded from the saved main capture and not used or reproduced.

Teaching and actual execution

All 34 original standard-library Python/SQLite checks passed. Eight synthetic import rows yield two inserts, one update, one unchanged result, two quarantines, one conflict and one stale result. Exact final records retain both vendors' ID42: VendorA revision2/North and VendorB revision1/East. Whole-batch replay makes no mutations. The SQLite primary key and integrity check are real local execution. Source normalization, case-sensitive external IDs, quarantine and revision rules are chosen policies, not ServiceNow's default coalesce or CMDB reconciliation. This sequential model proves no distributed/concurrent guarantee.

Trusted-label access cases require an explicit qualifying role and other conditions; table/field/derived-value distinctions motivate negative tests. This is not authentication or a full ACL engine. A deployment dependency model rejects omitted, repeated and cyclic steps. An expected-revision example preserves a later revision6 repair when a backout still expects5; it is not an update-set collision emulator or actual ServiceNow recovery. No account, PDI, plugin, privilege elevation, instance script, notification, paid lesson or external action was used.

Primary-source findings

The coalesce reference requires every selected field to match, updates only the first of multiple target matches and documents empty/case options. The guide now separates identity matching from preflight uniqueness and validation. Original cases demonstrate why total counts alone are inadequate evidence.

The Deny-Unless page gives priority over Allow-If but contradicts itself about access when no Allow-If matches. That case remains unresolved. Empty/invalid rules, no-match behavior, wildcard defaults and admin overrides must be assessed separately in the installed engine. The function-field page explains contributor and report access, but indexed Australia and fresh Brazil bodies both name their own selected family as the transition. Exact historical introduction is not established. Generated page summaries were excluded from the technical evidence.

The Business Rule main body describes before-save behavior, recursion risk, async ordering/previous limitations and bypass/context exceptions. Its Workflow Studio recommendation for new process automation does not remove Business Rules from the CSA blueprint or justify rewriting production logic here.

The tracked-customization reference distinguishes catalog definitions from orders, restricts update_synch manipulation and describes destructive schema/preview limits. The collision reference lists role/group record classes, qualifying the delta guide's broad nonmetadata shorthand without proving every membership/group change is automatically captured. Actual payload inspection remains necessary. Current batch guidance qualifies blanket Ignore advice: parent only, other sets Complete. XML import and remote retrieval are separate transfer paths; neither was executed.

Release and catalog boundaries

Current browser documentation identifies Brazil and September 10, while cached search bodies still identify Australia. The August 13 vendor Developer Passport article advertises eleven September 14–18 preview sessions. Its main article was read, but no linked videos were played. Product documentation/preview availability does not establish CSA exam cutover or general availability. The 2026 delta guide remains update-set maintenance material.

The old Welcome citation targeted a CSA certification path. The guide now links the actual Welcome course from the canonical blueprint and labels the earlier path correctly. Welcome/Fundamentals remain loading shells; earlier three-hour/three-day durations are unverified. Three resources returned403, so earlier Udemy8h17/May2026 and O'Reilly2h30/2025 metadata and infographic content were not revalidated.

MeasureUp's specific listing has120questions and domain counts6/12/26/23/33/20, versus generic roughly150 wording on the same page. That distribution differs from exam weights. Translation advertising and readiness percentages are not official exam language or scoring policy. No practice or paid lesson interiors were accessed. The community checklist is user-authored July2025; only its main and first comments were read, with no attachment or mock endorsement.

The Pearson FAQ was fully read for scheduling/registration context. English CMP payment dates differ by a day from supplemental public-language references; the official completion route remained empty/error. No assigned deadline, future window or account maintenance compliance was established. Thirty direct receipts show27HTTPsuccesses and3blocked requests; many successes are shells, and per-source browser/indexed reading boundaries are recorded separately.

Evidence and remaining work

The operation at ADLC_Docs/operations/2026-09-29-servicenow-csa-deep-review.json records prior records, objective mapping, source receipts, manual acceptance/archive hashes, exact local code/output and repository gates. Remaining limits are the source contradictions, assigned maintenance policy, live instance activities, inaccessible interiors and human review. The five reserved GitHub guides and disabled notification pilot remain untouched.