JSA-41-01 deep review — September 29, 2026

The study guide now maps all 40 numbered objectives, improves all 40 answer explanations, and supplies three complete scenarios with executable language/API evidence. This is a same-context AI review; independent human review remains pending.

Scope and source evidence

The complete indexed current English syllabus still shows active JSA-41-01 and September 22, 2025. The four objective groups contain 11/7/12/10 subjects, with item counts 11/7/12/10 and weights 25/23/27/25. Manual comparison supports retaining the existing abbreviated-ID snapshots after direct fetching and automated monitoring failed. No successful new hash comparison is claimed. An older 2022 PDF was visible only as indexed cover/selected excerpts and does not replace the current baseline. Normalized 70% passing does not establish an equal-item shortcut.

The complete indexed credential and certification-overview pages were read, alongside the complete public course and Pearson policy content and selected TestNow policy sections. TestNow's seven-day failed retake differs by channel from Pearson's fifteen-day rule. Free-retake eligibility wording and USD295/345 product prices do not establish a particular voucher's terms. Spanish has a TestNow qualification in the overview; an unrelated Python credential label on the main page is a copy issue. No credential expiration is inferred from silence, and no account, booking or checkout was entered.

There are 41 registered sources, with direct outcomes {'blocked': 2, 'error': 6, 'ok': 33}. HTTP success is not equivalent to complete content review: Cisco returned a 23-character application shell; technical reading is section-specific; paid interiors remain unseen. O'Reilly and Udemy returned HTTP403. The operation record lists each reading boundary and preserves the direct failures rather than converting indexed fallback reading into a fabricated HTTP success.

Teaching and executed evidence

Seven exact public JavaScript blocks were extracted and hashed. The core workbook passed 143 named checks in both Node 24.18.1 and Chrome 154.0.8037.58. The catalog and normalization scenarios passed 84 additional checks in each runtime. Three introductory examples ran in separate Chrome pages. No runtime or package was installed.

The changes distinguish new descriptor defaults from preserved attributes on existing properties; getter/setter effects and partial mutation during copying; null prototypes; class fields versus methods; bound/arrow/constructor receivers; UTF-16 units versus code points; numeric coercion and safe integer bounds; valid versus normalized calendar dates; sparse slots; Map/Set identity and order; JSON losses; regex state; and exact promise/async contracts. The workbook's unique symbol extension of Array.prototype is confined to a disposable process/page and removed in finally. Vendor global-prototype-freezing examples were not executed.

The catalog validates bounded integer prices/quantities and IDs/tags, copies and freezes tags, and uses descriptors, static factories, subclassing and receiver-preserving super calls. Its discount rounding is an explicit fictional rule. The JSON importer rejects malformed/root-invalid input, reports row errors, retains the first valid duplicate ID and uses strict decimal/date contracts. Bounds keep the largest aggregate at 10^12 cents; no pre-parse byte limit or arbitrary-object sandbox is claimed. Tests cover malformed roots, leap years/centuries, years 0000/9999, sparse tags, duplicates and preservation of caller data.

The browser client passed 36 network/UI checks using actual Fetch, XMLHttpRequest, AbortController, timers and DOM clicks with route-intercepted synthetic same-origin responses. Cases cover success, HTTP404/500, network failure, wrong content type, malformed JSON, wrong/null shape, empty204 body, pre/in-flight cancellation, XHR timeout, body consumption, concurrent calls and loading/retry recovery. Returned markup-like IDs remain literal text. This exercises browser API behavior, not live external transport, deployed CORS, authentication or cross-browser support. A test-harness issue initially invoked the listener-removal function while installing the UI through Playwright; the harness was corrected and the final full browser run passed.

Promise tests distinguish all/any/race on empty input, ordered results/errors, catch recovery, finally replacement, async prefix/continuation and promise identity. A finite sentinel observes empty race pending during the comparison; permanent pending follows the documented contract. Aggregate rejection does not cancel other work; actual request cancellation is tested separately. The small range check of Math.random is not a statistical/security evaluation.

Learning resources and remaining limits

JavaScript Essentials 2 publicly lists 50 hours and four modules, but visible module bullets are not a complete 40-objective lesson crosswalk. No account, lessons or assessments were inspected. Old unverified paid-resource runtime/page-count claims were removed or labeled as author planning budgets. MDN supplies primary technical references; selected javascript.info contents are supplementary navigation.

Eight broader learner activities remain proposed. Direct official-page/monitor failures, current voucher/booking details, paid interiors and human review keep the outcome reviewed-with-blockers. No external message, enrollment, purchase, installation, persistent server/service, reboot, runtime/authentication/permission change or production infrastructure action occurred.

The operation record preserves objective mappings, exact code hashes, original harnesses, execution receipts and selected-reading boundaries. Full unit, repository, catalog and strict generated-site gates are recorded after completion.