MongoDB Associate Developer deep review — September 29, 2026

The guide now answers 40 original prompts and includes 38 executed Python prediction checks, an original shell exercise checked for JavaScript syntax, three worked scenarios and eight expanded proposed activities. The retained 18 objective statements remain a historical baseline; current enrolled scope was not confirmed.

Scope and access evidence

The public exam page still displays 53 multiple-choice questions, 75 minutes, English, online proctoring, no prerequisite, USD 150 and five language variants. Direct automated requests return application shells; browser reading supplied that contract. The canonical study-guide landing advertises a free 30-minute guide and enrollment. Its publicly linked viewer failed with HTTP 500, and the document asset published by that viewer returned 403. No document content, login, enrollment or account action was obtained. Existing objective/status snapshots were preserved, and the automated monitor remains manual review. The public program-guide landing was read, but its policy interior was not.

Original examples and actual execution

The Python program uses standard-library data structures, Decimal and exhaustive enumeration. Separate array predicates admit A/B; a same-element predicate admits B. Tenant filtering excludes the other tenant. Scalar null/missing/existence fixtures are deliberately narrower than MongoDB's array and dotted-path behavior. Pagination examples show both an offset repeat after an earlier insertion and a keyset repeat after a mutable sort key changes.

Three open orders include one empty array. Flattening produces three lines from two represented orders, totaling 8,250 cents. The guide separates row grain, distinct order count and arithmetic. It also distinguishes Python decimal calculations from BSON codec testing.

For stock of five and two requests for four, four of six read-before-write schedules accept eight units while storing a remaining balance of one. Conservation exposes the error that a nonnegative-balance check misses. An indivisible guarded model preserves stock in both actor orders. That indivisibility is a stated model assumption, not a measured MongoDB concurrency result.

Typed input rejects nested operators, Boolean/negative quantities and client-supplied tenant scope. A sequential replay model returns recorded outcomes and rejects a changed payload under the same request ID. It caches declines by choice and is neither durable nor concurrent. Production receipt/state coordination, identity enforcement and ambiguous network recovery remain untested.

All 38 Python checks passed. Node parsed the original shell example without executing it. Expected database observations are marked as predictions. No MongoDB server/shell/driver, BSON codec, Atlas account, index plan, network or paid credit was used. Eight activities and independent human review remain pending.

Technical reading and discrepancies

Current primary references support the explanations of atomicity, element matching, document/type rules, projection, upsert/result semantics, unique/multikey indexes, ESR/ERS alternatives and PyMongo pools/cursors. Reading limits are recorded per source; long update/release references were reviewed in selected sections. Interactive query pages sometimes omitted their main examples, so indexed primary text or a separately readable current document reference supplied the stated rule.

The BSON type page contains an erroneous displayed addition result; the local original check confirms 0.1 + 0.1 is 0.2. A multikey whole-array discussion displays a scalar operand despite describing exact-array equality. Neither inconsistent example was copied. The guide teaches independently stated predicates and requires a real server run. OWASP's constrained-query guidance is supplemented with explicit scalar/schema validation; a query object or a dollar-sign string check is not accepted as a complete defense.

The 9.0 release opening labels patch entries Upcoming while the manual has current/preview wording. The guide therefore does not infer general availability, deployed compatibility or a changed certification scope. Selected PyMongo 4.18.1 through 4.13 entries supply release awareness without an installation or blanket upgrade claim.

Learning catalog

The Python path has a 20-hour headline. Visible required cards total 765 minutes and elective learning 210 minutes: 16h15 of learning. A separate 75-minute exam card makes 17h30 including the exam; neither reconciles to 20 hours. The guide shows these measures separately and records the advertised completion discount without claiming eligibility.

Pluralsight lists seven courses and four labs, with 504 course minutes and 142 lab minutes: 10h46 against an 11-hour headline. Public cards span May 2025 to September 21, 2026. No lesson quality or current exam alignment was inferred. The practice catalog lists one hour for four variants; the Node.js card omits a duration. O'Reilly and Udemy returned 403, so earlier metadata remains explicitly unverified.

Thirty direct source receipts record 28 HTTP successes and two blocked requests. Several successes contain only shells; browser/indexed readings are documented separately. The additional failed public document attempt is retained as a separate access receipt. No assessment questions or paid lesson interiors were used.

The operational record is ADLC_Docs/operations/2026-09-29-mongodb-associate-developer-deep-review.json. It preserves prior records, objective mapping, unchanged historical snapshot hashes, source/access boundaries, catalog arithmetic, public-code hashes and validation results. The five reserved GitHub guides and disabled notification pilot remain untouched.