Skip to content
Montex Alpha Rail System
$497
← Back to Quality Assurance

Operator brief · 106

Black-box testing: proving the workbooks behave, not that the formulas compute.

The key idea

The doctrine

Test from the visible inputs and outputs, assume nothing inside.

Black-box testing means auditing a target workbook through what it takes in and what it puts out, without assuming the internal formulas are correct — and the guide states the reason directly: it prevents a false sense of security from formulas that calculate but produce wrong operating behavior. The distinction matters because Excel is exceptionally good at producing plausible wrong answers. A formula referencing a stale range, a lookup keyed on a floating-point value, a rollup silently including future blank rows: all of these compute cleanly, display convincingly, and are wrong. Only a controlled input with a pre-stated expected output catches them.

The scenario contract

A controlled condition plus explicit expected behavior.

The QA workbook's unit of testing is the scenario contract — a controlled test condition with explicit expected behavior — and the guide names precisely what it buys: it turns testing into an objective comparison instead of opinion. This is the same contract doctrine the Scenario Manager applies to strategy questions, and it's no accident the two share a definition: in both cases, stating what should happen before the run is what prevents the run's results from redefining what was being tested. An expected-output row is required to be specific, observable, source-backed, maturity-aware, and testable — because, as the guide puts it, vague contracts produce vague audits.

FigureThe QA spine — eligibility through certification
built becomes provable ↓Test Matrixis this scenario/timeframe combination valid?Scenario Librarystable Scenario IDs for reuseExpected Outputsthe contract — what should happenTest Runs & CHK tabsexpected compared to actualError Register → Retest Logfailure, fix, proofMaster Controlreadiness, derived not asserted

The workbook's primary workflow. Each stage produces a stable ID the next stage consumes, which is what makes a failure traceable from discovery back to the exact scenario and contract that exposed it.

The two gates

Maturity gating and source-of-truth routing.

Two concepts do most of the doctrine's protective work. Maturity gating means outputs only activate when the timeframe has enough evidence — the guide's phrasing is that it prevents a one-week sample from pretending it can prove monthly, quarterly, annual, or twelve-week structural behavior. It's the blanks-not-zeros rule from the SDE, generalized into a testable property: under a one-week scenario, the twelve-week structural outputs must be blank, and a workbook that fills them has failed regardless of what numbers it produced. Source-of-truth routing means every output must trace back to its correct truth layer — journal, weekly summary, EV scorecard, gate state — which stops dashboards and derived panels from inventing values or mirroring stale rows.

What gets tested

The deepest target is the pipeline everything else depends on.

The workbook audits the major MARS families — Compliance Panel 3, the Weekly EV Scorecard, the Throttle Control Panel, and the Cycle Command Console — with CP3 as the deepest target, because it contains the main journal to weekly summary to EV scorecard to dashboard to SDE, regime, and gate pipeline. The prioritization follows dependency: a defect in the journal's rollup propagates into every downstream engine, while a defect in a leaf display affects one panel. The CHK tab inventory reflects this, running from journal inputs and weekly summary through the SDE pipelines, contradiction engine, downstream dependencies, and charts — testing not just each layer but the connections between them.

The key idea

The workbook's purpose is moving the system from built to provable.

That phrase from the operating doctrine is the whole argument. A built system is one where everything appears to work; a provable system is one where controlled scenarios have demonstrated it behaves correctly, and the demonstration is on record. For a stack this size — eleven workbooks whose outputs govern real capital — the difference isn't academic. Black-box testing, scenario contracts, maturity gating, and source routing are the machinery that turns confidence into evidence.

Connected inside MARS

Quality Assurance →
Scenario Manager →
Compliance Panel 3 →

Every brief documents the same shipped system.

The complete MARS package — eleven workbooks, three TradingView indicators, the full manual library — $497.

Get MARS Pro — $497Explore the Platform