Quality Assurance · The Principles

From built to provable.

Core concepts

Inspection tabs

First test

TEST 0

The key idea

Test zero

The cockpit is audited before it is allowed to audit anything.

Before the workbook certifies a production workbook, it validates itself: dropdowns, formulas, matrix routing, expected-output linkage, error counts, retest status, and readiness counters. The instruction is written as a do-not-skip because the failure it prevents is the most expensive one available. A quality cockpit with a broken counter will certify a broken workbook confidently, and the certificate will be indistinguishable from a real one.

Maturity gating

Silence is a correct output.

The hardest thing to test is a workbook that should be showing nothing. A no-data baseline check exists specifically to prove that blank conditions do not trigger fake analytics, false colours, stale dashboard values, future-period activation, or artificial structural states. An output that fills itself in before the evidence exists is not merely early; it is the mechanism by which a one-week sample gets read as a twelve-week structural finding.

Quarterly and annual blocks must stay inert until mature, and must not contaminate immature periods.
Charts and visuals must not plot fake zeros, stale data, or future periods.
Future-week and future-month silence is inspected as its own expected behaviour.

Source before mirror

Audit the truth layer first, always.

The inspection order is fixed: journal, weekly summary, expectancy scorecard, monthly rollup, and gate state are audited before any dashboard mirror. Testing a display tab first lets a downstream panel mask an upstream error, because a mirror faithfully reflecting a wrong number looks exactly like a mirror working correctly. This ordering rule shows up again in the trading workflow, where efficiency grading runs only after journal reconciliation, for the same reason.

Identity is permanent

A scenario identifier is audit identity, and is never repurposed.

Scenario identifiers are the join key connecting expected outputs, test runs, the error register, the retest log, and every inspection tab. Reusing one for a different scenario silently rewrites the history of every record that referenced it. When a variant is needed, a new instance is created under the same family rather than an existing identifier being edited, which is why the numbering scheme separates the archetype, the family, and the instance.

In practice

Six structural inputs are scored monthly into a composite, and the composite maps onto the regime ladder. MARS reads results in context: the same numbers mean different things earned in Expansion tailwind versus survived through Instability, and posture is adjusted to the regime rather than to the last trade.

Twelve months classified across the regime ladder - the system's own weather record, from Expansion down toward Drawdown Control.

Reference

The five concepts the workbook is built on

Concept	Meaning	Why it matters
Black-box testing	Test from visible inputs and outputs, assuming nothing about the formulas	Prevents false confidence from formulas that calculate but behave wrongly
Scenario contract	A controlled test condition with explicit expected behaviour	Turns testing into objective comparison instead of opinion
Maturity gating	Outputs activate only when the timeframe has enough evidence	Stops a one-week sample pretending it can prove annual behaviour
Source-of-truth routing	Every output traces back to the correct truth layer	Stops dashboards inventing values or mirroring stale rows
Error lifecycle	Failure, register, fix, retest, close	Forces repair proof instead of an assumption that it is fixed

Before you go deeper

Connected inside MARS

This module doesn't work alone.

Connected — 01

Certified Validation QA (full module)

Open system page →

Connected — 02

The Audit Flow

Open system page →

Connected — 03

Compliance Panel 3

Open system page →

Go deeper

Operator briefs on this territory.

Deep dive — 01

Maturity gating: an output that cannot yet be true stays dark.

One week cannot prove a quarter. So the quarterly output does not appear until a quarter exists.

Read the full brief →

Deep dive — 02

Every output must name where its truth came from.

A panel that invents a value and a panel that reports one look identical. Routing is what separates them.

Read the full brief →

Deep dive — 03

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

Formulas that compute can still be wrong. Black-box testing checks behavior, not arithmetic.

Read the full brief →

Every module ships in the complete MARS package.

One price. Eleven workbooks, three TradingView indicators, and the full manual library — $497.

Get MARS Pro — $497 Explore the Platform