Eleven workbooks, three families: who does what in the ecosystem.

The sorting principle

A module's family is defined by what its output may touch.

The ecosystem's taxonomy isn't about what a workbook computes — it's about what its computation is permitted to do. Evidence-family outputs enter the record and inform interpretation, but bind nothing: a weekly EV read, a regime label, an efficiency grade. Authority-family outputs bind capital directly: a gate state caps tiers whether or not anyone agrees with it; a throttle resolution is the deployment envelope, not a suggestion. R&D-family outputs touch nothing at all until promoted: a scenario result, however brilliant, has no operational surface. The same metric can appear in all three families at different privileges — EV as evidence in the scorecard, as a tactical selector inside the throttle's authority chain, and as a sandbox variable in the lab — and the privilege, not the metric, is what the family system tracks.

The three families

The roster, sorted by privilege.

The authority family is deliberately small: the Gate/Drawdown layer (capital-state routing, the hard ceiling, System Lock), the Throttle Control Panel with its 7-Tier Decision Engine (final tier, pool, per-trade risk), and the Trade Launch Checklist's narrow blocking power. The evidence family is the largest: Compliance Panel 3 as system of record, the Weekly Trading Scorecard as expectancy engine, the Cycle Command Console preparing cycle facts and exposure, the MAE/MFE lab grading execution, the Volatility Intelligence Panel and Distance Matrix informing stop and trail decisions, the SDE and Regime Classification Engine diagnosing structure, and the Monte Carlo benchmark supplying the ruler. The R&D family holds the Monte Carlo Lab itself — the ecosystem's newest workbook, where assumption sets are stress-tested and benchmark runs are built — alongside the Advanced EV Analytics Lab's sandbox, the Scenario Manager, and every sensitivity tool: production-shaped instruments with zero production power. Note the deliberate split: the MC benchmark tables the comparison workflow consumes sit in the evidence family as the ruler, while the lab that generates and re-generates them is R&D — the same distinction as everywhere between a published instrument and its experimental workshop.

Why authority stays small

Binding power is expensive to trust, so it's concentrated and audited.

The asymmetry in family sizes is a design statement. Every module granted binding power is a module whose errors reach capital directly, so the authority family is kept minimal — routing, resolving, and blocking, nothing more — and each member's power is scoped to one verb. Meanwhile the evidence family grows freely, because informative modules fail safely: a wrong efficiency grade misleads a review; a wrong gate state mis-sizes a cycle. This is also why promotion from evidence to authority essentially never happens — the SDE will never deploy, the benchmark will never permit — and why new capability lands in the evidence or R&D families by default. The system gets smarter by adding eyes, not hands.

• Authority modules own one verb each: route (gate), resolve (throttle), block (checklist). None owns two.
• Evidence modules may be rich, opinionated, and wrong — their failure mode is a bad interpretation, not a bad deployment.
• R&D modules may be reckless by design — the wall makes their recklessness free.

Reading the ecosystem in practice

Family-first is the operating habit.

The taxonomy earns its keep as a reading habit: before reacting to any output, place it. A crimson gate cell is authority — it has already acted, and the only question is compliance. A deteriorating branch EV is evidence — it informs the weekly read and binds nothing until the authority chain consumes it through the throttle's inputs. A spectacular scenario comparison is R&D — it proposes, and the promotion gauntlet disposes. Most operator confusion about 'what the system is telling me to do' dissolves under this habit, because the system's outputs come pre-sorted by exactly that question: authority tells, evidence informs, R&D asks. The workbook count stops mattering once the families are internalized.

The key idea

The ecosystem is organized by power, and that's what makes it one system.

Eleven workbooks could be a pile; what makes them an ecosystem is a privilege structure every module respects — a small core that binds, a broad base that informs, and a sealed wing that explores. Family membership answers the first question about any number before it's asked, keeps binding power rare and auditable, and lets the evidence and R&D families grow without ever making the system more dangerous. Count the modules if you like. The architecture is in the columns.