From lookup table to cockpit: what the Intelligence Panel actually is.

The job to be done

One question, answered defensibly: how much breathing room does the trail deserve?

The panel exists for a single downstream decision — the ATR coefficient that sets trail width on the trend branches' runners. That number is consequential in both directions: too tight and the trail surrenders continuation trades to ordinary retracement noise; too wide and giveback balloons while the position donates back its excursion. Before the cockpit, the choice was a lookup or a feeling. The cockpit's contribution is process: the same decision, made through the same surfaces, against the same evidence, with the same record — every time. Volatility judgment doesn't get automated away; it gets structured until it's auditable.

The surfaces

Control panel, evidence intake, coefficient model, decision, record.

The workflow crosses five surfaces in fixed order. The Control Panel sets the frame: symbol, trigger timeframe, branch, engine mode (Conservative, Balanced, or Aggressive — which changes tier interpretation, not the evidence), and authority mode. The Input Calculator takes the live VIE readings: ATR, Bollinger width, slope, outer-band persistence, confirmation state, news risk, spread risk. The Coefficient Model converts evidence into a tier and a base VIE coefficient, then applies the authority add-on for timeframe context. The Comparison Panel stages the actual decision — static versus VIE versus override. And the Audit Log records what was actually used, which is the surface that makes all the others honest: a cockpit whose decisions aren't logged is a lookup table with extra steps.

Evidence versus interpretation

Confidence is raw evidence; engine mode is how strictly you read it.

The panel maintains a separation that confuses first-time operators and protects everyone after: VIE confidence does not change when engine mode changes — the manual flags this as expected behavior, not a bug. Confidence reports what the evidence is: how strongly the live readings support a volatility state. Engine mode governs how that evidence maps to tiers: Conservative demands more proof for the higher, wider tiers; Aggressive grants them earlier. The separation mirrors the system-wide doctrine of keeping measurement and posture apart — the same evidence should read identically regardless of the operator's mood, and the operator's chosen strictness should be a visible dial, not an invisible thumb on the readings.

The boundary

A volatility read informs a trail. It authorizes nothing.

The cockpit inherits the volatility layer's standing limits, and they're worth restating because a confident coefficient read feels like more than it is. The panel cannot make a disallowed trade valid, cannot override a restricted gate or System Lock, and its stop distances must size inside throttle-approved risk. It also doesn't decide Normal versus Trend — that's the regime switch's job, covered one page over — and it doesn't fire until trend management is already legitimately in play. What the cockpit governs is exactly one parameter of one management structure, governed well: which is the MARS pattern in miniature — narrow authority, exercised rigorously, logged completely.

• Use the VIP for context first, then the Distance Matrix for exact pip/point translation — the manuals assign the order.
• A favorable volatility read plus a restricted gate resolves the way every such conflict resolves: the gate wins.
• The coefficient decision is captured for review because coefficient-versus-outcome is a standing maintenance check.

The key idea

Structure the judgment; keep the judge.

The cockpit doesn't remove the operator from the coefficient decision — the Comparison Panel exists precisely because static classification and live evidence can legitimately disagree, and someone has to choose. What it removes is unstructured choosing: evidence entered the same way, tiers assigned by declared thresholds, timeframe context applied by rule, and the final call written down beside what the model recommended. Months later, the audit log answers the only question that improves the next decision: when you overrode the model, were you right?