MODULE 02 · LESSON 4 · 7 MIN
The Fleet: the Second Opinion
Goal: Explain what the multi-agent fleet adds on top of signals, and what it never does.
Alongside the indicator, Novex runs a public multi-agent fleet — a pipeline of AI analysts that stress-tests market state from independent angles and publishes its read.
The pipeline shape
A technical engine reads structure first. Three analysts then work the angles — sentiment, flow, macro. Two debaters argue the bull and bear case against each other on purpose. An aggregator weighs the room, and a risk gate gets the final veto. The output is a published stance: aligned, split, or opposed — with the reasoning visible.
What the fleet is for
The signal engine is deliberately narrow: nine dimensions of chart truth, no narratives. The fleet is deliberately wide: it reads the things a chart cannot — crowd mood, news flow, macro context. When a structural signal fires AND the wide read aligns, you have two independent systems agreeing from different evidence. When they disagree, you have a flag worth respecting: the chart says one thing, the room says another.
Transparency is the point. The fleet publishes its calls in public, right or wrong. You grade it on its record, not its confidence.
What the fleet never does
It never places trades, never overrides a gate, never turns a STANDARD into a LEGENDARY. It is commentary with structure — a second opinion from a different discipline. Your decision framework (M03) decides what disagreement means for size, or for skipping entirely.
Takeaways
- ▸ The fleet reads what charts cannot: sentiment, flow, macro — published transparently.
- ▸ Signal + fleet alignment = two independent systems agreeing from different evidence.
- ▸ The fleet is commentary, never execution, never a gate override.
Checkpoint
Answer all questions correctly to complete the lesson. Misses reset for another pass — no penalty.
1. The fleet's role relative to signals:
2. Fleet opposes a fired signal. Correct posture:
3. Why publish fleet calls publicly?