Cognitive OS  →  Demo 2  →  Judgment to Decision

How judgment
supports decision
without being pushed by emotion

You researched a topic for three months. All your theses verified. One day it suddenly explodes — inner voice says "buy now." You almost did. Then you paused.

The real difference isn't how smart AI is. It's whether the rules you wrote when calm are wired into an "auto-scan pipeline" to remind you when emotion takes over. Below: how a judgment transforms from static conclusion into executable protocol.

STARTING POINT · THE PROTOCOL YOU WROTE WHEN CALM

Not AI's idea · A rule you defined for "future-self under emotional pressure"

07 Methodology / Investment / 05 Implementation / spike-pause-buying-protocol.md
§ 4.5.9 · Emotion-Trigger Pause Protocol
Origin: a real "almost-FOMO" feedback. When calm, you realized — when you buy on a day the price spikes, you're mostly buying FOMO, not the thesis.
Next day, you crystallized that reflex into a named protocol.
Trigger conditions (any one fires)
  • Single-day gain ≥ 10%
  • 5-day cumulative gain ≥ 25%
  • Approaching 52-week high
  • Major catalyst day
Action: Pause buying for 1-3 trading days · Write latest-spike-event field · Wait for ≥ 8-15% pullback before re-evaluating
ONE DAY SIGNAL FIRES · TWO PATHS COEXIST

Same situation · Same you · Two systems = completely different decision paths

Without protocolized system
Emotion-driven decision path
①09:30 · Ticker you researched opens with limit-up +10%. Push notifications. Group chats explode.
②09:45 · You open the research doc, confirm thesis still holds. "Yes — this is what I said!"
③10:00 · Price hits +15%. Someone in group: "just added." FOMO escalates.
④10:15 · You buy. "I did the research. The thesis is right."
⑤3 days later · Pullback -12%. "But the thesis is right? Why is it dropping?"
Problem: "thesis is right" and "should buy now" are not the same thing. You bought certainty-as-illusion that the current move provided, not the thesis.
With protocolized system
Protocol-driven decision path
①08:00 · Auto-scan finds single-day ≥ 10%. § 4.5.9 protocol conditions met.
②08:30 · Synthesis report auto-writes: latest-spike-event field + "Pause buying 1-3 days" alert.
③09:30 · You open screen, see the alert. Not AI's recommendation — your own rule from when you were calm.
④10:00 · Price hits +15%. You feel "buy now", but the protocol checklist is right there. FOMO blocked by a named rule.
⑤3 days later · Pullback -12%. Per protocol: wait for ≥ 8-15% pullback. Cold-window arrives.
Difference: the present "buy or not" doesn't require your willpower to fight FOMO. The rules you wrote when calm made the choice for you.
FROM JUDGMENT TO PROTOCOL UPGRADE · COMPLETE LOOP

Decision isn't the end · Feedback re-entering the loop is what makes the system permanently smarter

1
Protocol triggers
Auto-scan finds condition met
2
Decision log
Cognitive snapshot (immutable)
3
Execute + feedback
Review against expectation
4
Protocol upgrade
Feedback → methodology revised

Every protocol trigger + execution produces feedback into 01 Drafts/Feedback/. After distillation this feedback updates the methodology file — e.g. if "single-day ≥ 10% combined with fundamental breakthrough" turns out to be a new scenario, the protocol automatically adds a new conditional branch. The same mistake only happens once.

PROTOCOL UPGRADE · DIFF AFTER FIRST FEEDBACK

Real example: from v3.2.1 to v3.2.2

Trigger conditions · protocol diff

Single-day ≥ 10%
5-day cumulative ≥ 25%
Approaching 52-week high
Major catalyst day
+ Spike-day accompanied by fundamental breakthrough signal (added in v3.2.2)
+ Cross-topic consensus-stoploss trigger same-direction stacking (added in v3.2.4)

Trigger actions · protocol diff

Pause buying 1-3 trading days
Write latest-spike-event field
+ Extend cooling to 5 days when double-trigger stacked (added in v3.2.4)
+ Auto reverse-append evidence to all linked Cognition nodes (added in v3.2.4)

The protocol isn't written once and frozen. Every real trigger + feedback makes the rule finer, more accurate, more fitted to your judgment habits.
The protocol 90 days from now is not the same rule as the protocol today.

VS LLM WIKI

Same decision scenario · two systems can give you different things

What LLM Wiki gives you
A continuously updated research synthesis
✓ When prices jump, LLM Wiki gives you the latest synthesis of all research, historical precedents, expert views.
✓ It would point out "in this scenario markets often reverse", "Expert X says wait for pullback".
✗ But your in-the-moment decision is still you alone vs FOMO. More knowledge ≠ more calm.
Missing: Where is "the protocol I wrote when calm"? Who pushes it in front of me when emotion takes over?
What Cognitive OS adds
A protocol layer that executes rules for you
✓ The § 4.5.9 protocol you wrote when calm isn't "reference material" — it's a program scanning every ticker every day for rule matches.
✓ The moment a condition hits = protocol auto-fires + Synthesis writes "pause buying N days", visible the instant you open your screen.
✓ This isn't AI making the judgment for you. It's your past self making the judgment for your present self — AI just pushes the rule into view.
Difference: LLM Wiki helps you know more. Cognitive OS helps you obey the rules you set for yourself.
The real moat isn't "not making mistakes"
It's "every almost-mistake becomes a protocol"
Same mistake only happens once · That's the compounding of judgment quality
About this Demo: the dual-path comparison and protocol upgrade diff are abstracted from real methodology evolution. All specific details (ticker codes, position percentages, operation dates) have been anonymized. This page demonstrates how judgment becomes executable protocol and how protocols get upgraded by feedback, not investment advice.