🔺 Game-OS V2.5 — Decision AI Platform

A trustable decision-making AI platform that turns black-box AI into auditable, explainable, circuit-broken decision pipelines — built around the Decision Cone geometry and the universal 0.68 threshold.

Live: https://hellomind-star.github.io/ymine/ Portfolio: https://hellomind-star.github.io/ymine/portfolio.html Research paper: https://hellomind-star.github.io/ymine/research.html

Game-OS is not a game — it is an operating system for decisions.

Starting from game theory and extending through quantitative finance, cognitive modeling, airspace scheduling, and molecular mixology, Game-OS is built on the conjecture that every decision-making problem, regardless of domain, projects onto the same geometric skeleton: a Cone with a universal risk threshold at 0.68.

• Below the threshold — decisions are rigid, brittle, over-optimized.
• Above the threshold — decisions collapse into chaos, cascade risk, black swans.
• At the threshold — the system breathes: adaptive, auditable, evolvable.

64 fully interactive HTML labs let visitors feel the math instead of just reading it.

                   ┌─────────────────────────────────┐
                   │      Portfolio / Research       │  ← 面向求职 / 学术入口
                   │   portfolio.html · research.html│
                   └──────────────┬──────────────────┘
                                  │
        ┌─────────────────────────┼─────────────────────────┐
        ▼                         ▼                         ▼
┌───────────────┐       ┌──────────────────┐       ┌────────────────┐
│  Control Deck │       │  Core Engines    │       │  Private       │
│  index.html   │◄─────►│  - quant-engine  │◄─────►│  Engines (JS)  │
│  ymine-studio │       │  - mcn-alpha     │       │  - financemind │
│               │       │  - cone-game     │       │  - airmind     │
└───────┬───────┘       │  - triangle-audit│       │  - mindspeak   │
        │               │  - risk-circuit  │       │  - circle-cog  │
        │               └────────┬─────────┘       └────────┬───────┘
        │                        │                          │
        ▼                        ▼                          ▼
┌────────────────────────────────────────────────────────────────────┐
│                         Labs / Evidence (64 HTML)                  │
│  Cone Game · Chromosome · CAPM Pricing · 7D Vector · Sandbox …    │
└────────────────────────────────────────────────────────────────────┘
                                  │
                                  ▼
                   ┌─────────────────────────┐
                   │  Backend (FastAPI)      │
                   │  - chromosome router    │
                   │  - storm simulator      │
                   └─────────────────────────┘

Everything runs client-side by default (zero-dependency vanilla JS/HTML/CSS). The optional Python backend (backend/) powers a few heavier simulations; the Streamlit dashboard (dashboard/) is a separate ops view.

Start here for interviews, demos, or just to get oriented:

Every trade-off (explore/exploit, risk/return, order/chaos) lives on a cone:

• Vertex (0.0) — absolute certainty, dead.
• 0.48 line — capital-preservation floor.
• 0.50 axis — steady-state centerline.
• 0.68 threshold — fuse / circuit-breaker line (derived from φ + safety margin).
• Base (1.0) — pure chaos, maximum entropy.

Risk is not a single number — it's a layered fuse system. When any layer trips, the system degrades gracefully rather than cascading.

Every decision is cross-validated by three independent models before it can be acted upon — no single model ever has unchecked authority.

The Fibonacci–Chromosome isomorphism (see research.html) is the mathematical backbone — explaining why the same 0.68 constant appears across finance, poker, scheduling, cognition, and mixology.

ymine/
├── index.html                   # Control deck (main dashboard)
├── portfolio.html               # Author portfolio / resume page
├── research.html                # Academic paper landing page
├── ymine-studio.html            # Full-feature studio console
├── 404.html                     # Themed 404 page
│
├── assets/
│   ├── css/                     # base.css · components.css · poker-egg.css
│   └── js/                      # bus · components · quant-engine · poker-egg · …
│
├── labs/
│   ├── evidence/                # ★ Core interactive demos (20+ HTML labs)
│   ├── engineering/             # Engineering-domain simulations
│   ├── structural-mechanics/    # Structural mechanics experiments
│   ├── marketing/               # Marketing reinvention demos
│   └── finance-evidence-lab/    # Finance evidence lab
│
├── engines/                     # Domain engines (financemind · airmind · mindspeak · gamemind · …)
├── models/                      # Data models (calc · compute · gamemind · common)
├── game-os-main/                # Core engine modules & business modules
├── isomorphism-block-engine/    # Cross-domain isomorphism block demo
│
├── backend/                     # Optional FastAPI backend
├── dashboard/                   # Streamlit ops dashboard
├── docs/                        # Papers, audits, architecture docs, PDFs
│   └── Fibonacci-Chromosome_Isomorphism.pdf
├── tests/                       # Core engine tests (run in browser)
├── tools/                       # Internal dev tooling
└── video/                       # Video production assets

Design philosophy: zero external dependencies for the public-facing site — every lab loads and runs offline once cached.

The mathematical foundation behind Game-OS is formalized in the working paper:

Fibonacci–Chromosome Isomorphism: A Unified Geometric Skeleton for Cross-Domain Decision Systems — HelloMInd-star, 2026 (7-page IEEE double-column preprint, to appear on arXiv).

• 📥 Download PDF
• 📄 Research page (with abstract, framework, interactive experiments, BibTeX)
• arXiv link pending.

HelloMInd-star builds decision infrastructure at the intersection of game theory, cognitive science, and finance.

• 🎓 MBA (Singapore) · Finance undergrad
• 🎯 2026 AI Product Manager candidate — Game AI / Creative Tech
• 🔺 Game-OS V2.5 architect · Cross-domain isomorphism researcher
• ♠️♥️♦️♣️ Poker Face Arena — 扑克人格竞技场（MBTI 16型AI陪练 · Game-OS V2.5首个公开Demo）
• 🍸 Y.MINE "Before the Exit" — MBTI-personalized cocktail system
• 🎵 400h+/month music listener · Latin jazz + Trap Soul aesthetic

Core thesis: Every decision domain — from asset allocation to poker to mixology — shares the same geometric skeleton. The Decision Cone and the 0.68 universal threshold are not metaphors; they are measurable, auditable, and exploitable.

• 📁 Portfolio
• 📄 Research
• ⭐ GitHub

# Just serve the directory — any static server works
python3 -m http.server 8080
# then open http://localhost:8080/

# Optional backend (for chromosome / storm endpoints)
cd backend
pip install -r requirements.txt
uvicorn main:app --reload --port 8000

MIT — feel free to fork, study, and build on top. If this work helps your research or product, a citation or a star is always appreciated.

_{KMP → IPD → 7D Vector → 0.68 Cone · Triangle Audit · Seven-Layer Fuse · Cross-Domain Isomorphism}