Note: Copilot organized and set up this README for me. I'm still going through everything, evaluating it, and making edits where needed. Feel free to check the linked screenshots for verification. Right now this is mostly a temporary template to help me organize the project.
This repository contains architecture documentation, prompt frameworks, and validation materials for an AI-assisted TCG analysis framework built around the UROE v2 and H.E. design architectures.
The framework is designed to improve analytical consistency through structured evaluation layers, verification rules, state tracking, and context re-anchoring mechanisms. Included materials document deck analysis testing , complex MTG interactions, and framework validation examples.
This is an active development project. Documentation, testing, and framework components continue to evolve as new scenarios are evaluated and refined.
**A structured reasoning framework for AI-assisted trading card game analysis. The framework uses layered evaluation, verification rules, state tracking, and context re-anchoring to reduce drift and improve analytical consistency. Future versions are intended to integrate external structured databases to further reduce hallucinations and improve reproducibility.
The framework is:
- โ Designed to operate with modern AI systems
- โ Tested on complex MTG interactions and deck analysis scenarios
- โ Built around deterministic workflows, verification rules, and re-anchoring mechanisms
- โ Open source (MIT License)
- โ Ready for implementation and expansion
4 validated test cases demonstrating UROE v2 functionality:
| Test | Result | Evidence |
|---|---|---|
| System Initialization | โ PASS | View Screenshot |
| MTG Game Profile Recognition | โ PASS | View Screenshot |
| Complex Engine Classification | โ PASS | View Screenshot |
| UROE v2 Edge Case Resolution | โ PASS | View Screenshot |
๐ View Full Proof of Concept
Additional validation screenshots:
- Extended Test Trace 1
- Extended Test Trace 2
- Gemini Session 2 - Part 1
- Gemini Session 2 - Part 2
- Gemini Session 2 - Part 3
- Gemini Session 2 - Part 4
- UROE v2 Resolution Engine โ Deterministic resolution cycle: [S] โ [N] โ [I] โ [M] โ [W/V/T/B/H/C] โ [STATE UPDATE]
- Core Layers Framework โ R, A, C, N, M, Z, I, H, B, V, T framework
- Edge Case Library โ Game-specific behavior patterns
- Chat Bot v1 โ Universal card game model with resolution engine
- Chat Bot v2 โ Enhanced analysis directives
- Dual Personality Framework โ Analytical + Coaching AI modes
- Proof of Concept โ Validated test results with Gemini
- Thinking Model
- Teaching Model
- Load the system directive from
prompts/chat-bot-v1.mdinto Claude, GPT, or Gemini - Provide a game state and card interaction query
- AI analyzes through UROE v2 cycle and returns layer-by-layer resolution
Example Use Case:
Query: "I have Brainstorm in hand with 1 Blue Mana. Opponent has Chains of
Mephistopheles in play. What happens if I cast Brainstorm?"
Framework Response: [Full UROE v2 cycle trace with exact [S] โ [N] โ [I] โ [M]
sequence and impact on [Z], [H], [V], [C] layers]
The framework provides the architecture for building a full TCG analysis engine:
- Backend: Implement UROE v2 resolution logic in Python/TypeScript
- Game Profiles: Define MTG, Yu-Gi-Oh, Pokรฉmon rules as structured data
- Database: Store card definitions and interaction patterns
- Frontend: Web app or Discord bot for player queries
โ Multi-Game Support
- Magic: The Gathering (tested)
- Yu-Gi-Oh (architecture ready)
- Pokรฉmon TCG (architecture ready)
- One Piece TCG (architecture ready)
โ Advanced Resolution
- Stack/Chain handling
- Replacement effects
- Missed timing detection
- State-based effects
- Simultaneous triggers
โ Teaching Modes
- 6-stage progression from Foundation โ Meta Awareness
- Beginner-friendly explanations
- Advanced technical analysis
AI-PoweredTCGAnalysisFramework/
โโโ docs/
โ โโโ proof-of-concept.md โ Test results (START HERE)
โ โโโ assets/ โ Screenshots & evidence
โ โโโ [core specifications]
โโโ prompts/
โ โโโ chat-bot-v1.md โ Universal framework
โ โโโ chat-bot-v2.md โ Enhanced directives
โ โโโ dual-personality-framework.md
โ โโโ stress-proof-edge-case-library-v5.md
โโโ README.md โ You are here
โโโ CONTRIBUTING.md
โโโ LICENSE (MIT)
This project is in active development. We're looking for:
- Backend implementation (Python/TypeScript)
- Game rule database architecture
- Frontend (web app / Discord bot)
- LLM integration & prompt optimization
- Game profile definitions (MTG, Yu-Gi-Oh, Pokรฉmon, One Piece)
- Edge case validation
- Test case generation
- Rule accuracy review
- Test the framework with your favorite TCGs
- Report findings & edge cases
- Share use cases & ideas
See CONTRIBUTING.md for details.
Read: UROE v2 Overview
Read: Proof of Concept Test Results
Read: Chat Bot v1 - Complete Framework
- Load
prompts/chat-bot-v1.mdinto Claude/GPT/Gemini - Ask it to analyze a complex card interaction
- See the UROE v2 cycle trace
| Feature | Status | Details |
|---|---|---|
| Universal Framework | โ Complete | Works across all major TCGs |
| Deterministic Resolution | โ Validated | No hallucination, reproducible |
| Edge Case Handling | โ Tested | Handles complex interactions |
| AI Integration | โ Ready | Prompts provided for Claude/GPT/Gemini |
| Multi-Game Profiles | ๐ In Progress | MTG complete, others architected |
| Backend Implementation | โ Needed | Seeking developer partner |
| Frontend UI | โ Planned | Web app / Discord bot |
MIT License โ Free for personal, commercial, and research use.
- Technical questions? Open an issue or start a discussion
- Want to contribute? See CONTRIBUTING.md
- Looking to implement? Reach out โ we're building a team
If you use this framework in your project, please reference:
TCG Analysis Framework (UROE v2)
Repository: [https://github.com/oKarismatic/TCGAnalysisFramework]
Framework: Universal Resolution Order Engine v2
License: MIT
(Original README content preserved. Last updated: 2026-06-05)