AI Battleground ⚔️

Project Dashboard

🎯 Current Sprint Goals (July 14 - July 28)

This sprint, we will make a basic UI playable Godot game with static data, and start de-risking Game Design decisions, UI/UX wireframes, and deck generation via on-device LLMs.

1. Design & Art Team Outcome:

A V1 'Lore & World Guide' document. A detailed design for the 'Traits' system and a concrete progression path for the 'Lazeraptor' static deck, including all card and trait upgrades.

2. Godot Game Team Outcome:

A playable combat loop using our two static decks. A player can roll dice, see valid cards based on the result, play a card, and see the turn resolve against an opponent's action using prototype UI.

3. AI/Python Team Outcome:

The on-device Rust model is runnable via a new Python wrapper. A functional HTML prototype (`Python-Mimic`) of the deck generation UI exists, including an optional image generation step.

4. Product & UI/UX Team Outcome:

A complete Figma flow documenting the core combat screens from start to finish. Stretch goal: matchmaking, post-battle rewards, and creature inventory screens.

5. Marketing Team Outcome:

An initial plan for Alpha/Beta user acquisition. A documented list of 'key gameplay moments' suitable for creating early marketing assets (GIFs, short videos).

🔑 Key Areas of Investment

Player-Driven Novel Decks

  • Satisfying Coherence: Generated decks feel like a witty, surprising interpretation of the player's prompt, not a literal translation.
  • Strategic Viability: Every generated deck is playable out-of-the-box, offering novel synergies between its cards, trait, and dice.
  • High Replay Value: The variety of outputs consistently creates new gameplay puzzles, compelling players to create more creatures.

Strategic Depth

  • Meaningful Choices: The simultaneous-reveal system consistently creates dramatic anticipation and satisfying resolution.
  • Resource Management: Custom dice pools and Arena Effects present players with interesting, non-obvious spending decisions each round.
  • No Stale Metas: Headless evaluation and continuous tuning prevent dominant strategies from emerging, ensuring a dynamic play environment.

Compelling Progression

  • Rewarding Core Loop: Every match provides a satisfying amount of Upgrade Currency (UC), making consistent progress feel attainable.
  • Aspirational Creation: Creature Creation Currency (CCC) is scarce enough to feel valuable, making the choice of what to create next a meaningful one.
  • Visible Power Growth: Upgrading a creature provides a clear, noticeable improvement in its performance and Power Rating.

📊 Current Status (July 2025)

AI Generation

  • LLM deck generation is functional but cloud-only.
  • On-device model (Rust) runs on laptops; exhibits context-window issues on mobile testbeds.
  • Two static creature decks exist for internal testing.

Godot Game

  • Headless combat prototype is operational in Godot 4.
  • P2P multiplayer lobby and handoff are functional.
  • Authoritative host logic is implemented but not cloud-tested; no matchmaking service exists.

Design & Art

  • Core GDD is written, but progression, traits, and economy systems are high-level concepts only.
  • Art direction is still not agreed upon; artists are in research phase.

Product & UI/UX

  • All current UI is prototype-level implementation.
  • No formal design documentation (e.g., Figma) for game flows exists.

📚 Alpha Milestone Backlog

This backlog represents the full known scope of work required to complete the Alpha milestone. It is a living document and will be updated as new requirements emerge.
Priority Legend: 90-100 (Current Sprint) | 70-89 (Next Sprint) | <70 (Future Sprints)