A playable system, not a mock-up
Arena Rubra began as a tactical card-game idea and grew into a working digital rules environment. Matches take place on a hexagonal battlefield where cards, positioning, strategic points and faction-specific mechanics all interact.
The current public status is deliberately simple: Demo — Work in Progress. The project is playable and substantial, but it is still being iterated, tested and refined rather than presented as a finished commercial release.
What the demo already explores
Tactical play
Units are deployed onto an interactive hex map and fight for position, strategic points and pressure. Cards are not detached from the battlefield: movement, targeting, defensive states, abilities and territorial control all belong to the same turn-based system.
Cards, decks and content tools
The application includes a card library, deck-building tools and structured card data. Cards can be inspected as rendered objects while their gameplay statistics and ability text remain available as explicit data.
Maps and terrain
Arena Rubra is not tied to a single board. Its map system supports different layouts and terrain rules, while the technical Map Editor provides validation, import/export and controlled editing instead of treating custom maps as opaque artwork.
Guided learning
Tutorial scenarios use structured steps, narrator overlays and controlled interactions to teach the game without relying on a wall of rules text.
Built through iteration
The interesting part of Arena Rubra is not only the visible game. The project is developed through small baselines: a feature is introduced, tested, inspected, corrected and only then treated as stable enough to build on.
Bot-vs-bot matches, manual playtests, telemetry, runtime logs and dedicated diagnostic screens are used to expose behaviour that is difficult to judge from code alone. That makes the project a useful example of the broader GVibeDev approach: AI can accelerate implementation and analysis, while design choices, acceptance criteria and validation remain human responsibilities.
AI-assisted, human-directed
AI is used throughout development as an engineering and production partner: to reason about systems, propose implementations, investigate regressions and help turn requirements into code.
That does not remove the need for direction. Rules still have to be chosen. Builds still have to be tested. Unexpected behaviours still have to be reproduced. Features still have to earn their place.
Arena Rubra is therefore less interesting as a claim about what a prompt can generate, and more interesting as a record of what becomes possible when prompting is only one part of a disciplined iterative workflow.
Project status
This page documents the project as it exists during active development. Screenshots are taken from development builds and may change as the interface, balancing, assets and systems continue to evolve.
The source repository is public and remains the reference point for the code currently shared outside the local development archive.
