fraserscottmorrison/gameengines — explained in plain English
Analysis updated 2026-05-18
Run a simulated Carcassonne-style board game between AI personas from the command line
Watch turn-by-turn logs to see how tiles, roads, and cities get scored
Study a fully AI-generated Python game engine as a coding example
Play or continue a game session by asking an AI assistant through the bundled skill
| fraserscottmorrison/gameengines | 0xallam/my-recipe | 0xhassaan/nn-from-scratch | |
|---|---|---|---|
| Stars | 0 | — | 0 |
| Language | Python | Python | Python |
| Last pushed | — | 2022-11-22 | — |
| Maintenance | — | Dormant | — |
| Setup difficulty | easy | moderate | moderate |
| Complexity | 2/5 | 2/5 | 4/5 |
| Audience | vibe coder | general | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires Python 3.10 or newer, no external packages needed.
GameEngines is a text-based digital version of a tile-placement board game very similar to Carcassonne, built entirely by an AI language model. The author describes reading the rulebook, analyzing photos of the physical tiles, and writing all of the Python code through AI prompting alone, without personally writing any of it by hand. The game engine runs from the command line and simulates a full game between computer-controlled players. Ten different AI player personas are included, each with its own difficulty level and playing style, ranging from a cautious novice who builds small roads to an advanced player focused on blocking opponents. Players are set up as a comma-separated list when starting a new game, and the engine automatically draws tiles, places them, adds meeples where useful, and scores completed roads, cities, and monasteries turn by turn. Every game session is saved to its own folder containing a JSON file with the full game state and a readable Markdown log recording each turn's tile, placement, and score changes. A short summary table is produced at the end of a completed game showing final rankings and scores. The engine can be driven turn by turn, several turns at once, or all the way to completion in one command, and there are additional commands for checking legal moves, current scores, and the state of the tile deck. The project can also be played through an AI assistant using a bundled skill, letting a user simply ask the AI to start or continue a game rather than typing commands directly. The tool requires Python 3.10 or newer and has no external dependencies. As the author notes plainly, the game currently has no visual interface, no option for a human player to join, and the computer opponents do not always play well, especially at the higher difficulty settings.
A command-line, AI-built simulation of a Carcassonne-style tile placement board game played entirely by computer personas.
Mainly Python. The stack also includes Python.
No license information is provided in the repository, so default copyright rules apply and reuse permissions are unclear.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly vibe coder.
This repo across BitVibe Labs
Don't trust strangers blindly. Verify against the repo.