Lab Bloss0m Note 000 This project features specifications and web-based implementations of several Clubhouse Games (51 Worldwide Games). We have standardized the specifications for various classic games and implemented them as independent web pages. Through a unified “Games Overview Menu”, we integrate the entry points, allowing users to experience multiple mini-games directly in their browsers.
The project contains specifications for over 20 games ranging from cards and boards to puzzle and sports games. It includes implementations for classic titles such as Blackjack, Othello, and Klondike.
0. Next phase: One-shot / Few-shot Game Development Center
Clubhouse Games is evolving from a game collection into a public AI game-development lab. The goal is to test whether a model can turn a game specification into something playable, testable, and publishable from one or a few instructions.
Each experiment is intended to record:
- Input conditions: model, tools, initial prompt, reference assets, and shot count.
- Development process: generated scaffolding, interaction logic, asset handling, and revision count.
- Outcome evaluation: rule correctness, playability, completeness, elapsed time, and degree of human intervention.
- Public evidence: playable demo, source code, successful path, failure analysis, and next improvements.
The goal is not to claim that every game can be completed with one prompt. It is to build a comparable record of models, prompting strategies, and Agent workflows.
1. Game Overview and Classification Structure
For ease of management, the games are divided into multiple categories. The source code and specification documents are also arranged according to this structure:
- 01 Cards (4 games): Blackjack, FreeCell, Klondike, Last Card
- 02 Board (3 games): Othello, Checkers, Connect Four
- 03 Tiles and Dice (2 games): Dominoes, Yacht Dice
- 04 Sports and Arcade (5 games): Toy Tennis / Soccer / Boxing / Baseball, Badminton Smash Training
- 05 Puzzle (3 games): Mystery Liquid Sort, Takoyaki, Tetris
- 06 Mini-Games (5 games): Carrom, Slot Cars, Color Guessing, Tank Battle, Samurai Reflex Training
(For detailed specifications, please refer to the specification documents under each category folder in the GitHub Repository)
2. System Architecture and Overview Menu
The core architecture of this project uses a single index.html site homepage as the “Overview Menu”, while all game implementations are independently placed under the Games/ directory.
- Overview Menu: Responsible for displaying the game list and handling entry point redirection.
- Sub-game Implementations: Each game is independently developed and bundled, for example,
Games/Blackjack-main/. - Single Development Server: When developing locally, you only need to start one dev server. The menu and all built games will be served on the same port, eliminating the hassle of managing multiple microservices.
3. Implemented Game Examples
The game examples that are currently built and integrated into the system include:
- Blackjack (
Games/Blackjack-main/) - FreeCell (
Games/FreeCell/) - Last Card (
Games/Last-Card/) - Klondike (
Games/Klondike/) - Badminton Smash Training (
Games/Block-the-smash/) - Samurai Reflex Training (
Games/Instant-Flash/) - Mystery Liquid Sort (
Games/Mystery-Liquid-Sort/)
4. Local Development and Startup Process
The project adopts frontend technologies such as TypeScript, and each game can be built independently:
- Install root directory dependencies:
npm install - Install sub-project dependencies (if a new game is added):
cd Games/Blackjack-main && npm install && cd ../.. - Build the specified game (compile the sub-game output and make it ready):
npm run build:game Blackjack-main - Start the unified development server:
npm run dev - Open
http://localhost:3000to see the overview menu. Click “Enter Game” to seamlessly switch to the built sub-game.
5. Deployment to GitHub Pages
The project supports bundling the menu and all built sub-games, and publishing them to GitHub Pages. The related build processes, CI scripts, and the local build:pages task are documented in docs/DEPLOYMENT.md, ensuring the consistency and convenience of automated deployment.