Pyjamada

A tiny domestic adventure where getting ready in the morning can become a complete disaster.

Pyjamada is an Android-first React Native game experiment built around a deliberately small systemic sandbox. Wally needs to get dressed and find his keys, but every action costs time, energy or noise — and simple objects can combine into surprisingly bad decisions.

Six objects. Ten deterministic rules. One bedroom. No game engine hiding the interesting parts.

move → interact → consequence → adapt → succeed / fail → retry

Why this repo exists

Pyjamada is also a compact game-architecture playground:

• React Native owns the app shell.
• TypeScript owns the game rules and state.
• Skia owns the pixels.
• AsyncStorage owns persistence.
• Maestro owns the Android visual tour.

The goal is to see how much emergent gameplay can come from a very small deterministic model before adding more rooms, progression or monetization.

The current game

The bedroom contains six interactive objects:

bed · slippers · alarm clock · wardrobe · keys · window

Wally can be sleepy, normal, rushed or startled. Ten ordered rules connect those states with object interactions and movement, producing different routes through the same room: efficient escapes, near misses and full domestic chaos.

The objective is simple:

Get dressed + find the keys before the house wakes up, Wally runs out of energy, or time runs out.

There is one active gameplay path and one save model.

Architecture

                         PYJAMADA
                            │
                    React Native shell
                            │
             ┌──────────────┼──────────────┐
             ▼              ▼              ▼
         MainMenu      SettingsScreen   GameScreen
                                           │
                              input ────────┼──────── render
                                           │
                         ┌─────────────────┴─────────────────┐
                         ▼                                   ▼
                 SystemicRuntime                         GameCanvas
                  pure TypeScript                           Skia
                         │
              ┌──────────┼──────────┐
              ▼          ▼          ▼
           Objects      Rules    Objectives
              └──────────┼──────────┘
                         ▼
                       State
                         │
                 ┌───────┴────────┐
                 ▼                ▼
             Persistence       Telemetry

The systemic folder describes the architecture of the gameplay engine; it is not a separate game mode.

Stack

• React Native 0.86 + Expo 57
• TypeScript
• React Native Skia
• AsyncStorage
• Maestro
• Android-first development

Gameplay logic is UI-independent and deterministic, which keeps the interesting behavior testable without rendering a frame.

Run it

Requirements: Node.js >=22.13, Android tooling and Java 17.

npm install
npm run android

For the Expo development server separately:

npm start

Validate it

npm run test:all
npm run typecheck

The test suite covers the game rules, object interactions, success / near-miss / chaos paths, restart behavior, persistence validation, telemetry and settings.

Generate the Android visual tour

With an Android emulator running and Maestro installed:

npm run screenshots:android

If the release APK is already built:

SKIP_BUILD=1 npm run screenshots:android

The tour exercises the real game flow and writes its screenshots to:

artifacts/android-screenshots/

Useful entry points

App.tsx                         application composition + navigation
src/app/GameScreen.tsx          HUD, feedback and touch controls
src/game/systemic/              state, objects, rules, runtime and telemetry
src/game/render/GameCanvas.tsx  Skia bedroom renderer
src/platform/storage/           game persistence
src/settings/                   settings domain
maestro/screenshots.yaml        Android screenshot journey
tests/game.test.ts              deterministic gameplay coverage

Current scope

Pyjamada is intentionally small. The current question is not “how much content can we add?” but “does this tiny room make players curious enough to experiment and retry?”

The next expansion gate is human playtesting: objective comprehension, understandable cause/effect, at least one unexpected-but-logical consequence, and voluntary retry.

See docs/GAMEPLAY.md for the active gameplay contract and docs/ANDROID_SMOKE_TEST.md for device validation.