Work-Life Balance

The premise of Work-Life Balance is right there in the title: you're juggling a job and a life, and the game makes that literal by splitting the screen in half and running two unrelated minigames at once, both against the same shared clock. Spot the misspelled job title on one side while matching a dance pose on the other. Pick the correct balance-sheet total while rotating a lawn mower to cut every patch of grass. It's a genuinely clever structural idea for something built during a two-week game jam, and it's executed with more visual polish than most entries in that category ever get — which is presumably why it picked up an honorable mention in Gamedev.js Jam 2025's Open Source challenge.

Buttons all the way down, confirmed across four different minigames

The thing worth actually testing, rather than assuming, is whether a design built around split attention translates cleanly to a thumb on a phone. It does, because every single interaction in every minigame tested is a tap on a clearly-defined on-screen button. The typo-finding task lays out four word buttons and you tap the misspelled one. The balance-sheet task does the same with dollar amounts. The dance and lawn-mowing tasks use paired directional buttons rather than a drag gesture. Testing this directly — not reading the code and assuming it would work — a tap on the correct word in the typo-finding task registered immediately, incremented the visible score, and the game cleanly rotated to a new pair of tasks (a balance-sheet puzzle and a lawn to mow) without any lag, double-registration, or missed input. Reading BaseButton.ts confirms why: every interactive element is a Phaser container with setInteractive() and a pointerdown handler, and Phaser's own pointer system treats touch and mouse identically with no branching anywhere in the code for input type. This is about as low-risk a touch implementation as a game can have, because there's genuinely nothing touch-specific that could have been gotten wrong.

Two real, small issues

The repository as committed to GitHub doesn't build cleanly everywhere. Three image files — the frame border, mask, and outer decoration around the game's UI — are named with a capital Frame- prefix in the repository, while the TypeScript source imports them in lowercase. That distinction is invisible on the case-insensitive filesystems most individual developers use (Windows, default macOS), so it never surfaced for the author, but it means the project fails to build on a standard Linux CI or hosting pipeline until the filenames and imports agree. Adding correctly-cased copies of the three files resolved it and the production build completed cleanly afterward.

Separately, and unrelated to the build: the opening description text — the paragraph explaining the premise on the title screen — runs off the right edge of the canvas specifically at phone width. It reads in full on a 1280px desktop screen, but on a 390px phone the text box doesn't appear to reflow to the narrower available space, cutting off the second half of several sentences. It's a presentation issue rather than a functional one — nothing depends on reading that text to play — but it's a real, confirmed gap on exactly the device category this review is built around.

Where that leaves it

A 7.0 reflects an honestly small, honestly short jam entry that does the one clever thing it set out to do — force genuine split attention across two simultaneous minigames — and does it with controls that were built to work with a finger from the ground up rather than adapted after the fact. Four days and roughly five to ten minutes total is not a lot of game, and the clipped title text is a real if minor rough edge, but what's here works cleanly and is worth the short time it asks for.