Check game screen readability with AI
Check the frame where two events overlap and find messages that technically appeared but a player could not read.
Check game screen readability with AI when your tests say a notification fired but nobody remembers seeing it. A puzzle game might show a time bonus directly under its “level complete” banner. The auditing-game-screen-readability workflow asks the agent to capture that exact coincidence and compare what is technically drawn with what a player can make out.
A useful report names the frame, the competing elements and the missing information. Show an ordinary play frame to someone unfamiliar with the rules and ask what a particular counter measures. If they mistake a remaining-moves counter for points, more contrast alone will not solve the problem. Keep their interpretation with the screenshot so the designer can revisit the choice.
What it does
- Capture the exact frame in which two notices compete for attention.
- Check text and icons at native resolution.
- Ask an unfamiliar reader what a counter measures and whether an increase is good.
- Log clipping, occlusion and ambiguous meaning separately.
- Repeat the same event after a change and compare the evidence.
Install and use
- Use an authorised build that can reach chosen game states and capture frames at native size.
- Name the events you want to coincide and ask for screenshots of those moments.
- Record the visible problem, make the change and capture the matching frame again.
FAQ
Source
Made by abagames. Original skill: auditing-game-screen-readability, published on GitHub under the MIT licence.
View on GitHub