Install any skill in seconds. Free to start, no credit card required.
Get Started Free →zzafergok/threejs-skill
Build high-performance 3D web applications using Three.js (WebGL/WebGPU).
Passed 0 of 22 test cases with the skill, and 22 of 22 without it.
0 fixed19 passed either way3 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
This benchmark measured an earlier version of the skill.
The instructions your model reads when the skill loads.
Create a 3D experience that meets the product goal and target-device performance budget rather than maximizing scene complexity.
Identify target browsers and devices, input methods, frame-rate goal, loading budget, accessibility alternative, asset pipeline, visual requirements, and whether WebGL, WebGPU, or XR behavior is actually required. Inspect the installed Three.js version and existing renderer architecture.
Choose asset formats, compression, texture sizes, color spaces, tone mapping, lights, shadows, and post-processing based on measured quality and performance. Avoid shipping editor-scale assets directly. Provide progress, error, cancellation, and fallback behavior for asset loading.
For custom shaders, document uniforms, spaces, precision, fallback behavior, and the visual reason the shader is necessary. Treat shader source and generated values as untrusted when user-controlled.
Define pointer, touch, keyboard, controller, and XR input as required. Make selection and focus states visible. Provide a meaningful non-3D representation or equivalent controls when the experience communicates essential information.
Measure before optimizing. Inspect frame time, draw calls, triangles, texture memory, shader compilation, asset transfer, CPU work, and garbage collection. Apply level of detail, instancing, culling, batching, simplified materials, or lower render resolution only where evidence supports them.
Test representative devices and browsers, slow loading, failed assets, resize, background and foreground transitions, navigation cleanup, reduced motion, input modes, and context recovery. Report observed metrics rather than generic performance claims.
test cases passed, vs 22 / 22 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.