mcpField
R
renderer

renderer

github.com/dzhngDevelopment1.4k260 installsUpdated today
Repository

About renderer

renderer is a GPU-rendering skill for projects using TypeGPU and WebGPU. It guides developers in keeping rendering costs bounded by the view, maintaining shared engine contracts, managing GPU resource lifetimes, and validating visual changes on real hardware.

Key features

  1. 01Bound geometry and per-frame work by view
  2. 02Keep engine contracts under one owner
  3. 03Share depth and lighting conventions
  4. 04Manage GPU resources through a registry
  5. 05Validate results with real-GPU images

Use cases

  1. 01Reviewing renderer changes for world-size scaling
  2. 02Adding a render pass under shared conventions
  3. 03Checking GPU resource cleanup after rebuilds
  4. 04Verifying an optimization with image evidence

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/dzhng/skills --skill renderer

After installing, type/rendererin Claude Code to confirm it loaded.

Details

From the project README

GPU renderer

The renderer is TypeGPU over WebGPU. It draws plain frame inputs the application hands it and never reaches into simulation or app state. These are the flaws that sink a 3D experience; check a change against each before writing it.

  • Fixed-view cost never grows with world size. For a fixed camera and render size, submitted geometry, shadow casters, and per-frame lists are bounded by what the view can see or cast into, and screen-space work by pixels. Store the world map-wide; select from it spatially. Frustum clipping does not make vertices free.
  • One owner per contract. Frame orchestration, camera, depth convention, targets, culling, lighting, timing, and resource lifetime each have one owner every pass shares. Extend it; a second copy drifts.
  • Shared contracts change together. Depth and lighting are engine-wide: every pass reads the same convention and every material the same shading path. A per-pass exception is a future visual bug.
  • No GPU resource without a lifetime owner. Everything is allocated through the registry, rebuilds swap in whole, and repeated resize/rebuild/reset returns to baseline.
  • Pixels prove correctness, not counters. An optimization is accepted only with an interleaved A/B on the real GPU and an image that shows the subject still drawn correctly.

Comments

Markdown supported