ShaderSitter
A self-hosted CLI that graphics programmers drop into their Unity/Unreal build pipelines to catch shader breakage at PR time, with a 'translate-and-visually-verify' workflow before merge.
Graphics programmers at small-to-mid game studios
- Pre-merge shader-translation gate that fails the PR if SPIR-V won't validate
- Side-by-side Metal vs Vulkan render diffs uploaded as PR artifacts
- Project-aware translation cache to skip already-verified shaders in seconds
- Configurable 'warn-only' mode for experimental shader stages
The trending native-Rust emitter finally makes pipeline-grade translation realistic without an external LLVM dependency, but no turnkey CI tool wraps it for studios yet.
Web search tooling failed (403s), so I could not confirm live demand; based on general knowledge, shader cross-compilation is a known pain but most studios lean on engine-native (Unity's shader compiler, Unreal's shader compile worker) pipelines rather than a third-party CLI.Trending Rust repositories on GitHub today ↗
No live competitor scan possible, but established free/OSS tools (naga, SPIRV-Cross, Google's Tint, RenderDoc) cover the translation/validation pieces; the 'turnkey Unity/Unreal PR-time CLI' wrapper niche does appear relatively open, though Unity Cloud Build and Unreal's own automation also touch shader validation tangentially.Trending Rust repositories on GitHub today ↗
Self-hosted CLIs targeting small-to-mid game studios are notoriously hard to monetize — studios expect infra tooling to be cheap or OSS, and naga/SPIRV-Cross are already free; per-seat or per-pipeline pricing rarely clears mid-five-figures in this segment.Trending Rust repositories on GitHub today ↗
Cross-platform shader portability is a durable concern (especially with Apple Silicon and Vulkan-on-Mac momentum), but the specific 'Metal AIR' framing is fragile as Apple reshapes Metal's shading language; engines are also gradually converging on HLSL/Slang as a portability layer, which could erode the wedge over 3-5 years.
Wrapping an existing naga/SPIRV-Cross translation step into a CI CLI is tractable, but the 'visually verify' half is the hard part — it requires headless render targets, deterministic golden-image diffing, and per-platform shader-variant matrices, which materially raise build and maintenance cost.