Add a CUDA/C++ AOT kernel to sgl-kernel with tests and benchmarks.
npx skills add https://github.com/OpenMOSS/MOSS-VL --skill add-sgl-kernel-openmoss
Please help me install this Agent Skill. Skill: add-sgl-kernel Source: https://github.com/OpenMOSS/MOSS-VL/tree/main/sglang/.claude/skills/add-sgl-kernel Command: npx skills add https://github.com/OpenMOSS/MOSS-VL --skill add-sgl-kernel-openmoss
The Skill helps developers add a heavyweight CUDA/C++ AOT kernel to the sgl-kernel project, including tests and benchmarks, enabling end-to-end kernel integration.
Build the project with make build in the sgl-kernel directory and run the tests to verify the new kernel.
High-intent search queries and answers about installing and using this skill.
To add a CUDA kernel to sgl-kernel, implement the C++/CUDA source, declare headers, register the torch extension, integrate CMake, expose the Python API, and write unit tests and benchmarks for validation.
Integrating an AOT CUDA kernel into a PyTorch extension involves writing the C++/CUDA implementation, adding header declarations, configuring the CMake build system, and registering the operation to ensure reproducible builds and deterministic behavior.
Yes, building C++ kernels in sgl-kernel requires CMake for build integration and PyTorch for the torch extension registration and Python API exposure to ensure proper end-to-end kernel deployment.
The best way to validate CUDA kernel performance is by running the integrated benchmarks after building the project with make build, ensuring the new kernel meets correctness and performance standards reproducibly.
Build failures when adding a new C++ operation often stem from incomplete CMake integration or missing header declarations, which are required to properly register the torch extension and expose the Python API.
Yes, you can add deterministic CUDA kernels to sgl-kernel for new PyTorch operations by following the complete workflow of C++/CUDA implementation, library registration, CMake integration, and testing.
randDiscover and load build-system skills for Make, CMake, Gradle, Maven, and Bazel.
spacecat2002Add and validate lightweight CUDA JIT kernels in SGLang's jit_kernel module.
mindspore-aiCompile and call CUDA C kernels from PyTorch using load_inline.
moirai-internalAdds custom CUDA/C++ kernels to the sgl-kernel library with PyTorch extension integration and CMake build system support.
