Custom Kernels for All from Codex and Claude

Custom Kernels for All from Codex and Claude Custom Kernels for All from Codex and Claude Published February 13, 2026 Update on GitHub Upvote 81 ben burtenshaw burtenshaw Sayak Paul sayakpaul Aritra Roy Gosthipaty ariG23498 shaun smith evalstate tl;dr: We built an agent skill that teaches coding agents how to write production CUDA kernels. Then we pointed Claude and Codex at two real targets: a diffusers pipeline and a transformers model.

This ProductUpdate is relevant to the technology intelligence record because it involves GitHub, NVIDIA, DeepSeek, Meta. The source article should remain the factual reference for follow-up coverage.

• Custom Kernels for All from Codex and Claude Published February 13, 2026 Update on GitHub Upvote 81 ben burtenshaw burtenshaw Sayak Paul sayakpaul Aritra Roy Gosthipaty ariG23498 shaun smith evalstate tl;dr: We built an agent skill that teaches coding agents how to write production CUDA kernels.
• Then we pointed Claude and Codex at two real targets: a diffusers pipeline and a transformers model.
• The agents produced working kernels for both, with correct PyTorch bindings and benchmarks, end to end.
• Writing CUDA kernels that correctly integrate with transformers and diffusers is harder.
• There are architecture-specific memory access patterns, vectorization strategies, warp shuffle reductions, and a dozen integration pitfalls that trip up even experienced developers.
• It is exactly the kind of specialized, high-stakes problem where agent skills shine.