Kernel Triton Specialist — Development agent for Claude Code
name: kernel-triton-specialist description: Expert in writing optimized Triton kernels for PyTorch operators.
How to install Kernel Triton Specialist
Installs to ~/.claude/agents/nvidia-tensorrt-llm-kernel-triton-specialist.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/NVIDIA/TensorRT-LLM/HEAD/.claude/agents/kernel-triton-specialist.md -o ~/.claude/agents/nvidia-tensorrt-llm-kernel-triton-specialist.md Restart Claude Code, or start a new session, for it to be picked up.
What Kernel Triton Specialist does
name: kernel-triton-specialist description: > Expert in writing optimized Triton kernels for PyTorch operators. Delegate to this agent for: (1) Analyzing operators for Triton suitability, (2) Writing fused Triton kernels (element-wise, reductions, attention), (3) Verifying kernel correctness against reference, (4) Benchmarking performance. skills:
- kernel-triton-writing
- perf-workload-profiling license: Apache-2.0 metadata: author: NVIDIA Corporation
You are a Triton ker
Alternatives in Development
- GPU Kernel Baseline 95 ★
- Ascend Kernel Developer — Ascend kernel 开发专家 Agent，通过 TileLang 设计表达和 AscendC 落地完成算子优化任务 39 ★
- Strategy Critic — PROACTIVELY use this subagent immediately after the user writes or revises any strategy artifact: Strategy Blo 24 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Fk Kernel Engineer
Writes and integrates GPU kernels (Triton, TileLang, CuTe DSL, CUDA C++, CUDA graphs, torch.compile) under can
Fk Librarian
Finds the right API pattern or prior evidence - Triton/TileLang/CuTe/CUDA docs, the starter templates in fastk
Kernel Cuda Specialist
Kernel Cute Specialist
Sap Basis Consultant
SAP Basis 장애 상황을 체계적으로 라우팅하고 진단하는 한국어 전문 에이전트. ABAP 덤프(ST22), Work Process 행(SM50/SM66), Transport 실패(STMS), R
Linus Code Reviewer
Use this agent when you need rigorous code review focused on quality, performance, and best practices from a L
Related Skills
Triton Dynamic Launcher Tiling
Build Triton kernel launchers that pick tile sizes, warps, and stages at runtime based on problem shape, dtype
ML Production
Use when shipping ML models (classical or LLM) to production - quantization (torch.ao.quantization, AWQ/GPTQ/A
Claudes C Compiler
Claude Opus 4.6 wrote a dependency-free C compiler in Rust, with backends targeting x86 (64- and 32-bit), ARM,
