tilegym-cutile-python — Claude Code skill: cuTile GPU kernels
Turns the assistant into a cuTile specialist: writing, debugging and tuning GPU kernels in NVIDIA's Python DSL. Useful when porting tensor ops off PyTorch by hand.
npx skills add NVIDIA/skills --skill tilegym-cutile-python -a claude-codeWhat tilegym-cutile-python does
The skill gives the agent working knowledge of cuTile, a domain-specific language for parallel computing on NVIDIA hardware. It covers the execution and memory models, the available operations (load and store, reductions, scans, matmul, atomics, metaprogramming, autotuning), and the rules for emitting code the compiler accepts.
Reference notes ship inside the skill, along with a folder of worked examples. The recommended first move on any new task is a search for something close: TileGym is the main source, and the bundled examples/ cover ops it lacks, such as convolution, pooling, GEMV, 4D matmul and split-k GEMM. When the skill lives outside a TileGym checkout, the repo gets cloned to ~/.cache/tilegym/TileGym, and TILEGYM_SKILL_CACHE_DIR points it elsewhere.
A single kernel takes one pass: state the goal, generate, validate. Bigger jobs — say a transformer block with attention and a feed-forward layer — get split across an orchestrator, an analyzer, parallel kernel agents and a composer. Pass the tensor shapes and dtype when you invoke it. Skip that and the agent guesses, and a wrong shape wrecks the result.
What it helps with
- write a cuTile GPU kernel from scratch
- port a tensor operation off PyTorch to cuTile
- fix a cuTile kernel that crashes
- speed up an existing kernel without a full rewrite
- split a multi-layer
nn.Moduleinto separate kernels
Example prompts for the agent
- Write a cuTile kernel for softmax, tensor shape (1024, 1024), dtype float16
- Port this operation to cuTile and verify the output
- Break the forward pass of this
nn.Moduleinto cuTile kernels and wire them together
How to install tilegym-cutile-python
1Skills CLI (Vercel)
npx skills add NVIDIA/skills --skill tilegym-cutile-python -a claude-code2Manually: put the files in place
~/.claude/skills/tilegym-cutile-python/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/tilegym-cutile-python/
1Skills CLI (Vercel)
npx skills add NVIDIA/skills --skill tilegym-cutile-python -a codex2Manually: put the files in place
.agents/skills/tilegym-cutile-python/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add NVIDIA/skills --skill tilegym-cutile-python -a cursor2Manually: put the files in place
.cursor/skills/tilegym-cutile-python/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install NVIDIA/skills tilegym-cutile-python2Manually: put the files in place
.github/skills/tilegym-cutile-python/SKILL.md1Skills CLI (Vercel)
npx skills add NVIDIA/skills --skill tilegym-cutile-python -a gemini-cli2Platform CLI
gemini skills install https://github.com/NVIDIA/skills --path skills/tilegym-cutile-pythonManually: put the files in place
.devin/skills/tilegym-cutile-python/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add NVIDIA/skills --skill tilegym-cutile-python -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about tilegym-cutile-python
How do I install tilegym-cutile-python in Claude Code?
Run npx skills add NVIDIA/skills --skill tilegym-cutile-python -a claude-code; the skill lands at ~/.claude/skills/tilegym-cutile-python/SKILL.md.
What does tilegym-cutile-python do?
It guides the assistant through writing and tuning cuTile kernels, and through adapting TileGym samples to new shapes.
Where do examples come from when the skill sits outside the TileGym repo?
TileGym is cloned into ~/.cache/tilegym/TileGym; set TILEGYM_SKILL_CACHE_DIR to change that location.
