Claude Skill MartBrowse skillsQuick linesLearn by videoTerminal guideWhat is a Skill?
 
← Back to list

Add uint Support to PyTorch Operators

Updates PyTorch AT_DISPATCH_V2 macros to enable uint16/uint32/uint64 type coverage for operators and kernels.

Dev & CodingAdvanced★ 403⑂ 55AI score 8/10Last updated: Sep 6, 2026

What it does

Locates AT_DISPATCH_V2 macros in PyTorch C++ sources and extends them with barebones unsigned integer types (kUInt16, kUInt32, kUInt64).

  • Two documented methods: explicitly add AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES), or swap AT_INTEGRAL_TYPES for AT_INTEGRAL_TYPES_V2
  • Detects legacy AT_DISPATCH_* macros and routes you through the at-dispatch-v2 skill first
  • Sweeps every dispatch site in a file (CPU/CUDA, multiple functions) for consistent coverage
  • Handles edge cases: float-only operators, files that already have uint, complex types present

Who it's for

  • PyTorch core contributors touching ATen/native kernels
  • ML infrastructure engineers maintaining an internal PyTorch fork
  • C++ developers doing bulk refactors of macro/template dispatch code

Examples

  1. Open a CUDA reduce min/max kernel and ask "add uint support" — every dispatch site gains AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES)
  2. A file listing AT_INTEGRAL_TYPES + AT_FLOATING_TYPES gets the tidier AT_INTEGRAL_TYPES_V2 substitution
  3. Legacy AT_DISPATCH_ALL_TYPES_AND2(...) code is first converted to V2, then extended with unsigned types

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/Microck/ordinary-claude-skills/HEAD/skills_all/add-uint-support/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the skills_all/add-uint-support folder from the GitHub repo Microck/ordinary-claude-skills into my ~/.claude/skills/add-uint-support/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/Microck/ordinary-claude-skills.git && mkdir -p ~/.claude/skills && cp -r ordinary-claude-skills/skills_all/add-uint-support ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal in your working directory.
  2. Clone the repo: git clone https://github.com/Microck/ordinary-claude-skills.git
  3. Create the skills folder: mkdir -p ~/.claude/skills
  4. Copy the skill: cp -r ordinary-claude-skills/skills_all/add-uint-support ~/.claude/skills/
  5. Also copy the companion at-dispatch-v2 skill so legacy dispatch macros can be converted automatically.
  6. Run Claude Code inside your PyTorch source tree and ask it to "add uint16/32/64 support to this kernel".
  7. Build and run your own tests afterwards — the skill does not verify functionality.
View source on GitHub ↗License: NOASSERTION