Can this skill document C-bound functions?

Yes, it includes specific implementation patterns for using `_add_docstr` and `add_docstr_all`, which are standard for documenting PyTorch functions defined in C++.

Why does this skill use raw strings for all docstrings?

PyTorch documentation frequently uses LaTeX and math backslashes. Raw strings (r""") prevent Python from interpreting these as escape characters, ensuring they render correctly in Sphinx.

Does it support cross-referencing other PyTorch modules?

Absolutely. It uses Sphinx roles like :class:, :func:, and :meth: with the tilde prefix to create proper links within the PyTorch documentation ecosystem.

How does it handle tensor shapes in documentation?

It follows the PyTorch convention of using LaTeX math notation within the Args section to describe dimensions, such as :math:`(\text{minibatch} , \text{in\_channels})`.

PyTorch Docstring Formatter

Name: PyTorch Docstring Formatter
Author: danielzayas

bydanielzayas

0•

Learning & Documentation

Writes and updates PyTorch-compliant docstrings using specific Sphinx, reStructuredText, and LaTeX conventions.

The PyTorch Docstring skill is a specialized tool designed to ensure that code documentation meets the rigorous standards of the PyTorch project. It automates the creation of raw-string docstrings to protect LaTeX backslashes, standardizes function signatures, and formats parameter lists (Args/Keyword Args) according to PyTorch's specific indentation and typing rules. Whether you are documenting new neural network layers, C-bound functions, or mathematical operations, this skill provides the correct patterns for cross-referencing, mathematical notation, and example blocks to ensure high-quality, render-ready documentation.

Key Features

01Automates Example block formatting with Python prompt conventions

02Supports specialized PyTorch patterns like _add_docstr and add_docstr_all

03Enforces raw-string (r""") usage for LaTeX and math compatibility

04Generates precise parameter descriptions with LaTeX tensor shape notation

05Standardizes Sphinx and reStructuredText (reST) cross-references

060 GitHub stars

Use Cases

01Converting existing Python docstrings to PyTorch-compliant reST format

02Standardizing documentation for a new PyTorch pull request

03Documenting complex mathematical functions with proper LaTeX and cross-references

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add danielzayas/pytorch__pytorch.c8a37a2d docstring

For use in Claude.ai and ChatGPT

Download Skill