Does it support C++ bound functions?

Yes, the skill includes specific implementation patterns for documenting C-bound functions using the _add_docstr pattern common in PyTorch.

Does it generate usage examples automatically?

It provides the structure and guidelines to ensure every docstring includes at least one valid Python example following the 'Examples::' directive.

Does this skill handle LaTeX math?

Yes, it specifically instructs the use of raw strings and Sphinx math directives for tensor shapes and mathematical formulas to ensure correct rendering.

Is this skill only for the PyTorch core library?

While it follows PyTorch's internal conventions, it is ideal for any deep learning library that wants to maintain high-quality, Sphinx-compatible documentation.

PyTorch Docstring Writer

Name: PyTorch Docstring Writer
Author: arthur0824hao

byarthur0824hao

学習とドキュメンテーション

Generates high-quality docstrings for PyTorch functions and methods following official Sphinx and reStructuredText conventions.

概要

This skill automates the creation of standardized PyTorch documentation, ensuring that functions and methods adhere to the specific formatting rules used in torch/_tensor_docs.py and torch/nn/functional.py. It guides Claude in using raw strings for LaTeX math, structured argument blocks, cross-references via Sphinx roles, and comprehensive usage examples, making it indispensable for contributors to the PyTorch ecosystem or developers building PyTorch-compatible libraries that require professional-grade documentation.

主な機能

Sphinx-compliant cross-referencing for classes, functions, and methods
0 GitHub stars
Automated LaTeX math integration for tensor shapes and formulas
Structured 'Args', 'Returns', and 'Examples' sections following reST
Standardized PyTorch signature formatting with type annotations
Support for C-bound functions (add_docstr) and in-place method variants

ユースケース

Documenting new modules or functions within the PyTorch core repository
Standardizing documentation for custom PyTorch-based deep learning libraries
Updating legacy PyTorch code to modern Sphinx documentation standards

概要

主な機能

Sphinx-compliant cross-referencing for classes, functions, and methods
0 GitHub stars
Automated LaTeX math integration for tensor shapes and formulas
Structured 'Args', 'Returns', and 'Examples' sections following reST
Standardized PyTorch signature formatting with type annotations
Support for C-bound functions (add_docstr) and in-place method variants

ユースケース

Documenting new modules or functions within the PyTorch core repository
Standardizing documentation for custom PyTorch-based deep learning libraries
Updating legacy PyTorch code to modern Sphinx documentation standards