ascendc-operator-doc-gen

Extracts AscendC operator interface data from source files into PyTorch-style README documentation drafts.

31|1|Updated Mar 28, 2026
One-click install
npx skills add https://github.com/Ascend/agent-skills --skill ascendc-operator-doc-gen
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ascendc-operator-doc-gen
Source: https://github.com/Ascend/agent-skills/tree/main/skills/ascendc-operator-doc-gen
Command: npx skills add https://github.com/Ascend/agent-skills --skill ascendc-operator-doc-gen

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

将 AscendC 算子源代码中的接口信息提取为 PyTorch 风格的 API 文档,自动生成 README.md 的文档草案。

Core Features & Use Cases

  • 自动提取 register.cpp、ops.h、design.md、op_host、tests 中的信息并生成结构化文档草案。
  • 为不同算子生成一致的文档格式,方便团队审阅与发布。

Quick Start

触发该技能以为 AscendC 算子生成 PyTorch 风格的接口文档。

Frequently Asked Questions about ascendc-operator-doc-gen

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate PyTorch-style documentation for AscendC operators?

To generate PyTorch-style documentation for AscendC operators, trigger this Skill to extract interface data from source files like register.cpp and ops.h, assembling a structured README.md draft automatically.

What source files are needed to extract operator API information into a README?

Extracting operator API information requires source files including register.cpp, ops.h, design.md, op_host directories, and tests, which are analyzed together to assemble a consistent README-ready documentation draft.

Can I automate README generation during the compile-debug workflow for custom C++ operators?

Yes, you can automate README generation during the compile-debug workflow; the Skill is explicitly applied during compile-debug workflows to extract interface data and format it into consistent API documentation for custom operators.

How does generating PyTorch-style operator documentation ensure consistency across different C++ operators?

Generating PyTorch-style operator documentation ensures consistency by automatically extracting and formatting interface data from specified source files into a standardized structure, providing a uniform README format for team review and publication.

Do I need design.md to generate a complete API documentation draft for my operator?

Yes, design.md is required; the Skill analyzes register.cpp, ops.h, design.md, op_host, and tests to assemble a complete README-ready document, ensuring all design specifications are included in the API documentation draft.