review-doc-style

Featured

Review NeMo Relay documentation, examples, or public text for NVIDIA technical writing style, terminology, and repository accuracy. Do not use for ordinary implementation review.

AI & Automation 177 stars 75 forks Updated today Apache-2.0

Install

View on GitHub

Quality Score: 87/100

Stars 20%
75
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
73
Issue Health 10%
50
License 10%
100
Description 5%
100

Skill Content

# Review Documentation Style Prioritize factual accuracy over copy polish. Confirm commands, package names, APIs, paths, supported bindings, and public behavior against the current repository before reporting style findings. ## Review Flow 1. Identify the changed documentation, examples, or public strings and any corresponding entry points such as `README.md`, `docs/index.yml`, or package READMEs. 2. Read `references/nvidia-style-guide.md` for the fast-path checklist and severity model. 3. Open only the focused reference that resolves an actual ambiguity: - `references/nvidia-style-technical-docs.md` for structure, procedures, examples, links, tables, UI text, or accessibility. - `references/nvidia-style-language-mechanics.md` for voice, grammar, punctuation, dates, numbers, units, or plain English. - `references/nvidia-style-brand-terminology.md` for NVIDIA and product names, trademarks, acronyms, legal copy, or SEO. 4. Verify MDX files use JSX delimiters for top-of-file SPDX comments. 5. Report findings in severity order with a file and line, reader impact, and a concrete rewrite or direction. Omit preference-only findings unless the user requested a deep copyedit. If no issues are found, say so and identify any commands or examples that were not executed. ## References - `CONTRIBUTING.md` - `references/nvidia-style-guide.md` - Open only the focused NVIDIA style reference selected in step 3.

Details

Author
NVIDIA
Repository
NVIDIA/NeMo-Relay
Created
5 months ago
Last Updated
today
Language
Rust
License
Apache-2.0

Integrates with

Similar Skills

Semantically similar based on skill content — not just same category