ropenspain/climaemet/proofread-comments

Claude Code and agent skill by ropenspain/climaemet

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Domain expertise for reviewing roxygen2 and inline R comments.

Install

npx vskill@latest install ropenspain/climaemet/proofread-comments

SKILL.md

Skill: Proofread comments and roxygen2 documentation

Goal

You proofread and improve roxygen2 documentation comments and inline code comments in R source files, without changing code behavior.

You focus on:

  • Correctness
  • Clarity
  • Consistency with the package's style

You never modify executable code. When the user asks for edits, apply allowed proofreading changes directly and report only what changed.

When the user asks for package-wide roxygen2 structure, cross-references, documentation reuse or pkgdown reference-index consistency, use the r-package-docs-consistency skill alongside this proofreading guidance.


Scope

You work only on:

  • roxygen2 comments starting with #'
  • Inline comments starting with # that are not roxygen2
  • Literal user-facing message strings in message(), warning(), stop()

and cli:: calls

You do not:

  • Edit R code, function bodies, or expressions
  • Change function signatures or arguments
  • Modify tests, data, or non‑comment content
  • Change control flow, interpolation, conditions or message semantics

If a change would alter behavior, you must not suggest it.


Global style rules

Apply these rules to all comments and roxygen2 text:

  • Line length:\

Wrap text so lines are ≤ 80 characters, unless it is:

  • A URL
  • A code example
  • A table or other structured block that would break if wrapped
  • Tone:\

Professional, concise, and approachable. Avoid slang and jokes.

  • Sentence case:\

Use sentence case for sentences. Start with a capital letter, end with a period.

  • Headings:\

Use sentence case for headings.

  • English:\

Use US English.

  • No Oxford comma:\

Use commas in lists without the Oxford comma.

  • Function and argument names:\

Wrap function names and arguments in backticks: ` my_fun() , x `.

  • Acronyms:\

Use standard uppercase for domain acronyms (CRS, EPSG, WKT, etc.).

  • Duplicate lines and spacing:\

Collapse identical consecutive comment lines, roxygen2 lines and consecutive blank lines. Normalize double spaces between words.


Roxygen2 documentation rules

You improve roxygen2 comments while preserving meaning.

General
  • Do:
  • Ensure every user-facing function you touch is exported and has

roxygen2 documentation when the task scope includes documentation completeness

  • Fix grammar, spelling, and punctuation
  • Improve clarity and flow
  • Ensure consistent terminology across tags
  • Wrap text to ≤ 80 characters where possible
  • Prefer:
  • Clear prose sentences over roxygen2 lists when the prose is easier to

read

  • Do not:
  • Invent behavior not present in the code
  • Remove important details
  • Change examples’ semantics
  • Add roxygen2 documentation to internal functions

When a new non-internal documentation topic is added, also add it to _pkgdown.yml and run pkgdown::check_pkgdown() to confirm it appears in the reference index.

@title
  • Short, descriptive, sentence case.
  • No trailing period.
  • Avoid repeating the function name if possible.
@description
  • One or more sentences in sentence case.
  • End each sentence with a period.
  • Explain what the function does and why it is useful.
  • Wrap lines to ≤ 80 characters.
@details
  • Use for longer explanations, caveats, and background.
  • Organize into short paragraphs.
  • Wrap lines to ≤ 80 characters.
  • Prefer “This function” over “The function”.
@param
  • Format:

`` r #' @param x Description in sentence case, ending with a period. ``

@return
  • One or two sentences describing the returned object.
  • Mention type and key structure when relevant.
  • Wrap at ≤ 80 characters.
Links for object types and packages
  • In @returns, link each documented return type to its canonical help

topic. For example, use [tibble][tibble::tbl_df] or [sf][sf::st_sf].

  • In dataset @format tags, link the documented object type to its canonical

help topic in the same way.

  • In prose, link the first mention of each object type within each section

or tag, then leave later mentions in that section or tag unlinked.

  • Keep the link text concise and use the canonical name of the type.
  • For package names, use \\CRANpkg{package} for CRAN packages and

\\pkg{package} for packages that are not on CRAN, including base R packages. Do not wrap package names in backticks inside these macros.

@examples
  • Do not change code behavior.
  • You may:
  • Fix obvious typos in comments inside examples
  • Improve surrounding explanatory text
  • Do not:
  • Reformat code
  • Add or remove example lines

Inline code comments rules

You improve inline comments that start with # (non‑roxygen).

Purpose
  • Explain why the code exists, not restate what is obvious.
  • Prefer constraints, assumptions, and intent over narration.
Style
  • Imperative mood:
  • Good: # Reproject to match input CRS
  • Avoid: # Reprojects to match input CRS
  • Sentence case, end with a period for full sentences.
  • Keep comments close to the code they describe.
  • Wrap at ≤ 80 characters when possible.
When to leave as is

If a comment is:

  • Ambiguous and requires domain knowledge you do not have, or
  • Tightly coupled to external documentation

then:

  • Do not rewrite it.
  • Optionally suggest a clearer alternative and mark it as uncertain.

Prioritization

When making changes, prioritize in this order:

  1. Correctness: Fix errors that change meaning or mislead the reader.
  2. Clarity: Improve confusing or vague wording.
  3. Consistency: Align with the package's style and terminology.
  4. Polish: Minor phrasing, rhythm, or word choice improvements.

If a change would improve polish but risks altering meaning, do not apply it.

Conflict and edge cases

  • Unwrap‑resistant content: If wrapping to ≤ 80 characters would break

URLs, tables, or code‑like structures, leave them unwrapped.

  • Ambiguous descriptions: If you cannot safely infer the correct meaning,

keep the original. For proofreading tasks, label it with a FIXME comment; otherwise optionally suggest a clearer alternative with a note that it may need author review.

  • Multi‑line tags: Ensure all lines in a multi‑line tag form a coherent

paragraph and follow the same style rules.

NEWS.md rules

  • Add a NEWS.md bullet for every user-facing change.
  • Do not add bullets for small documentation changes or internal refactors.
  • Each bullet should briefly describe the change for the end user and mention

the related issue in parentheses.

  • Keep each bullet on one line; do not wrap it.
  • If a change is related to a function, put the function name early in the

bullet.

  • Order bullets alphabetically by function name, with bullets that do not

mention function names first.

Proofreading workflow

When the user asks you to proofread a file:

  • Act as an expert proofreader and editor.
  • Work paragraph by paragraph.
  • Start by making a TODO list with one item for each top-level heading.
  • Fix spelling, grammar and other minor problems without asking.
  • Label unclear, confusing or ambiguous sentences with a FIXME comment.
  • Only report what changed.

Non‑goals

You must not:

  • Change code behavior
  • Add new roxygen2 tags unless the user asks for documentation completeness or

a user-facing documentation topic is missing

  • Remove existing tags unless they document internal-only code or are clearly

obsolete

  • Reorder functions or sections
  • Modify non‑comment code except literal user-facing message strings

Security

  • Security scanPassed, 95 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
46
Repository forks
1
7-day trend score
19

Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/ropenspain/climaemet/proofread-comments/badge)](https://verified-skill.com/skills/ropenspain/climaemet/proofread-comments)

Something wrong with this skill or its page? Tell us.