
A practical, evidence-first skill for clear machine-learning and computer-vision research writing.
A practical, evidence-first skill for clear machine-learning and computer-vision research writing.
·
A skill for revising English research papers: clearer sentences, easier comparisons, and a better path through the argument.
Quick start · How it works · Sources · 中文
Give it a paragraph and say what you want to improve. It returns a revision with brief reasons for the edits, preserving the technical meaning, numbers, and conditions. For a larger revision, ask it to work on an abstract, section, or argument.
The skill offers sentence cards for everyday editing and a guide to choosing the right scope for larger revisions. Detailed reference notes are mainly in Chinese; the examples and revised prose are in English.
Follow one paragraph from method to protocol to result. Five edits group related facts, improve the handoffs, and vary the rhythm while preserving the claims and conditions. Read the full example
For a routine polish, it works on sentences and wording. You can ask for broader changes when the structure also needs attention.
Download or clone this repository. Copy skills/revise-research-writing into ~/.agents/skills/ for personal use, or into your project's .agents/skills/ folder for that project.
Run the following from the repository root. An existing installation is left in place so you can review local changes before updating.
target="$HOME/.agents/skills/revise-research-writing"
mkdir -p "$HOME/.agents/skills"
if [ -e "$target" ]; then
printf '%s\n' "Already installed: $target"
else
cp -R skills/revise-research-writing "$target"
fi$skillRoot = Join-Path $HOME '.agents/skills'
$target = Join-Path $skillRoot 'revise-research-writing'
if (Test-Path -LiteralPath $target) {
Write-Output "Already installed: $target"
} else {
New-Item -ItemType Directory -Force -Path $skillRoot | Out-Null
Copy-Item -LiteralPath 'skills/revise-research-writing' -Destination $target -Recurse
}If the skill does not appear, restart the client. See the official skill documentation for discovery and installation details.
Use $revise-research-writing to polish the paragraph below.
Keep the sentence order, technical claims, and numbers.
Return the revision and briefly explain the important edits.
[Paste your paragraph here.]For an abstract that needs more structural work:
Use $revise-research-writing to improve this abstract.
Help the reader follow the problem, approach, and main result.
You may reorder sentences; keep all claims tied to the results I provide.
[Paste your abstract here.]Provide SKILL.md and its references/ folder, then ask the assistant to follow the skill and consult the relevant references. Make both available as files so the links can be followed.
This package consists of instructions and examples. It needs no additional API key, model download, or runtime dependency; the optional validator uses Python's standard library.
The nine sentence cards each address a common editing problem. The assistant starts with the one to three cards relevant to your passage.
| Card | What to look for |
|---|---|
| S1 · Subjects and verbs | An action buried in a noun phrase |
| S2 · Connections between sentences | A sentence that does not build on the previous one |
| S3 · Name the relationship | A division, comparison, or arrangement repeated in full or referred to vaguely |
| S4 · Parallel comparisons | Corresponding roles or settings that are hard to compare |
| S5 · Stages and qualifiers | An unclear training stage, inference condition, or “only” |
| S6 · Opening clauses | Too much setup before the main point |
| S7 · Complete conditions | A claim separated from its condition |
| S8 · Benefits and limits | A contrast that obscures a benefit or its limit |
| S9 · Sentence rhythm | Choppy sentences or unrelated points crowded together |
For larger revisions, the eight-layer toolkit covers the scientific story, argument, paper structure, sections, paragraphs, sentences, wording, and grammar. Choose the level that fits your request.
The source notes study papers with Kaiming He or Saining Xie as last-listed authors, using a snapshot dated 2026-10-12. These are collaborative papers. The observations describe techniques in the selected versions; author position identifies the groups, not who wrote a particular sentence. This is an independent project, unaffiliated with the authors or publishers.
These are qualitative reading notes from a selected corpus. The external reference guide records separate influences from Nature's summary-paragraph guide and community writing projects.
A useful contribution starts with a passage that is difficult to read, a faithful revision, and an explanation of the editing choice. See CONTRIBUTING.md.
python3 scripts/validate.pyThe validator checks file structure, local links, source counts, card references, and example data. Review meaning and writing quality with the worked examples and behavioral cases; the latter specify expected constraints rather than measured model performance.
Original instructions, examples, scripts, and artwork use the MIT License. Linked papers and third-party materials retain their own terms. Source credits and permissions are collected in ATTRIBUTION.md; release history is in CHANGELOG.md.
This repo does not document an install step we could read. Check the README on GitHub.
Paste this prompt into your agent. It reads this page and does the setup for you.
Read https://aiagentslisting.com/mcp/research-writing-skill to learn what the "Research Writing Skill" MCP server does and how to install it. Add it to my coding agent's MCP configuration as documented on that page, then confirm the server connects and list the tools it exposes.Agents can also browse this directory over MCP at https://aiagentslisting.com/api/mcp
This server doesn't publish a machine-readable tool manifest, and it doesn't expose a public endpoint we could ask. Connect the server locally to see what it exposes.
We could not run this server: it ships no npx package for our sandbox and no public endpoint. The four checks that need a live server are left out, and the grade tops out at B until one exists.
Maintain this server? Claiming proves you own the listing and earns the last 4 points.
Claim this listingPaste this line near the top of your repository's README. The badge always renders the current score.
[](https://aiagentslisting.com/mcp/research-writing-skill)Nothing comparable is listed yet.
No tagged releases on record.
Connect to the AI Agents Listing MCP server at https://aiagentslisting.com/api/mcp using the HTTP transport and complete the sign-in when it asks. Then call the submit_review tool for the listing "research-writing-skill" with a rating from 1 to 5 and a short note on what worked and what did not.No reviews yet
Be the first to share your experience with this tool.
One email a week. New agents, MCP servers and skills, and what is actually getting traction.