Best-Paper Writing
Drafts or rewrites a paper's title, Abstract, Introduction, Method, Experiments and figures in the style of CVPR/ICCV/NeurIPS award papers, grounded in 55 annotated award papers.
Content & WritingIntermediate★ 281⑂ 17AI score 9/10Last updated: Oct 29, 2026
What it does
- Turns a project description or a draft abstract into a title, Abstract and Introduction that read like a recent CVPR/ICCV/NeurIPS award paper.
- Forces an explicit prep stage first: claim, tension type, insight, handle, closest work and headline numbers; then a paragraph skeleton, a claim index, and drafting against retrieved award-paper exemplars.
- On request, extends into Method and Experiments sections and a full figure plan (Figure 1, method overview, result figures), drawing mechanism schematics in TikZ and leaving results figures as planned placeholders.
- Ships checkers (
check_draft.py,style_metrics.py,check_figures.py) that flag dashes, hedges, boosters, sentence-length outliers and AI-draft habits against corpus percentiles. - Hard rule: never fabricate numbers, datasets, baselines, citations or figure content; gaps become
[TODO].
Who it's for
- ML/CV researchers and PhD students submitting to CVPR, ICCV, NeurIPS or ICLR
- Non-native English authors who want the rhetorical structure of strong papers, not just grammar fixes
- Anyone with an existing abstract who wants a sentence-by-sentence change table with reasons
Examples
- From scratch: fill in
assets/intake_form.html, export JSON, and get a brief plus Abstract + Introduction and areport.mdwith the claim map and TODO list. - Abstract rewrite: paste your abstract; it is reverse-engineered into intake fields with a present/vague/missing table, then revised with an original → revised → reason change table.
- LaTeX project: hand it the CVPR author kit and it writes straight into
main.tex, draws Figure 1 and the method overview in TikZ, and returns thecheck_figuressummary.
· · · Install guide · · ·
Try it now, no install
Paste this into Claude to use the skill without installing anything.
Read the instructions in this file and follow them to help me: https://raw.githubusercontent.com/gulucaptain/Best-Paper-Writing-Skill/HEAD/skills/bestpaper-writing/SKILL.md What I want: (describe your task here)
If Claude can't open the link, open it yourself and paste the contents instead.
↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.
Install in the Claude app (no terminal)
- Download the ZIP with the button below.
- In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
- Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
Install in Claude Code
Let Claude do it — paste this into Claude Code
Install the skill I found on AI Skill Mart. Copy the skills/bestpaper-writing folder from the GitHub repo gulucaptain/Best-Paper-Writing-Skill into my ~/.claude/skills/bestpaper-writing/. When it's done, tell me in one line what this skill can do.
Install with a command instead
git clone https://github.com/gulucaptain/Best-Paper-Writing-Skill.git /tmp/bpws && mkdir -p ~/.claude/skills && cp -r /tmp/bpws/skills/bestpaper-writing ~/.claude/skills/⚠ This is a third-party skill. Check the source repository before installing.
- Open a terminal (Terminal on macOS/Linux, WSL or Git Bash on Windows).
- Confirm Python 3 is available with
python3 --version; the skill's helper and checker scripts require it. - Clone the repository:
git clone https://github.com/gulucaptain/Best-Paper-Writing-Skill.git - Copy the skill into place:
mkdir -p ~/.claude/skills && cp -r Best-Paper-Writing-Skill/skills/bestpaper-writing ~/.claude/skills/ - Verify with
ls ~/.claude/skills/bestpaper-writingthatreferences/,scripts/andassets/came along. - Restart Claude Code and ask something like "rewrite my abstract in the style of a CVPR award paper" to trigger the skill.
- (Optional) If you have no material prepared, open
assets/intake_form.htmlin a browser, fill it in, export the JSON, and hand that file to Claude (note: the form is in Chinese).
View source on GitHub ↗License: NOASSERTION