Research Paper Writing
Helps an agent draft, polish, review, and compile academic ML/CV/NLP papers to a publication standard.
- Repository
- XiaomiMiMo/MiMo-Code
- Stars
- 13,743
- Licence
- MIT
- Files
- 15 files
What it does
Once installed, the agent behaves like an experienced co-author when you work on an academic paper. It routes your request to a matching guide (Abstract, Introduction, Method, Experiments, Conclusion, a hostile-reviewer checklist, or LaTeX-to-PDF compilation), checks that every claim has evidence, and refuses to invent numbers or citations, using visible placeholders instead. It can also edit .tex files in place while keeping citations, labels and math intact.
Great for
- Researchers preparing a conference submission (CVPR, ICCV, NeurIPS, ICLR, ACL style) who want a pre-submission critique
- Non-native English writers turning rough bullet points or Chinese notes into a clear English section
- Students who need a paragraph polished or want to understand how a strong abstract or introduction is structured
Skip it if
- Ghostwriting a paper you did not do the research for: the skill asks for facts and will not invent results, but it can still produce text that misrepresents your work if the inputs are wrong
- Non-academic writing or venues outside ML/CV/NLP, where the templates and review dimensions may not fit
Try asking
- “Polish this paragraph and tell me if the flow works.”
- “Here are my bullet points for the Introduction of my CVPR submission; draft it in English with a claim-evidence map.”
- “Review my paper as a skeptical NeurIPS reviewer and give me a prioritized fix list.”
- “Compile my main.tex to PDF and tell me how to fix the LaTeX error.”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 15 original files, the MIT licence and an attribution note.
research-paper-writing.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip research-paper-writing.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip research-paper-writing.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Give the agent the real numbers, baselines and dataset facts up front; without them it will leave placeholders like [XX.X] rather than guess.
- Ask for one section or one paragraph at a time to get the most focused edits.
- Check every citation and claim yourself before submission, since the skill flags unsupported claims but cannot verify references.
What's inside
- agents/openai.yaml
- references/abstract.md
- references/conclusion.md
- references/examples/abstract/template-a.md
- references/examples/abstract/template-b.md
- references/examples/abstract/template-c.md
- references/examples/introduction/novel-task-challenge-decomposition.md
- references/examples/introduction/pipeline-not-recommended-abstract-only.md
- references/examples/introduction/pipeline-version-1-one-contribution-multi-advantages.md
- references/examples/introduction/pipeline-version-2-two-contributions.md
- references/examples/introduction/pipeline-version-3-new-module-on-existing-pipeline.md
- references/examples/introduction/pipeline-version-4-observation-driven.md
- references/examples/introduction/technical-challenge-version-1-existing-task.md
- references/examples/introduction/technical-challenge-version-2-existing-task-insight-backed-by-traditional.md
- SKILL.md
Redistributed unmodified from XiaomiMiMo/MiMo-Code under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.