Nature Skills - An open-source AI research skills library focusing on the entire paper writing process.

What are Nature Skills?

Nature Skills is an open-source research skills library for AI scholars worldwide, focusing on the entire workflow of Nature/CNS papers. Nature Skills uses a routed SKILL.md architecture and includes 18 built-in skills covering paper writing, editing, peer review, figures and tables, citations, literature retrieval, and patent transfer. Nature Skills supports one-click installation of multiple agents such as Codex and Claude Code, and its rules are based on first-hand journal guidelines, with output ready for immediate use.

The main functions of Nature Skills

• Thesis writingDraft Nature-style manuscripts chapter by chapter and polish and translate them, supporting the entire process from research proposal to final manuscript.
• Review and revisionSimulates Nature reviewers' output of review reports and automatically generates point-by-point responses and cover letters.
• Chart StatisticsGenerate submission-level research figures and review p-values, replication counts, and effect sizes in statistical reports.
• Document ManagementIt covers the entire management chain, including retrieval, citation, verification, download, and automated document delivery.
• In-depth reading reportGenerates a bilingual (Chinese and English) reader with illustrations, structured reading cards, and presentation slides.
• Assisted conversionIt handles data availability claims, experimental log recording, and supports converting papers into patent drafts.

How to use Nature Skills

• Install Node.js 18+Ensure that the local environment meets the operational requirements.
• View available skills:run npx skills add Yuan1z0825/nature-skills --list List all skills.
• Global installation skill pack:implement npx skills add Yuan1z0825/nature-skills --global --agent codex --skill '*' --yes --copy Deploy all skills to Codex.
• Start a new sessionRestart the Agent session after installation to ensure the skills take effect.
• Describe the research task: Directly tell the Agent what you want to do using natural language, such as "polish this abstract in Nature style".
• Agent automatic routingThe system automatically matches and loads the corresponding skill rules based on the task content.
• Obtain ready-to-use products: Receive outputs from the Agent, such as text, charts, PPTs, or review responses.

Follow us on WeChat and reply with "open source",join inAI open source project discussion group

The core advantages of Nature Skills

• Route loadingAutomatically matches and loads only the necessary rule fragments based on tasks, chapters, and journals, avoiding redundancy.
• Nature system nativeAll rules are based on the official guidelines of Nature/CNS and published papers, not general aesthetic preferences.
• Full-process closed loopIt provides one-stop coverage from topic selection and argumentation, writing and polishing, peer review simulation to revision response, chart creation and literature management.
• Multi-Agent CompatibilitySupports one-click installation and automatic updates for mainstream agents such as Codex, Claude Code, and OpenClaw.
• Output is productIt directly returns usable deliverables such as pasteable text, submission-level charts, PPTs, and cover letters.
• The rules are transparent and explainable.Each writing/censorship rule explains its source and rationale, and explicitly marks any missing evidence and the boundaries of its application.
• international influenceThe design philosophy has been adopted by Google DeepMind and used to launch Science Skills, which has been recognized by top institutions.
• Modular and scalableEach skill is self-contained, new features are added without intruding on the existing system, and the community can contribute freely.

Nature Skills project address

• GitHub repository：https://github.com/Yuan1z0825/nature-skills

Comparison of Nature Skills with similar products

Application scenarios of Nature Skills

• Nature System Paper DraftBased on research data and figures, it automatically generates abstracts, introductions, methods, discussions, and other chapters in the Nature style, and reconstructs the argumentation logic.
• Pre-submission peer review simulationBefore formal submission, simulate the output of review reports by three mutually blind Nature reviewers to identify weaknesses in the manuscript and classify them as Major/Minor comments in advance.
• Response to each point of the repair feedbackIt analyzes journal revision emails and generates individual point-by-point responses, cover letters, and highlighted revised manuscripts for each reviewer.
• Submission-level scientific charts and statistical reviewGenerate Python/R research graphs that meet Nature's requirements and review the compliance of p-values, number of replicates, effect sizes, etc. in the statistical reports.
• Full-chain management and in-depth reading report of literatureIt automates retrieval, verifies citations, generates bilingual (Chinese and English) readers, and converts papers into Chinese group meeting presentation PPTs with a single click.