{"id":4163,"date":"2026-08-18T22:07:09","date_gmt":"2026-08-18T14:07:09","guid":{"rendered":"https:\/\/ai.wuaishare.cn\/hub\/?p=4163"},"modified":"2026-08-18T22:07:09","modified_gmt":"2026-08-18T14:07:09","slug":"nature-skills","status":"publish","type":"post","link":"https:\/\/ai.wuaishare.cn\/hub\/en\/4163","title":{"rendered":"Nature Skills"},"content":{"rendered":"
\n

\n \"Nature\n <\/p>\n

\n \"License\"<\/a>
\n \"Install\"<\/a>
\n \"Skills\"<\/a>
\n \"Language\"<\/a>\n <\/p>\n

\n Website<\/a>
\n \u00b7 Install<\/a>
\n \u00b7 Quick Start<\/a>
\n \u00b7 Skill Index<\/a>
\n \u00b7 Other Install<\/a>
\n \u00b7 Contributing<\/a>
\n \u00b7 \u4e2d\u6587<\/a>\n <\/p>\n<\/div>\n

—<\/p>\n

`nature-skills` collects reusable research skills for AI scholars worldwide, with an emphasis on real problem solving, verifiable workflows, and research outputs that can be used directly.<\/p>\n

## Table of Contents<\/p>\n

– [1. Project Founder and Operations](#1-project-founder-and-operations)
\n – [1.1 Founder Introduction](#11-founder-introduction)
\n – [1.2 Knowledge Planet](#12-knowledge-planet)
\n – [1.3 Repository Store](#13-repository-store)
\n – [1.4 Business Cooperation](#14-business-cooperation)
\n– [2. Nature Skills Core Developers](#2-nature-skills-core-developers)
\n– [3. Project Philosophy and Community](#3-project-philosophy-and-community)
\n– [4. Quick Start](#4-quick-start)
\n– [5. Installation](#5-installation)
\n – [5.1 `npx skills` Installation](#51-npx-skills-installation)
\n – [5.2 Claude Code Installation](#52-claude-code-installation)
\n – [5.3 Codex Installation](#53-codex-installation)
\n – [5.4 Other Agent Scenarios](#54-other-agent-scenarios)
\n– [6. Skill Index](#6-skill-index)
\n– [7. Contribution and Development](#7-contribution-and-development)
\n– [8. Star History](#8-star-history)<\/p>\n

## 1. Project Founder and Operations<\/p>\n

### 1.1 Founder Introduction<\/p>\n

Hello, I am Yizhe Yuan, the founder of `nature-skills`. Thank you for following this project. We have published many video tutorials on Douyin; search by topic name to find them, and I sincerely hope they help with real research work.<\/p>\n

### 1.2 Knowledge Planet<\/p>\n

Knowledge Planet name: **Nature Skills and the Philosophy Behind It**.<\/p>\n

\"Nature<\/p>\n

### 1.3 Nature AI GPT \/ Claude Top-ups and Ready-Made Accounts<\/p>\n

We carefully screen service channels and provide legitimate top-up channels and
\nsupport. Visit **Nature AI Recharge Card Store**:<\/p>\n\n\n
\n
\n \"Nature
\n <\/a>
\n
\n Nature AI Recharge Card Store<\/strong>
\n
\n https:\/\/apiciyuan.top\/<\/a>\n <\/td>\n		\n \"Nature
\n
\n WeChat Support<\/strong>
\n
\n Scan to add<\/sub>\n <\/td>\n<\/tr>\n<\/table>\n

### 1.4 Business Cooperation<\/p>\n

For business cooperation, please contact us at [natureskills2026@outlook.com](mailto:natureskills2026@outlook.com).<\/p>\n

## 2. Nature Skills Core Developers<\/p>\n

| Developer | Project Role | Main Focus and Contributions | Profiles and Contact |
\n|—|—|—|—|
\n| **Yizhe Yuan** | Founder \/ Maintainer | Project initiation, skill-system design, and community operations | \u2014 |
\n| **Xin-Rui Ma** | Core Developer | Core development and architecture maintenance | [Gmail](mailto:travisma2233@gmail.com) |
\n| **Bin Hu** | Major Contributor | Agentic agents and AI for Science | [GitHub](https:\/\/github.com\/Flyme886) \u00b7 [Email](mailto:mhoang12205@gmail.com) |<\/p>\n

## 3. Project Philosophy and Community<\/p>\n

### 3.1 Some Personal Views<\/p>\n

– Recently, I noticed that the Nature Skills design has drawn attention from
\n Google DeepMind and has been referenced by them. They drew on its citation
\n system, script ideas, and skill-design philosophy to launch Science Skills. To
\n be honest, this makes me pleased: when leading international AI institutions
\n begin to draw inspiration from our work, it means original ideas from Chinese
\n developers are being seen by the world. This is not a feeling of loss from
\n being copied, but a sign of Chinese strength taking root in open source and
\n naturally growing outward.
\n– The focus of our skill design has never been to require every user to fully
\n master the whole philosophy. The point is that the philosophy itself can be
\n understood and reused by machines. If you want to create a new skill or adapt
\n this system to your own field, you can give the Nature Skills GitHub repository
\n to Codex and let it learn the design pattern, then help you create or modify a
\n skill. This is how ideas become operational rather than remaining oral
\n explanations.
\n– The real value of Nature Skills may not be limited to any individual module.
\n It may be that the project has quietly opened a new door: many people realize
\n for the first time that Codex or other agents can operate a local computer for
\n research. I have been fortunate to witness and accompany many people through
\n this shift in research workflow. When they say, “so research can be done this
\n way,” that break in understanding and liberation of thought matters more to me
\n than the skills themselves. This is not merely the success of a tool, but the
\n beginning of a new way of thinking spreading among people.
\n– In practice, almost every useful tool can be distilled into a standardized
\n process, and standardized processes can be packaged as reusable skills.<\/p>\n

### 3.2 Tutorials and Community<\/p>\n\n\n
\n Follow Douyin for video tutorials<\/b>
\n \"Douyin\n <\/td>\n		\n Agent Research Community<\/b>
\n \"Agent\n <\/td>\n		\n Yuan’s personal WeChat<\/b>
\n \"personal\n <\/td>\n<\/tr>\n<\/table>\n

## 4. Quick Start<\/p>\n

After installation, you can give the agent a paper, paragraph, reviewer letter,
\nor task description directly. These prompts are ready to copy:<\/p>\n

| Goal | Prompt |
\n| — | — |
\n| Read a paper \/ bilingual reader | `Turn this PDF into a figure-aware Chinese-English Markdown reader.` |
\n| Generate a paper presentation | `Create a Chinese journal-club PPT from this paper, keeping key figures and source labels.` |
\n| Polish or translate a manuscript paragraph | `Rewrite this Chinese paragraph into Nature-style academic English without changing the meaning.` |
\n| Draft an abstract, introduction, or discussion | `Using these results and figures, draft a Nature-style abstract and introduction.` |
\n| Simulate pre-submission review | `Evaluate this manuscript from a Nature reviewer perspective, produce three mutually blind reviewer reports, and synthesize them only after all are final.` |
\n| Respond to reviewer comments | `Use this revision email to draft a separate point-by-point response for each mutually blind reviewer and a cover letter, and mark redline locations in the revised manuscript.` |
\n| Search literature, strict citations, and citer profiles | `Create a table with this paper’s citation count, strict external citation count, DOI, and whether major scholars or Fellows cited it.` |
\n| Create scientific figures or schematics | `Use this method and result description to draft a publication-ready scientific figure or manuscript schematic.` |<\/p>\n

If you are unsure which skill to use, describe the task naturally. If you already
\nknow the skill name, explicitly say “use `nature-reader`” or “use
\n`nature-response`” in the prompt.<\/p>\n

## 5. Installation<\/p>\n

`nature-skills` is a collection of reusable skill packages organized around
\n`SKILL.md`. Each top-level skill directory under `skills\/` is an installable unit,
\nsuch as `nature-*`; `nature-shared` is an installable support package read by
\nother skills.<\/p>\n

### 5.1 `npx skills` Installation<\/p>\n

Install [Node.js 18 or later](https:\/\/nodejs.org\/) first. The CLI does not need
\nto be installed globally. List the skill names available in this repository:<\/p>\n

“`bash
\nnpx skills add Yuan1z0825\/nature-skills –list
\n“`<\/p>\n

Install every skill globally for Codex. The complete selection includes
\n`nature-shared`, so skills that use the common references remain functional:<\/p>\n

“`bash
\nnpx skills add Yuan1z0825\/nature-skills –global –agent codex –skill ‘*’ –yes –copy
\n“`<\/p>\n

Omit `–global` to install one independent skill in the current project:<\/p>\n

“`bash
\nnpx skills add Yuan1z0825\/nature-skills –agent codex –skill nature-figure –yes –copy
\n“`<\/p>\n

When installing `nature-reader`, `nature-paper2ppt`, `nature-polishing`, or
\n`nature-writing` alone, select the shared support package as well:<\/p>\n

“`bash
\nnpx skills add Yuan1z0825\/nature-skills –global –agent codex \\
\n –skill nature-reader –skill nature-shared –yes –copy
\n“`<\/p>\n

Install all skills for every agent supported by the CLI:<\/p>\n

“`bash
\nnpx skills add Yuan1z0825\/nature-skills –all
\n“`<\/p>\n

Verify the global Codex installation and update it later:<\/p>\n

“`bash
\nnpx skills list –global –agent codex –json
\nnpx skills update –global –yes
\n“`<\/p>\n

Update one skill, or update only the current project’s skills:<\/p>\n

“`bash
\nnpx skills update nature-reader –global –yes
\nnpx skills update –project –yes
\n“`<\/p>\n

Pass the frontmatter name displayed by `–list` to `–skill`; for example, the
\n`nature-proposal-writer` directory is currently listed as `researchwrite`.
\n`npx skills` manages skill files only. Optional Python, R, browser, and MCP
\nruntime dependencies still need the separate setup described below.<\/p>\n

### 5.2 Claude Code Installation<\/p>\n

Claude Code cannot use `scripts\/update-codex-skills.sh` directly because that
\nscript only syncs skills into Codex’s `~\/.codex\/skills\/`. For Claude Code, keep a
\nstable local clone and create a subagent or slash command wrapper that points to
\nthe real `skills\/*\/SKILL.md`. This preserves the skill directory structure and
\nlets the workflow keep using `references\/`, `static\/`, `manifest.yaml`, scripts,
\nassets, and `skills\/nature-shared\/`.<\/p>\n

If Claude Code is not installed yet:<\/p>\n

“`bash
\nnpm install -g @anthropic-ai\/claude-code
\nclaude
\n“`<\/p>\n

Clone the repository to a stable path:<\/p>\n

“`bash
\nmkdir -p ~\/ai-skills
\ncd ~\/ai-skills
\ngit clone https:\/\/github.com\/Yuan1z0825\/nature-skills.git
\n“`<\/p>\n

Recommended method: create a Claude Code subagent wrapper for the skills you use
\noften. Example for `nature-reader`:<\/p>\n

“`bash
\nmkdir -p ~\/.claude\/agents
\ncat > ~\/.claude\/agents\/nature-reader.md <<'EOF'\n---\nname: nature-reader\ndescription: Use for Chinese-English paper reading, figure-aware translation, and source-grounded paper notes.\n---\n\nWhen invoked, first read `~\/ai-skills\/nature-skills\/skills\/nature-reader\/SKILL.md` and follow it as the governing workflow.\nRead supporting files from `~\/ai-skills\/nature-skills\/skills\/nature-reader\/` and `~\/ai-skills\/nature-skills\/skills\/nature-shared\/` only when needed.\nDo not replace this skill with a generic paper-reading response.\nEOF\n```\n\nThen start a new Claude Code session and ask for the subagent explicitly:\n\n```text\nUse the nature-reader subagent to turn this paper into a Chinese-English Markdown reader.\n```\n\nIf you prefer a slash command, create a command wrapper instead:\n\n```bash\nmkdir -p ~\/.claude\/commands\ncat > ~\/.claude\/commands\/nature-reader.md <<'EOF'\nRead `~\/ai-skills\/nature-skills\/skills\/nature-reader\/SKILL.md` first and follow it strictly.\nRead directly needed supporting files from `~\/ai-skills\/nature-skills\/skills\/nature-reader\/` and `~\/ai-skills\/nature-skills\/skills\/nature-shared\/`.\n\n$ARGUMENTS\nEOF\n```\n\nUse it inside Claude Code:\n\n```text\n\/nature-reader Turn this paper into a full Chinese-English side-by-side Markdown reader.\n```\n\nTo install other skills, replace `nature-reader` with the target directory name,\nsuch as `nature-polishing`, `nature-writing`, `nature-reviewer`,\n`nature-response`, or `nature-figure`. To update later:\n\n```bash\ncd ~\/ai-skills\/nature-skills\ngit pull\n```\n\nAs long as the wrapper still points to this stable clone path, no repeated file\ncopy is needed.\n\n**Auto-Update (Optional)**\n\nIf you want Claude Code to pull upstream updates automatically on every session\nstart, use `scripts\/autoupdate-skills.sh` together with a `SessionStart` hook.\n\nThis approach **copies** the skills straight into `~\/.claude\/skills\/` (Claude Code\nauto-discovers that directory and loads each skill by its directory name) instead\nof using the wrappers above. Pick whichever one you prefer.\n\nKeep a **dedicated** stable clone (used only to sync skills \u2014 don't make dev\ncommits inside it):\n\n```bash\nmkdir -p ~\/ai-skills\ngit clone https:\/\/github.com\/Yuan1z0825\/nature-skills.git ~\/ai-skills\/nature-skills\n```\n\nInstall once, copying the skills into Claude Code's skills directory:\n\n```bash\n~\/ai-skills\/nature-skills\/scripts\/autoupdate-skills.sh --force\n```\n\nThen add a `SessionStart` hook to `~\/.claude\/settings.json` (merge this entry into\nan existing `hooks` block rather than replacing it):\n\n```json\n{\n \"hooks\": {\n \"SessionStart\": [\n {\n \"hooks\": [\n {\n \"type\": \"command\",\n \"command\": \"$HOME\/ai-skills\/nature-skills\/scripts\/autoupdate-skills.sh\",\n \"async\": true,\n \"timeout\": 120\n }\n ]\n }\n ]\n }\n}\n```\n\n`async: true` runs it in the background so it never blocks startup. The script is\nsafe to run constantly: it skips the network if it already checked within the last\nhour, silently skips when offline or the pull fails (`exit 0`, never stalling a\nsession), re-syncs when the upstream HEAD changes or the installed copy drifts, and refuses to\nfast-forward a clone that has uncommitted changes. New skills take effect on the\n**next** session (the current one already loaded its skills). Each destination\nhas a separate log at\n`~\/.local\/state\/nature-skills\/\/autoupdate.log`.<\/p>\n

The destination and check interval are both configurable:<\/p>\n

“`bash
\n# Defaults to ~\/.claude\/skills; use –dest for another location, e.g. Codex:
\n~\/ai-skills\/nature-skills\/scripts\/autoupdate-skills.sh –dest ~\/.codex\/skills
\n# Check the network at most once per hour; still verify the local install each run:
\n~\/ai-skills\/nature-skills\/scripts\/autoupdate-skills.sh –throttle 3600
\n“`<\/p>\n

### 5.3 Codex Installation<\/p>\n

Use the repository script to install or update Codex skills. It syncs every
\ntop-level skill directory under `skills\/` and verifies the copied contents with
\n`diff`. It does not overwrite unrelated Codex skills.<\/p>\n

“`bash
\ngit clone https:\/\/github.com\/Yuan1z0825\/nature-skills.git
\ncd nature-skills
\nscripts\/update-codex-skills.sh –pull
\n“`<\/p>\n

If you already have a clone:<\/p>\n

“`bash
\ncd nature-skills
\nscripts\/update-codex-skills.sh –pull
\n“`<\/p>\n

Verify that the current Codex installation matches this checkout:<\/p>\n

“`bash
\nscripts\/update-codex-skills.sh –check
\n“`<\/p>\n

If you use this script for long-term updates and want to remove directories that
\nwere previously managed by this installer but no longer exist upstream:<\/p>\n

“`bash
\nscripts\/update-codex-skills.sh –pull –prune
\n“`<\/p>\n

`–prune` only removes directories recorded by this installer. It will not guess
\nor delete unrelated skills.<\/p>\n

You can also ask Codex to install the repository for you:<\/p>\n

“`text
\nInstall Codex skills from this repository:
\nhttps:\/\/github.com\/Yuan1z0825\/nature-skills.git<\/p>\n

Clone the repository, run scripts\/update-codex-skills.sh –pull, and then run
\nscripts\/update-codex-skills.sh –check to verify the installation. Keep complete
\nskill directories under skills\/; do not copy only SKILL.md.
\n“`<\/p>\n

To install only one skill, specify the skill name:<\/p>\n

“`text
\nInstall only nature-reader from this repository:
\nhttps:\/\/github.com\/Yuan1z0825\/nature-skills.git<\/p>\n

If the skill needs shared files, install skills\/nature-shared as well.
\n“`<\/p>\n

Key rule: keep the full directory structure. Many skills depend on
\n`references\/`, `static\/`, `manifest.yaml`, scripts, assets, or shared files.<\/p>\n

The installer does not install Python dependencies automatically. Install them
\nonly when you need the corresponding scripts or MCP services:<\/p>\n

“`bash
\npython -m pip install -r skills\/nature-paper-to-patent\/requirements.txt
\npython -m pip install -r skills\/nature-paper-to-patent\/scripts\/disclosure\/requirements-cnipa.txt # optional CNIPA published-patent search
\npython -m pip install -r skills\/nature-academic-search\/mcp-server\/requirements.txt
\n“`<\/p>\n

If you enable `nature-paper-to-patent` CNIPA published-patent search, also run `python -m playwright install chromium`.<\/p>\n

`nature-academic-search` also requires `PUBMED_EMAIL`. Optional Scopus,
\nScienceDirect, and other provider credentials should be configured locally and
\nmust not be committed to the repository.<\/p>\n

After installation, start a new Codex session and describe your task naturally,
\nfor example:<\/p>\n

“`text
\nTurn this paper into a full Chinese-English side-by-side Markdown reader.
\n“`<\/p>\n

“`text
\nCreate a Chinese PPT deck from this paper.
\n“`<\/p>\n

For OpenClaw, OpenCode, Hermes, and other open-source agent frameworks, see the [OpenClaw \/ OpenCode \/ Hermes integration guide](docs\/open-source-agent-frameworks_EN.md).<\/p>\n

**Auto-Update (Optional)**<\/p>\n

Codex supports a global `SessionStart` hook. With a dedicated clone, it can check
\nfor updates whenever a Codex session starts or resumes and sync new versions into
\n`~\/.codex\/skills\/`.<\/p>\n

Create the dedicated clone and perform the initial sync:<\/p>\n

“`bash
\nmkdir -p ~\/.codex
\ngit clone https:\/\/github.com\/Yuan1z0825\/nature-skills.git ~\/.codex\/.nature-skills-src
\n~\/.codex\/.nature-skills-src\/scripts\/autoupdate-skills.sh \\
\n –dest ~\/.codex\/skills –force
\n“`<\/p>\n

Then create or merge `~\/.codex\/hooks.json`:<\/p>\n

“`json
\n{
\n “hooks”: {
\n “SessionStart”: [
\n {
\n “matcher”: “startup|resume”,
\n “hooks”: [
\n {
\n “type”: “command”,
\n “command”: “\/bin\/bash \\”$HOME\/.codex\/.nature-skills-src\/scripts\/autoupdate-skills.sh\\” –dest \\”$HOME\/.codex\/skills\\””,
\n “timeout”: 120,
\n “statusMessage”: “Checking Nature Skills updates”
\n }
\n ]
\n }
\n ]
\n }
\n}
\n“`<\/p>\n

If `hooks.json` already contains other hooks, merge the `SessionStart` entry
\ninstead of replacing the file. After enabling or changing the hook, run `\/hooks`
\nin Codex to review and trust it. Codex currently runs command hooks synchronously,
\nso this setup relies on the script’s built-in 1-hour throttle, 60-second network
\nguard, and offline-safe exit to avoid repeated network checks or blocking startup
\nwhen an update cannot be fetched. Even when upstream is unchanged, the script
\nverifies the destination and repairs detected drift.<\/p>\n

Each destination has a separate log at
\n`~\/.local\/state\/nature-skills\/\/autoupdate.log`. Newly fetched
\nskills normally take full effect in the next session.<\/p>\n

### 5.4 Other Agent Scenarios<\/p>\n

For OpenClaw, OpenCode, and Hermes, see the dedicated [integration guide](docs\/open-source-agent-frameworks_EN.md).<\/p>\n

For other agents, keep a stable repository clone and create a lightweight
\nsubagent, slash command, or custom prompt wrapper that points to the real
\n`skills\/*\/SKILL.md` files. Preserve `skills\/nature-shared\/`.<\/p>\n

For manual or other-agent use:<\/p>\n

1. Copy complete skill directories into your prompt library or project.
\n2. Preserve `SKILL.md`, `manifest.yaml`, `static\/`, `references\/`, scripts,
\n assets, and required `skills\/nature-shared\/` files.
\n3. If the target agent has its own format requirements, adjust the frontmatter
\n and body structure.<\/p>\n

## 6. Skill Index<\/p>\n

The current `skills\/` directory contains the following triggerable skills.
\n`skills\/nature-shared\/` is shared content and is not counted in the skill index. Click a skill name or the “Details” link to open its dedicated documentation page.<\/p>\n

| Skill | Status | Purpose | Example Triggers | Details |
\n|—|—|—|—|—|
\n| [`nature-figure`](skills\/nature-figure\/README_EN.md) | Stable | Submission-grade Python or R scientific figure workflow for Nature \/ high-impact journals, with separately noticed third-party figures4papers references, original templates, and OpenRouter GPT Image 2 schematic drafts | “Nature figure”, “submission-grade figure”, “publication plot”, “scientific figure”, “figures4papers”, “paper schematic”, “GPT Image 2” | [Details](skills\/nature-figure\/README_EN.md) |
\n| [`nature-polishing`](skills\/nature-polishing\/README_EN.md) | Stable | Polish, restructure, or translate academic prose into Nature-style English, with manuscript-wide terminology, unit, precision, and claim-drift checks | “Nature style”, “polishing”, “academic writing”, “English manuscript” | [Details](skills\/nature-polishing\/README_EN.md) |
\n| [`nature-writing`](skills\/nature-writing\/README_EN.md) | Draft | Draft Nature-style manuscript sections and rebuild a paper argument | “Nature writing”, “write an abstract”, “write introduction”, “manuscript draft”, “paper writing” | [Details](skills\/nature-writing\/README_EN.md) |
\n| [`nature-reviewer`](skills\/nature-reviewer\/README_EN.md) | Draft | Simulate Nature-style reviewer assessment with three mutually blind reports, tiered Major\/Minor comments, and manuscript-internal consistency checks | “Nature reviewer”, “pre-submission review”, “reviewer report”, “reviewer-perspective assessment” | [Details](skills\/nature-reviewer\/README_EN.md) |
\n| [`nature-citation`](skills\/nature-citation\/README_EN.md) | Beta | Search support literature strictly within Nature \/ CNS families and export ENW, RIS, or Zotero RDF | “Nature citation”, “CNS citation”, “segmented citation”, “supporting references”, “Zotero RDF” | [Details](skills\/nature-citation\/README_EN.md) |
\n| [`nature-data`](skills\/nature-data\/README_EN.md) | Draft | Prepare Data Availability statements, data repository plans, and FAIR checks | “Data Availability”, “data availability”, “repository”, “FAIR metadata” | [Details](skills\/nature-data\/README_EN.md) |
\n| [`nature-statistics`](skills\/nature-statistics\/README_EN.md) | Draft | Audit, revise, or draft statistical reporting, covering experimental units, replicates, p values, multiple comparisons, effect sizes, confidence intervals, figure statistics, and cross-section numeric consistency | “Nature statistics”, “statistical analysis”, “p value”, “sample size”, “replicates”, “multiple comparisons”, “figure statistics”, “statistics review” | [Details](skills\/nature-statistics\/README_EN.md) |
\n| [`nature-reader`](skills\/nature-reader\/README_EN.md) | Beta | Generate full-paper Markdown readers with source anchors, figure-text alignment, rendered equations, and Chinese-English side-by-side translation | “nature reader”, “full Markdown”, “source-aligned text”, “figure-text alignment”, “rendered equations”, “full translation” | [Details](skills\/nature-reader\/README_EN.md) |
\n| [`nature-paper-card`](skills\/nature-paper-card\/README_EN.md) | Beta | Deep-read one paper into a source-grounded Sections 01\u201316 Paper Card covering method logic, experiment-to-claim evidence, conclusion boundaries, critical analysis, and testable research ideas | “nature paper card”, “deep-read paper”, “Paper Card”, “evidence chain”, “conclusion boundaries” | [Details](skills\/nature-paper-card\/README_EN.md) |
\n| [`nature-response`](skills\/nature-response\/README_EN.md) | Beta | Parse revision emails; create separate mutually blind reviewer responses, cover letters, red-marked manuscripts, LaTeX templates, and revision-package consistency checks | “response to reviewers”, “rebuttal letter”, “cover letter”, “major revision”, “revision email”, “reviewer-comment response”, “LaTeX template” | [Details](skills\/nature-response\/README_EN.md) |
\n| [`nature-paper2ppt`](skills\/nature-paper2ppt\/README_EN.md) | Beta | Generate Chinese PPTX journal-club or paper-presentation decks from research papers | “paper PPT”, “journal club”, “paper to slides”, “paper presentation” | [Details](skills\/nature-paper2ppt\/README_EN.md) |
\n| [`nature-image2ppt`](skills\/nature-image2ppt\/README_EN.md) | Beta | Reconstruct slide images, scanned PDFs, and image-only PPTX files as object-level editable PowerPoint with rendered QA | “image to editable PowerPoint”, “reconstruct slide screenshot”, “scanned PDF to PPTX”, “editable PPT” | [Details](skills\/nature-image2ppt\/README_EN.md) |
\n| [`nature-paper-to-patent`](skills\/nature-paper-to-patent\/README_EN.md) | Beta | Generate evidence-constrained Chinese invention patent drafts and support patent-point mining, prior-art search, and iterative technical disclosure drafting | “paper to patent”, “Chinese patent”, “paper-to-patent”, “claims drafting”, “technical disclosure”, “patent points” | [Details](skills\/nature-paper-to-patent\/README_EN.md) |
\n| [`nature-ref-verifier`](skills\/nature-ref-verifier\/README_EN.md) | Stable | Cross-check references across multiple sources and flag author, title, year, volume, issue, and page inconsistencies | “verify refs”, “check references”, “reference verification”, “ref check” | [Details](skills\/nature-ref-verifier\/README_EN.md) |
\n| [`nature-academic-search`](skills\/nature-academic-search\/README_EN.md) | Beta | Multi-source literature search, citation verification, strict other-citation audits, article-level citation metric tables, influential citer profiling, and reference management | “search papers”, “find articles”, “literature search”, “literature lookup”, “verify DOI”, “strict other citation”, “article citation table”, “influential citer” | [Details](skills\/nature-academic-search\/README_EN.md) |
\n| [`nature-downloader`](skills\/nature-downloader\/README_EN.md) | Beta | Legally obtain academic full text\/PDFs through library access, Chrome login state, and open-access routes | “download papers”, “library paper download”, “CARSI”, “Web of Science”, “PDF download” | [Details](skills\/nature-downloader\/README_EN.md) |
\n| [`nature-literature-pipeline`](skills\/nature-literature-pipeline\/README_EN.md) | Stable | Automated literature discovery pipeline: multi-source retrieval, six-axis scoring, deep-reading delivery, and local archiving | “literature pipeline”, “daily literature”, “literature push”, “daily literature push”, “cron” | [Details](skills\/nature-literature-pipeline\/README_EN.md) |
\n| [`nature-experiment-log`](skills\/nature-experiment-log\/README_EN.md) | Draft | Standardize experiment images, voice, and text into Obsidian experiment logs with YAML frontmatter and archived source materials | “experiment log”, “record experiment”, “Obsidian vault”, “Feishu research group” | [Details](skills\/nature-experiment-log\/README_EN.md) |
\n| [`nature-proposal-writer`](skills\/nature-proposal-writer\/README_EN.md) | Beta | Proposal-first research writing state machine: establish evidence, argument, and section contracts before drafting or reviewing text | “researchwrite”, “proposal”, “opening report”, “research plan”, “research writing QA” | [Details](skills\/nature-proposal-writer\/README_EN.md) |<\/p>\n

## 7. Contribution and Development<\/p>\n

### 7.1 Shared Design Principles<\/p>\n

1. **Prefer primary sources**: rules should be grounded in published Nature
\n content, official journal guidance, or explicit local sources rather than
\n generic taste.
\n2. **Make rules explicit**: explain the reason behind each rule instead of
\n giving unsupported assertions.
\n3. **Respect section and task context**: writing, figures, citations, and
\n responses depend on the manuscript section and task.
\n4. **Output first**: every skill should produce something directly usable, such
\n as paste-ready text, `.svg`, `.pptx`, `.docx`, or concrete instructions.
\n5. **Keep skills extensible**: each skill should be self-contained, and adding a
\n new skill should not require modifying existing skills.<\/p>\n

### 7.2 Repository Layout<\/p>\n

“`text
\nskills\/
\n\u251c\u2500\u2500 nature-shared\/ # keep this when skills reference ..\/nature-shared
\n\u251c\u2500\u2500 nature-\/
\n\u2502 \u251c\u2500\u2500 README.md
\n\u2502 \u251c\u2500\u2500 README_EN.md
\n\u2502 \u251c\u2500\u2500 SKILL.md
\n\u2502 \u251c\u2500\u2500 manifest.yaml # present in router-style skills
\n\u2502 \u251c\u2500\u2500 static\/ # present in router-style skills
\n\u2502 \u2514\u2500\u2500 references\/…
\n\u2514\u2500\u2500 nature-proposal-writer\/
\n \u251c\u2500\u2500 README.md
\n \u251c\u2500\u2500 README_EN.md
\n \u251c\u2500\u2500 SKILL.md
\n \u251c\u2500\u2500 scripts\/…
\n \u251c\u2500\u2500 templates\/…
\n \u2514\u2500\u2500 references\/…
\n“`<\/p>\n

### 7.3 Adding a Skill<\/p>\n

When adding a skill to this repository, follow this process.<\/p>\n

**1. Create the skill directory**<\/p>\n

“`text
\nskills\/nature-\/
\n“`<\/p>\n

**2. Add required files**<\/p>\n

| File | Required | Purpose |
\n|—|—:|—|
\n| `SKILL.md` | Yes | Frontmatter (`name`, `description`) plus rules and workflow loaded by the agent |
\n| `README.md` | Yes | Human-facing Chinese documentation |
\n| `README_EN.md` | Yes | English documentation paired with the Chinese details page |
\n| `references\/*.md` | Recommended for complex skills | Modular rule files, API references, design theory, tutorials, chart types, and similar material |<\/p>\n

**3. Write the Chinese and English READMEs**<\/p>\n

Every new skill must include both `README.md` and `README_EN.md`. The README is a human-facing entry page, not a duplicate of `SKILL.md` and not an installation manual. Its job is to help users decide within 30 seconds whether the skill fits their task, what they need to provide, what it will output, and where its boundaries are.<\/p>\n

Basic rules:<\/p>\n

– The Chinese and English README files must be one-to-one mirrors: same heading count, same order, and same information points. Do not let the English page become a separate template.
\n– Start with the skill name, language switch link, and one positioning sentence.
\n– Use the base structure below by default; add optional sections only when they are genuinely needed.
\n– Do not repeat repository installation instructions, author bios, changelogs, development history, full file trees, or long internal implementation details inside a single skill README.
\n– Put complex rules, API parameters, long tutorials, script explanations, and template indexes in `references\/`, `static\/`, `scripts\/`, or `SKILL.md`. Keep only navigational pointers in the README.
\n– If the skill has visual assets, a small preview table is fine; do not turn the README into a large gallery or long technical manual.<\/p>\n

Chinese README base structure:<\/p>\n

“`markdown
\n# `nature-` \u6280\u80fd<\/p>\n

[English](README_EN.md)<\/p>\n

\u4e00\u53e5\u8bdd\u8bf4\u660e\u8fd9\u4e2a\u6280\u80fd\u7684\u5b9a\u4f4d\u3001\u4e3b\u8981\u4efb\u52a1\u548c\u4f7f\u7528\u8fb9\u754c\u3002<\/p>\n

## \u9002\u5408\u7528\u5b83\u505a\u4ec0\u4e48
\n## \u5178\u578b\u8bf7\u6c42
\n## \u4f60\u9700\u8981\u63d0\u4f9b
\n## \u4ea7\u51fa
\n## \u8fb9\u754c
\n## \u76f8\u5173\u6280\u80fd
\n“`<\/p>\n

The English README must mirror it as:<\/p>\n

“`markdown
\n# `nature-` Skill<\/p>\n

[\u4e2d\u6587\u8bf4\u660e](README.md)<\/p>\n

One sentence describing the skill’s role, main task, and usage boundary.<\/p>\n

## What To Use It For
\n## Typical Requests
\n## What You Need To Provide
\n## Outputs
\n## Boundaries
\n## Related Skills
\n“`<\/p>\n

Optional sections must be inserted in both languages and in the same order. Common optional sections:<\/p>\n

| Chinese Section | English Section | Use Case |
\n|—|—|—|
\n| `## \u5de5\u4f5c\u65b9\u5f0f` | `## Workflow` | Explain the core workflow or routing behavior |
\n| `## \u8fd0\u884c\u548c\u4f9d\u8d56` | `## Runtime and Dependencies` | Scripts, MCP services, API keys, local config, or external dependencies |
\n| `## \u793a\u4f8b\u9884\u89c8` | `## Example Preview` | A few figures, screenshots, or visual assets are worth showing |
\n| `## \u5185\u7f6e\u53c2\u8003` | `## Built-In References` | Point to `references\/`, `assets\/`, or demos |
\n| `## \u65b9\u6cd5\u6765\u6e90` | `## Method Sources` | Writing, review, or analysis rules come from specific sources |
\n| `## \u4e09\u79cd\u6a21\u5f0f` | `## Three Modes` | The skill has clear compose\/revise\/hybrid-style modes |
\n| `## \u4e0e … \u7684\u5173\u7cfb` | `## Relationship With …` | The skill is easy to confuse with another skill and needs a division-of-labor note |<\/p>\n

Before submitting, run at least these README checks:<\/p>\n

“`bash
\npython scripts\/validate-readmes.py
\ngit diff –check
\nfor d in skills\/nature-*; do
\n [ -f “$d\/README.md” ] && [ -f “$d\/README_EN.md” ] || continue
\n rg -q ‘^\\[English\\]\\(README_EN\\.md\\)$’ “$d\/README.md”
\n rg -q ‘^\\[\u4e2d\u6587\u8bf4\u660e\\]\\(README\\.md\\)$’ “$d\/README_EN.md”
\n test “$(rg -c ‘^## ‘ “$d\/README.md”)” = “$(rg -c ‘^## ‘ “$d\/README_EN.md”)”
\ndone
\n“`<\/p>\n

**4. Record a usage tutorial**<\/p>\n

When submitting a PR, please also record a short usage tutorial explaining what
\nproblem the skill solves, how to trigger it, what inputs it needs, and what
\noutputs it produces. Add the video, screencast link, or public tutorial URL to
\nthe PR description.<\/p>\n

**5. Configure `SKILL.md` frontmatter**<\/p>\n

“`yaml
\n—
\nname: nature-
\ndescription: >-
\n One sentence explaining what the skill does, when it should trigger, primary
\n outputs, and core use cases.
\n—
\n“`<\/p>\n

**6. Update the skill index**<\/p>\n

After adding a skill, update the [Skill Index](#6-skill-index) table:<\/p>\n

“`markdown
\n| [`nature-`](skills\/nature-\/README_EN.md) | Draft \/ Stable | One-sentence purpose | Trigger terms | [Details](skills\/nature-\/README_EN.md) |
\n“`<\/p>\n

**7. Set the status label**<\/p>\n

| Status | Meaning |
\n|—|—|
\n| `Draft` | Rules are defined but not yet tested on real cases |
\n| `Beta` | Tested on examples, with possible edge-case issues |
\n| `Stable` | Validated on real academic content and relatively stable |<\/p>\n

—<\/p>\n

## 8. Star History<\/p>\n

[![Star History Chart](assets\/star-history-20260816T024352Z.svg)](https:\/\/star-history.com\/#Yuan1z0825\/nature-skills&Date)<\/p>\n","protected":false},"excerpt":{"rendered":"

Nature Skills is a collection of reusable research skills for AI scholars, focused on real problem solving, verifiable workflows, and directly usable research outputs.<\/p>\n","protected":false},"author":0,"featured_media":3855,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ai_share_source_url":"https:\/\/github.com\/Yuan1z0825\/nature-skills","ai_share_original_content_language":"en","ai_share_i18n_original_source_language":"en","ai_share_tool_icon_attachment_id":0,"ai_share_author_label":"yuan1z0825","ai_share_tool_type":"tool","ai_share_resource_type":"workflow_skill_pack","ai_share_platform_support":"Claude Skills,GitHub,Python","ai_share_provider_observed_at":"2026-08-18T08:31:00.711Z","ai_share_last_commit_at":"2026-08-16T08:15:00Z","ai_share_license":"Apache-2.0","ai_share_language_support":"zh-CN,en","ai_share_github_topics":"codex-skills, nature, nature-skills","ai_share_repo_default_branch":"main","ai_share_pricing_model":"open_source","ai_share_deployment_mode":"hybrid","ai_share_delivery_mode":"python","ai_share_region_availability":"global","ai_share_source_platform":"GitHub","ai_share_github_releases_json":"[]","ai_share_first_party_view_count":53,"ai_share_github_stars":35775,"ai_share_github_watchers":44,"ai_share_github_forks":2002,"ai_share_github_open_issues":1,"ai_share_open_source":true,"ai_share_repo_url":"https:\/\/github.com\/Yuan1z0825\/nature-skills","ai_share_documentation_url":"https:\/\/star-history.com","creator_id":"github-yuan1z0825","creator_slug":"yuan1z0825","creator_name":"yuan1z0825","creator_type":"person","creator_status":"external","creator_intro":"Yuan1z0825 maintains this GitHub repository.","creator_github_url":"https:\/\/github.com\/yuan1z0825","creator_avatar_url":"https:\/\/avatars.githubusercontent.com\/u\/140061814?v=4","ai_share_github_readme_path":"README_EN.md","ai_share_github_readme_hash":"6f53117372bd09bb7e1bd11e8771dfbb508bcf5a5a91d1116c360a8d4ad77a9c","ai_share_github_readme_language":"en","ai_share_github_readme_format":"markdown","ai_share_github_description":"A Skill for Nature Journal\u2013Style Academic Writing and Scientific Figure Design","ai_share_one_sentence_definition_en":"Reusable research skills for AI scholars, with verifiable workflows and directly usable research outputs.","ai_share_github_readme_url":"https:\/\/github.com\/Yuan1z0825\/nature-skills\/blob\/main\/README_EN.md","ai_share_github_readme_raw_url":"https:\/\/raw.githubusercontent.com\/Yuan1z0825\/nature-skills\/main\/README_EN.md","ai_share_github_readme_markdown":"

\n

\n \"Nature\n <\/p>\n

\n \"License\"<\/a>\n \"Install\"<\/a>\n \"Skills\"<\/a>\n \"Language\"<\/a>\n <\/p>\n

\n Website<\/a>\n \u00b7 Install<\/a>\n \u00b7 Quick Start<\/a>\n \u00b7 Skill Index<\/a>\n \u00b7 Other Install<\/a>\n \u00b7 Contributing<\/a>\n \u00b7 \u4e2d\u6587<\/a>\n <\/p>\n<\/div>\n\n---\n\n`nature-skills` collects reusable research skills for AI scholars worldwide, with an emphasis on real problem solving, verifiable workflows, and research outputs that can be used directly.\n\n## Table of Contents\n\n- [1. Project Founder and Operations](#1-project-founder-and-operations)\n - [1.1 Founder Introduction](#11-founder-introduction)\n - [1.2 Knowledge Planet](#12-knowledge-planet)\n - [1.3 Repository Store](#13-repository-store)\n - [1.4 Business Cooperation](#14-business-cooperation)\n- [2. Nature Skills Core Developers](#2-nature-skills-core-developers)\n- [3. Project Philosophy and Community](#3-project-philosophy-and-community)\n- [4. Quick Start](#4-quick-start)\n- [5. Installation](#5-installation)\n - [5.1 `npx skills` Installation](#51-npx-skills-installation)\n - [5.2 Claude Code Installation](#52-claude-code-installation)\n - [5.3 Codex Installation](#53-codex-installation)\n - [5.4 Other Agent Scenarios](#54-other-agent-scenarios)\n- [6. Skill Index](#6-skill-index)\n- [7. Contribution and Development](#7-contribution-and-development)\n- [8. Star History](#8-star-history)\n\n## 1. Project Founder and Operations\n\n### 1.1 Founder Introduction\n\nHello, I am Yizhe Yuan, the founder of `nature-skills`. Thank you for following this project. We have published many video tutorials on Douyin; search by topic name to find them, and I sincerely hope they help with real research work.\n\n### 1.2 Knowledge Planet\n\nKnowledge Planet name: **Nature Skills and the Philosophy Behind It**.\n\n\"Nature\n\n### 1.3 Nature AI GPT \/ Claude Top-ups and Ready-Made Accounts\n\nWe carefully screen service channels and provide legitimate top-up channels and\nsupport. Visit **Nature AI Recharge Card Store**:\n\n\n \n
\n \n \"Nature\n <\/a>\n
\n Nature AI Recharge Card Store<\/strong>\n
\n https:\/\/apiciyuan.top\/<\/a>\n <\/td>\n 		\n \"Nature\n
\n WeChat Support<\/strong>\n
\n Scan to add<\/sub>\n <\/td>\n <\/tr>\n<\/table>\n\n### 1.4 Business Cooperation\n\nFor business cooperation, please contact us at [natureskills2026@outlook.com](mailto:natureskills2026@outlook.com).\n\n## 2. Nature Skills Core Developers\n\n| Developer | Project Role | Main Focus and Contributions | Profiles and Contact |\n|---|---|---|---|\n| **Yizhe Yuan** | Founder \/ Maintainer | Project initiation, skill-system design, and community operations | \u2014 |\n| **Xin-Rui Ma** | Core Developer | Core development and architecture maintenance | [Gmail](mailto:travisma2233@gmail.com) |\n| **Bin Hu** | Major Contributor | Agentic agents and AI for Science | [GitHub](https:\/\/github.com\/Flyme886) \u00b7 [Email](mailto:mhoang12205@gmail.com) |\n\n## 3. Project Philosophy and Community\n\n### 3.1 Some Personal Views\n\n- Recently, I noticed that the Nature Skills design has drawn attention from\n Google DeepMind and has been referenced by them. They drew on its citation\n system, script ideas, and skill-design philosophy to launch Science Skills. To\n be honest, this makes me pleased: when leading international AI institutions\n begin to draw inspiration from our work, it means original ideas from Chinese\n developers are being seen by the world. This is not a feeling of loss from\n being copied, but a sign of Chinese strength taking root in open source and\n naturally growing outward.\n- The focus of our skill design has never been to require every user to fully\n master the whole philosophy. The point is that the philosophy itself can be\n understood and reused by machines. If you want to create a new skill or adapt\n this system to your own field, you can give the Nature Skills GitHub repository\n to Codex and let it learn the design pattern, then help you create or modify a\n skill. This is how ideas become operational rather than remaining oral\n explanations.\n- The real value of Nature Skills may not be limited to any individual module.\n It may be that the project has quietly opened a new door: many people realize\n for the first time that Codex or other agents can operate a local computer for\n research. I have been fortunate to witness and accompany many people through\n this shift in research workflow. When they say, \"so research can be done this\n way,\" that break in understanding and liberation of thought matters more to me\n than the skills themselves. This is not merely the success of a tool, but the\n beginning of a new way of thinking spreading among people.\n- In practice, almost every useful tool can be distilled into a standardized\n process, and standardized processes can be packaged as reusable skills.\n\n### 3.2 Tutorials and Community\n\n\n \n
\n Follow Douyin for video tutorials<\/b>
\n \"Douyin\n <\/td>\n 		\n Agent Research Community<\/b>
\n \"Agent\n <\/td>\n 		\n Yuan's personal WeChat<\/b>
\n \"personal\n <\/td>\n <\/tr>\n<\/table>\n\n## 4. Quick Start\n\nAfter installation, you can give the agent a paper, paragraph, reviewer letter,\nor task description directly. These prompts are ready to copy:\n\n| Goal | Prompt |\n| --- | --- |\n| Read a paper \/ bilingual reader | `Turn this PDF into a figure-aware Chinese-English Markdown reader.` |\n| Generate a paper presentation | `Create a Chinese journal-club PPT from this paper, keeping key figures and source labels.` |\n| Polish or translate a manuscript paragraph | `Rewrite this Chinese paragraph into Nature-style academic English without changing the meaning.` |\n| Draft an abstract, introduction, or discussion | `Using these results and figures, draft a Nature-style abstract and introduction.` |\n| Simulate pre-submission review | `Evaluate this manuscript from a Nature reviewer perspective, produce three mutually blind reviewer reports, and synthesize them only after all are final.` |\n| Respond to reviewer comments | `Use this revision email to draft a separate point-by-point response for each mutually blind reviewer and a cover letter, and mark redline locations in the revised manuscript.` |\n| Search literature, strict citations, and citer profiles | `Create a table with this paper's citation count, strict external citation count, DOI, and whether major scholars or Fellows cited it.` |\n| Create scientific figures or schematics | `Use this method and result description to draft a publication-ready scientific figure or manuscript schematic.` |\n\nIf you are unsure which skill to use, describe the task naturally. If you already\nknow the skill name, explicitly say \"use `nature-reader`\" or \"use\n`nature-response`\" in the prompt.\n\n## 5. Installation\n\n`nature-skills` is a collection of reusable skill packages organized around\n`SKILL.md`. Each top-level skill directory under `skills\/` is an installable unit,\nsuch as `nature-*`; `nature-shared` is an installable support package read by\nother skills.\n\n### 5.1 `npx skills` Installation\n\nInstall [Node.js 18 or later](https:\/\/nodejs.org\/) first. The CLI does not need\nto be installed globally. List the skill names available in this repository:\n\n```bash\nnpx skills add Yuan1z0825\/nature-skills --list\n```\n\nInstall every skill globally for Codex. The complete selection includes\n`nature-shared`, so skills that use the common references remain functional:\n\n```bash\nnpx skills add Yuan1z0825\/nature-skills --global --agent codex --skill '*' --yes --copy\n```\n\nOmit `--global` to install one independent skill in the current project:\n\n```bash\nnpx skills add Yuan1z0825\/nature-skills --agent codex --skill nature-figure --yes --copy\n```\n\nWhen installing `nature-reader`, `nature-paper2ppt`, `nature-polishing`, or\n`nature-writing` alone, select the shared support package as well:\n\n```bash\nnpx skills add Yuan1z0825\/nature-skills --global --agent codex \\\n --skill nature-reader --skill nature-shared --yes --copy\n```\n\nInstall all skills for every agent supported by the CLI:\n\n```bash\nnpx skills add Yuan1z0825\/nature-skills --all\n```\n\nVerify the global Codex installation and update it later:\n\n```bash\nnpx skills list --global --agent codex --json\nnpx skills update --global --yes\n```\n\nUpdate one skill, or update only the current project's skills:\n\n```bash\nnpx skills update nature-reader --global --yes\nnpx skills update --project --yes\n```\n\nPass the frontmatter name displayed by `--list` to `--skill`; for example, the\n`nature-proposal-writer` directory is currently listed as `researchwrite`.\n`npx skills` manages skill files only. Optional Python, R, browser, and MCP\nruntime dependencies still need the separate setup described below.\n\n### 5.2 Claude Code Installation\n\nClaude Code cannot use `scripts\/update-codex-skills.sh` directly because that\nscript only syncs skills into Codex's `~\/.codex\/skills\/`. For Claude Code, keep a\nstable local clone and create a subagent or slash command wrapper that points to\nthe real `skills\/*\/SKILL.md`. This preserves the skill directory structure and\nlets the workflow keep using `references\/`, `static\/`, `manifest.yaml`, scripts,\nassets, and `skills\/nature-shared\/`.\n\nIf Claude Code is not installed yet:\n\n```bash\nnpm install -g @anthropic-ai\/claude-code\nclaude\n```\n\nClone the repository to a stable path:\n\n```bash\nmkdir -p ~\/ai-skills\ncd ~\/ai-skills\ngit clone https:\/\/github.com\/Yuan1z0825\/nature-skills.git\n```\n\nRecommended method: create a Claude Code subagent wrapper for the skills you use\noften. Example for `nature-reader`:\n\n```bash\nmkdir -p ~\/.claude\/agents\ncat > ~\/.claude\/agents\/nature-reader.md <<'EOF'\n---\nname: nature-reader\ndescription: Use for Chinese-English paper reading, figure-aware translation, and source-grounded paper notes.\n---\n\nWhen invoked, first read `~\/ai-skills\/nature-skills\/skills\/nature-reader\/SKILL.md` and follow it as the governing workflow.\nRead supporting files from `~\/ai-skills\/nature-skills\/skills\/nature-reader\/` and `~\/ai-skills\/nature-skills\/skills\/nature-shared\/` only when needed.\nDo not replace this skill with a generic paper-reading response.\nEOF\n```\n\nThen start a new Claude Code session and ask for the subagent explicitly:\n\n```text\nUse the nature-reader subagent to turn this paper into a Chinese-English Markdown reader.\n```\n\nIf you prefer a slash command, create a command wrapper instead:\n\n```bash\nmkdir -p ~\/.claude\/commands\ncat > ~\/.claude\/commands\/nature-reader.md <<'EOF'\nRead `~\/ai-skills\/nature-skills\/skills\/nature-reader\/SKILL.md` first and follow it strictly.\nRead directly needed supporting files from `~\/ai-skills\/nature-skills\/skills\/nature-reader\/` and `~\/ai-skills\/nature-skills\/skills\/nature-shared\/`.\n\n$ARGUMENTS\nEOF\n```\n\nUse it inside Claude Code:\n\n```text\n\/nature-reader Turn this paper into a full Chinese-English side-by-side Markdown reader.\n```\n\nTo install other skills, replace `nature-reader` with the target directory name,\nsuch as `nature-polishing`, `nature-writing`, `nature-reviewer`,\n`nature-response`, or `nature-figure`. To update later:\n\n```bash\ncd ~\/ai-skills\/nature-skills\ngit pull\n```\n\nAs long as the wrapper still points to this stable clone path, no repeated file\ncopy is needed.\n\n**Auto-Update (Optional)**\n\nIf you want Claude Code to pull upstream updates automatically on every session\nstart, use `scripts\/autoupdate-skills.sh` together with a `SessionStart` hook.\n\nThis approach **copies** the skills straight into `~\/.claude\/skills\/` (Claude Code\nauto-discovers that directory and loads each skill by its directory name) instead\nof using the wrappers above. Pick whichever one you prefer.\n\nKeep a **dedicated** stable clone (used only to sync skills \u2014 don't make dev\ncommits inside it):\n\n```bash\nmkdir -p ~\/ai-skills\ngit clone https:\/\/github.com\/Yuan1z0825\/nature-skills.git ~\/ai-skills\/nature-skills\n```\n\nInstall once, copying the skills into Claude Code's skills directory:\n\n```bash\n~\/ai-skills\/nature-skills\/scripts\/autoupdate-skills.sh --force\n```\n\nThen add a `SessionStart` hook to `~\/.claude\/settings.json` (merge this entry into\nan existing `hooks` block rather than replacing it):\n\n```json\n{\n \"hooks\": {\n \"SessionStart\": [\n {\n \"hooks\": [\n {\n \"type\": \"command\",\n \"command\": \"$HOME\/ai-skills\/nature-skills\/scripts\/autoupdate-skills.sh\",\n \"async\": true,\n \"timeout\": 120\n }\n ]\n }\n ]\n }\n}\n```\n\n`async: true` runs it in the background so it never blocks startup. The script is\nsafe to run constantly: it skips the network if it already checked within the last\nhour, silently skips when offline or the pull fails (`exit 0`, never stalling a\nsession), re-syncs when the upstream HEAD changes or the installed copy drifts, and refuses to\nfast-forward a clone that has uncommitted changes. New skills take effect on the\n**next** session (the current one already loaded its skills). Each destination\nhas a separate log at\n`~\/.local\/state\/nature-skills\/\/autoupdate.log`.\n\nThe destination and check interval are both configurable:\n\n```bash\n# Defaults to ~\/.claude\/skills; use --dest for another location, e.g. Codex:\n~\/ai-skills\/nature-skills\/scripts\/autoupdate-skills.sh --dest ~\/.codex\/skills\n# Check the network at most once per hour; still verify the local install each run:\n~\/ai-skills\/nature-skills\/scripts\/autoupdate-skills.sh --throttle 3600\n```\n\n### 5.3 Codex Installation\n\nUse the repository script to install or update Codex skills. It syncs every\ntop-level skill directory under `skills\/` and verifies the copied contents with\n`diff`. It does not overwrite unrelated Codex skills.\n\n```bash\ngit clone https:\/\/github.com\/Yuan1z0825\/nature-skills.git\ncd nature-skills\nscripts\/update-codex-skills.sh --pull\n```\n\nIf you already have a clone:\n\n```bash\ncd nature-skills\nscripts\/update-codex-skills.sh --pull\n```\n\nVerify that the current Codex installation matches this checkout:\n\n```bash\nscripts\/update-codex-skills.sh --check\n```\n\nIf you use this script for long-term updates and want to remove directories that\nwere previously managed by this installer but no longer exist upstream:\n\n```bash\nscripts\/update-codex-skills.sh --pull --prune\n```\n\n`--prune` only removes directories recorded by this installer. It will not guess\nor delete unrelated skills.\n\nYou can also ask Codex to install the repository for you:\n\n```text\nInstall Codex skills from this repository:\nhttps:\/\/github.com\/Yuan1z0825\/nature-skills.git\n\nClone the repository, run scripts\/update-codex-skills.sh --pull, and then run\nscripts\/update-codex-skills.sh --check to verify the installation. Keep complete\nskill directories under skills\/; do not copy only SKILL.md.\n```\n\nTo install only one skill, specify the skill name:\n\n```text\nInstall only nature-reader from this repository:\nhttps:\/\/github.com\/Yuan1z0825\/nature-skills.git\n\nIf the skill needs shared files, install skills\/nature-shared as well.\n```\n\nKey rule: keep the full directory structure. Many skills depend on\n`references\/`, `static\/`, `manifest.yaml`, scripts, assets, or shared files.\n\nThe installer does not install Python dependencies automatically. Install them\nonly when you need the corresponding scripts or MCP services:\n\n```bash\npython -m pip install -r skills\/nature-paper-to-patent\/requirements.txt\npython -m pip install -r skills\/nature-paper-to-patent\/scripts\/disclosure\/requirements-cnipa.txt # optional CNIPA published-patent search\npython -m pip install -r skills\/nature-academic-search\/mcp-server\/requirements.txt\n```\n\nIf you enable `nature-paper-to-patent` CNIPA published-patent search, also run `python -m playwright install chromium`.\n\n`nature-academic-search` also requires `PUBMED_EMAIL`. Optional Scopus,\nScienceDirect, and other provider credentials should be configured locally and\nmust not be committed to the repository.\n\nAfter installation, start a new Codex session and describe your task naturally,\nfor example:\n\n```text\nTurn this paper into a full Chinese-English side-by-side Markdown reader.\n```\n\n```text\nCreate a Chinese PPT deck from this paper.\n```\n\nFor OpenClaw, OpenCode, Hermes, and other open-source agent frameworks, see the [OpenClaw \/ OpenCode \/ Hermes integration guide](docs\/open-source-agent-frameworks_EN.md).\n\n**Auto-Update (Optional)**\n\nCodex supports a global `SessionStart` hook. With a dedicated clone, it can check\nfor updates whenever a Codex session starts or resumes and sync new versions into\n`~\/.codex\/skills\/`.\n\nCreate the dedicated clone and perform the initial sync:\n\n```bash\nmkdir -p ~\/.codex\ngit clone https:\/\/github.com\/Yuan1z0825\/nature-skills.git ~\/.codex\/.nature-skills-src\n~\/.codex\/.nature-skills-src\/scripts\/autoupdate-skills.sh \\\n --dest ~\/.codex\/skills --force\n```\n\nThen create or merge `~\/.codex\/hooks.json`:\n\n```json\n{\n \"hooks\": {\n \"SessionStart\": [\n {\n \"matcher\": \"startup|resume\",\n \"hooks\": [\n {\n \"type\": \"command\",\n \"command\": \"\/bin\/bash \\\"$HOME\/.codex\/.nature-skills-src\/scripts\/autoupdate-skills.sh\\\" --dest \\\"$HOME\/.codex\/skills\\\"\",\n \"timeout\": 120,\n \"statusMessage\": \"Checking Nature Skills updates\"\n }\n ]\n }\n ]\n }\n}\n```\n\nIf `hooks.json` already contains other hooks, merge the `SessionStart` entry\ninstead of replacing the file. After enabling or changing the hook, run `\/hooks`\nin Codex to review and trust it. Codex currently runs command hooks synchronously,\nso this setup relies on the script's built-in 1-hour throttle, 60-second network\nguard, and offline-safe exit to avoid repeated network checks or blocking startup\nwhen an update cannot be fetched. Even when upstream is unchanged, the script\nverifies the destination and repairs detected drift.\n\nEach destination has a separate log at\n`~\/.local\/state\/nature-skills\/\/autoupdate.log`. Newly fetched\nskills normally take full effect in the next session.\n\n### 5.4 Other Agent Scenarios\n\nFor OpenClaw, OpenCode, and Hermes, see the dedicated [integration guide](docs\/open-source-agent-frameworks_EN.md).\n\nFor other agents, keep a stable repository clone and create a lightweight\nsubagent, slash command, or custom prompt wrapper that points to the real\n`skills\/*\/SKILL.md` files. Preserve `skills\/nature-shared\/`.\n\nFor manual or other-agent use:\n\n1. Copy complete skill directories into your prompt library or project.\n2. Preserve `SKILL.md`, `manifest.yaml`, `static\/`, `references\/`, scripts,\n assets, and required `skills\/nature-shared\/` files.\n3. If the target agent has its own format requirements, adjust the frontmatter\n and body structure.\n\n## 6. Skill Index\n\nThe current `skills\/` directory contains the following triggerable skills.\n`skills\/nature-shared\/` is shared content and is not counted in the skill index. Click a skill name or the \"Details\" link to open its dedicated documentation page.\n\n| Skill | Status | Purpose | Example Triggers | Details |\n|---|---|---|---|---|\n| [`nature-figure`](skills\/nature-figure\/README_EN.md) | Stable | Submission-grade Python or R scientific figure workflow for Nature \/ high-impact journals, with separately noticed third-party figures4papers references, original templates, and OpenRouter GPT Image 2 schematic drafts | \"Nature figure\", \"submission-grade figure\", \"publication plot\", \"scientific figure\", \"figures4papers\", \"paper schematic\", \"GPT Image 2\" | [Details](skills\/nature-figure\/README_EN.md) |\n| [`nature-polishing`](skills\/nature-polishing\/README_EN.md) | Stable | Polish, restructure, or translate academic prose into Nature-style English, with manuscript-wide terminology, unit, precision, and claim-drift checks | \"Nature style\", \"polishing\", \"academic writing\", \"English manuscript\" | [Details](skills\/nature-polishing\/README_EN.md) |\n| [`nature-writing`](skills\/nature-writing\/README_EN.md) | Draft | Draft Nature-style manuscript sections and rebuild a paper argument | \"Nature writing\", \"write an abstract\", \"write introduction\", \"manuscript draft\", \"paper writing\" | [Details](skills\/nature-writing\/README_EN.md) |\n| [`nature-reviewer`](skills\/nature-reviewer\/README_EN.md) | Draft | Simulate Nature-style reviewer assessment with three mutually blind reports, tiered Major\/Minor comments, and manuscript-internal consistency checks | \"Nature reviewer\", \"pre-submission review\", \"reviewer report\", \"reviewer-perspective assessment\" | [Details](skills\/nature-reviewer\/README_EN.md) |\n| [`nature-citation`](skills\/nature-citation\/README_EN.md) | Beta | Search support literature strictly within Nature \/ CNS families and export ENW, RIS, or Zotero RDF | \"Nature citation\", \"CNS citation\", \"segmented citation\", \"supporting references\", \"Zotero RDF\" | [Details](skills\/nature-citation\/README_EN.md) |\n| [`nature-data`](skills\/nature-data\/README_EN.md) | Draft | Prepare Data Availability statements, data repository plans, and FAIR checks | \"Data Availability\", \"data availability\", \"repository\", \"FAIR metadata\" | [Details](skills\/nature-data\/README_EN.md) |\n| [`nature-statistics`](skills\/nature-statistics\/README_EN.md) | Draft | Audit, revise, or draft statistical reporting, covering experimental units, replicates, p values, multiple comparisons, effect sizes, confidence intervals, figure statistics, and cross-section numeric consistency | \"Nature statistics\", \"statistical analysis\", \"p value\", \"sample size\", \"replicates\", \"multiple comparisons\", \"figure statistics\", \"statistics review\" | [Details](skills\/nature-statistics\/README_EN.md) |\n| [`nature-reader`](skills\/nature-reader\/README_EN.md) | Beta | Generate full-paper Markdown readers with source anchors, figure-text alignment, rendered equations, and Chinese-English side-by-side translation | \"nature reader\", \"full Markdown\", \"source-aligned text\", \"figure-text alignment\", \"rendered equations\", \"full translation\" | [Details](skills\/nature-reader\/README_EN.md) |\n| [`nature-paper-card`](skills\/nature-paper-card\/README_EN.md) | Beta | Deep-read one paper into a source-grounded Sections 01\u201316 Paper Card covering method logic, experiment-to-claim evidence, conclusion boundaries, critical analysis, and testable research ideas | \"nature paper card\", \"deep-read paper\", \"Paper Card\", \"evidence chain\", \"conclusion boundaries\" | [Details](skills\/nature-paper-card\/README_EN.md) |\n| [`nature-response`](skills\/nature-response\/README_EN.md) | Beta | Parse revision emails; create separate mutually blind reviewer responses, cover letters, red-marked manuscripts, LaTeX templates, and revision-package consistency checks | \"response to reviewers\", \"rebuttal letter\", \"cover letter\", \"major revision\", \"revision email\", \"reviewer-comment response\", \"LaTeX template\" | [Details](skills\/nature-response\/README_EN.md) |\n| [`nature-paper2ppt`](skills\/nature-paper2ppt\/README_EN.md) | Beta | Generate Chinese PPTX journal-club or paper-presentation decks from research papers | \"paper PPT\", \"journal club\", \"paper to slides\", \"paper presentation\" | [Details](skills\/nature-paper2ppt\/README_EN.md) |\n| [`nature-image2ppt`](skills\/nature-image2ppt\/README_EN.md) | Beta | Reconstruct slide images, scanned PDFs, and image-only PPTX files as object-level editable PowerPoint with rendered QA | \"image to editable PowerPoint\", \"reconstruct slide screenshot\", \"scanned PDF to PPTX\", \"editable PPT\" | [Details](skills\/nature-image2ppt\/README_EN.md) |\n| [`nature-paper-to-patent`](skills\/nature-paper-to-patent\/README_EN.md) | Beta | Generate evidence-constrained Chinese invention patent drafts and support patent-point mining, prior-art search, and iterative technical disclosure drafting | \"paper to patent\", \"Chinese patent\", \"paper-to-patent\", \"claims drafting\", \"technical disclosure\", \"patent points\" | [Details](skills\/nature-paper-to-patent\/README_EN.md) |\n| [`nature-ref-verifier`](skills\/nature-ref-verifier\/README_EN.md) | Stable | Cross-check references across multiple sources and flag author, title, year, volume, issue, and page inconsistencies | \"verify refs\", \"check references\", \"reference verification\", \"ref check\" | [Details](skills\/nature-ref-verifier\/README_EN.md) |\n| [`nature-academic-search`](skills\/nature-academic-search\/README_EN.md) | Beta | Multi-source literature search, citation verification, strict other-citation audits, article-level citation metric tables, influential citer profiling, and reference management | \"search papers\", \"find articles\", \"literature search\", \"literature lookup\", \"verify DOI\", \"strict other citation\", \"article citation table\", \"influential citer\" | [Details](skills\/nature-academic-search\/README_EN.md) |\n| [`nature-downloader`](skills\/nature-downloader\/README_EN.md) | Beta | Legally obtain academic full text\/PDFs through library access, Chrome login state, and open-access routes | \"download papers\", \"library paper download\", \"CARSI\", \"Web of Science\", \"PDF download\" | [Details](skills\/nature-downloader\/README_EN.md) |\n| [`nature-literature-pipeline`](skills\/nature-literature-pipeline\/README_EN.md) | Stable | Automated literature discovery pipeline: multi-source retrieval, six-axis scoring, deep-reading delivery, and local archiving | \"literature pipeline\", \"daily literature\", \"literature push\", \"daily literature push\", \"cron\" | [Details](skills\/nature-literature-pipeline\/README_EN.md) |\n| [`nature-experiment-log`](skills\/nature-experiment-log\/README_EN.md) | Draft | Standardize experiment images, voice, and text into Obsidian experiment logs with YAML frontmatter and archived source materials | \"experiment log\", \"record experiment\", \"Obsidian vault\", \"Feishu research group\" | [Details](skills\/nature-experiment-log\/README_EN.md) |\n| [`nature-proposal-writer`](skills\/nature-proposal-writer\/README_EN.md) | Beta | Proposal-first research writing state machine: establish evidence, argument, and section contracts before drafting or reviewing text | \"researchwrite\", \"proposal\", \"opening report\", \"research plan\", \"research writing QA\" | [Details](skills\/nature-proposal-writer\/README_EN.md) |\n\n## 7. Contribution and Development\n\n### 7.1 Shared Design Principles\n\n1. **Prefer primary sources**: rules should be grounded in published Nature\n content, official journal guidance, or explicit local sources rather than\n generic taste.\n2. **Make rules explicit**: explain the reason behind each rule instead of\n giving unsupported assertions.\n3. **Respect section and task context**: writing, figures, citations, and\n responses depend on the manuscript section and task.\n4. **Output first**: every skill should produce something directly usable, such\n as paste-ready text, `.svg`, `.pptx`, `.docx`, or concrete instructions.\n5. **Keep skills extensible**: each skill should be self-contained, and adding a\n new skill should not require modifying existing skills.\n\n### 7.2 Repository Layout\n\n```text\nskills\/\n\u251c\u2500\u2500 nature-shared\/ # keep this when skills reference ..\/nature-shared\n\u251c\u2500\u2500 nature-\/\n\u2502 \u251c\u2500\u2500 README.md\n\u2502 \u251c\u2500\u2500 README_EN.md\n\u2502 \u251c\u2500\u2500 SKILL.md\n\u2502 \u251c\u2500\u2500 manifest.yaml # present in router-style skills\n\u2502 \u251c\u2500\u2500 static\/ # present in router-style skills\n\u2502 \u2514\u2500\u2500 references\/...\n\u2514\u2500\u2500 nature-proposal-writer\/\n \u251c\u2500\u2500 README.md\n \u251c\u2500\u2500 README_EN.md\n \u251c\u2500\u2500 SKILL.md\n \u251c\u2500\u2500 scripts\/...\n \u251c\u2500\u2500 templates\/...\n \u2514\u2500\u2500 references\/...\n```\n\n### 7.3 Adding a Skill\n\nWhen adding a skill to this repository, follow this process.\n\n**1. Create the skill directory**\n\n```text\nskills\/nature-\/\n```\n\n**2. Add required files**\n\n| File | Required | Purpose |\n|---|---:|---|\n| `SKILL.md` | Yes | Frontmatter (`name`, `description`) plus rules and workflow loaded by the agent |\n| `README.md` | Yes | Human-facing Chinese documentation |\n| `README_EN.md` | Yes | English documentation paired with the Chinese details page |\n| `references\/*.md` | Recommended for complex skills | Modular rule files, API references, design theory, tutorials, chart types, and similar material |\n\n**3. Write the Chinese and English READMEs**\n\nEvery new skill must include both `README.md` and `README_EN.md`. The README is a human-facing entry page, not a duplicate of `SKILL.md` and not an installation manual. Its job is to help users decide within 30 seconds whether the skill fits their task, what they need to provide, what it will output, and where its boundaries are.\n\nBasic rules:\n\n- The Chinese and English README files must be one-to-one mirrors: same heading count, same order, and same information points. Do not let the English page become a separate template.\n- Start with the skill name, language switch link, and one positioning sentence.\n- Use the base structure below by default; add optional sections only when they are genuinely needed.\n- Do not repeat repository installation instructions, author bios, changelogs, development history, full file trees, or long internal implementation details inside a single skill README.\n- Put complex rules, API parameters, long tutorials, script explanations, and template indexes in `references\/`, `static\/`, `scripts\/`, or `SKILL.md`. Keep only navigational pointers in the README.\n- If the skill has visual assets, a small preview table is fine; do not turn the README into a large gallery or long technical manual.\n\nChinese README base structure:\n\n```markdown\n# `nature-` \u6280\u80fd\n\n[English](README_EN.md)\n\n\u4e00\u53e5\u8bdd\u8bf4\u660e\u8fd9\u4e2a\u6280\u80fd\u7684\u5b9a\u4f4d\u3001\u4e3b\u8981\u4efb\u52a1\u548c\u4f7f\u7528\u8fb9\u754c\u3002\n\n## \u9002\u5408\u7528\u5b83\u505a\u4ec0\u4e48\n## \u5178\u578b\u8bf7\u6c42\n## \u4f60\u9700\u8981\u63d0\u4f9b\n## \u4ea7\u51fa\n## \u8fb9\u754c\n## \u76f8\u5173\u6280\u80fd\n```\n\nThe English README must mirror it as:\n\n```markdown\n# `nature-` Skill\n\n[\u4e2d\u6587\u8bf4\u660e](README.md)\n\nOne sentence describing the skill's role, main task, and usage boundary.\n\n## What To Use It For\n## Typical Requests\n## What You Need To Provide\n## Outputs\n## Boundaries\n## Related Skills\n```\n\nOptional sections must be inserted in both languages and in the same order. Common optional sections:\n\n| Chinese Section | English Section | Use Case |\n|---|---|---|\n| `## \u5de5\u4f5c\u65b9\u5f0f` | `## Workflow` | Explain the core workflow or routing behavior |\n| `## \u8fd0\u884c\u548c\u4f9d\u8d56` | `## Runtime and Dependencies` | Scripts, MCP services, API keys, local config, or external dependencies |\n| `## \u793a\u4f8b\u9884\u89c8` | `## Example Preview` | A few figures, screenshots, or visual assets are worth showing |\n| `## \u5185\u7f6e\u53c2\u8003` | `## Built-In References` | Point to `references\/`, `assets\/`, or demos |\n| `## \u65b9\u6cd5\u6765\u6e90` | `## Method Sources` | Writing, review, or analysis rules come from specific sources |\n| `## \u4e09\u79cd\u6a21\u5f0f` | `## Three Modes` | The skill has clear compose\/revise\/hybrid-style modes |\n| `## \u4e0e ... \u7684\u5173\u7cfb` | `## Relationship With ...` | The skill is easy to confuse with another skill and needs a division-of-labor note |\n\nBefore submitting, run at least these README checks:\n\n```bash\npython scripts\/validate-readmes.py\ngit diff --check\nfor d in skills\/nature-*; do\n [ -f \"$d\/README.md\" ] && [ -f \"$d\/README_EN.md\" ] || continue\n rg -q '^\\[English\\]\\(README_EN\\.md\\)$' \"$d\/README.md\"\n rg -q '^\\[\u4e2d\u6587\u8bf4\u660e\\]\\(README\\.md\\)$' \"$d\/README_EN.md\"\n test \"$(rg -c '^## ' \"$d\/README.md\")\" = \"$(rg -c '^## ' \"$d\/README_EN.md\")\"\ndone\n```\n\n**4. Record a usage tutorial**\n\nWhen submitting a PR, please also record a short usage tutorial explaining what\nproblem the skill solves, how to trigger it, what inputs it needs, and what\noutputs it produces. Add the video, screencast link, or public tutorial URL to\nthe PR description.\n\n**5. Configure `SKILL.md` frontmatter**\n\n```yaml\n---\nname: nature-\ndescription: >-\n One sentence explaining what the skill does, when it should trigger, primary\n outputs, and core use cases.\n---\n```\n\n**6. Update the skill index**\n\nAfter adding a skill, update the [Skill Index](#6-skill-index) table:\n\n```markdown\n| [`nature-`](skills\/nature-\/README_EN.md) | Draft \/ Stable | One-sentence purpose | Trigger terms | [Details](skills\/nature-\/README_EN.md) |\n```\n\n**7. Set the status label**\n\n| Status | Meaning |\n|---|---|\n| `Draft` | Rules are defined but not yet tested on real cases |\n| `Beta` | Tested on examples, with possible edge-case issues |\n| `Stable` | Validated on real academic content and relatively stable |\n\n---\n\n## 8. Star History\n\n[![Star History Chart](assets\/star-history-20260816T024352Z.svg)](https:\/\/star-history.com\/#Yuan1z0825\/nature-skills&Date)"},"categories":[2462],"tags":[2072,2075,2078],"ai_series":[],"ai_topic":[2642],"ai_creator":[738],"ai_collection":[383],"ai_resource_type":[105],"ai_tool_tag":[679,567,680,681],"class_list":["post-4163","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-workflow_skill_pack","tag-codex-skills","tag-nature","tag-nature-skills","ai_topic-search-research","ai_creator-yuan1z0825","ai_collection-383","ai_resource_type-workflow_skill_pack","ai_tool_tag-claude-skills","ai_tool_tag-567","ai_tool_tag-680","ai_tool_tag-681"],"source_url":"","preserve_modified":false,"lang":"en","translations":{"en":4163,"zh-cn":387},"pll_sync_post":{},"_links":{"self":[{"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/posts\/4163","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/comments?post=4163"}],"version-history":[{"count":2,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/posts\/4163\/revisions"}],"predecessor-version":[{"id":4171,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/posts\/4163\/revisions\/4171"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/media\/3855"}],"wp:attachment":[{"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/media?parent=4163"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/categories?post=4163"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/tags?post=4163"},{"taxonomy":"ai_series","embeddable":true,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/ai_series?post=4163"},{"taxonomy":"ai_topic","embeddable":true,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/ai_topic?post=4163"},{"taxonomy":"ai_creator","embeddable":true,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/ai_creator?post=4163"},{"taxonomy":"ai_collection","embeddable":true,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/ai_collection?post=4163"},{"taxonomy":"ai_resource_type","embeddable":true,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/ai_resource_type?post=4163"},{"taxonomy":"ai_tool_tag","embeddable":true,"href":"https:\/\/ai.wuaishare.cn\/hub\/wp-json\/wp\/v2\/ai_tool_tag?post=4163"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}