yuan1z0825/nature-skills
11 skills64.6k installs367k starsGitHub
Install
npx skills add https://github.com/yuan1z0825/nature-skillsSkills in this repo
1Nature FigureThe nature-figure skill >-. Nature Figure Making - Router This skill is split into two layers: - A static layer under static/ that holds versioned, reusable content fragments (the figure contract and default stance, plus a per-backend quick-start for Python and R). - A dynamic layer (this file plus manifest.yaml ) that detects the plotting backend and loads only the fragment needed for the current job. The large design, API, pattern, and QA material lives in on-demand references. Do not try to apply the figure logic from memory or from this router. Always load fragments from disk as described below. Routing protocol Follow these five steps every time the skill is invoked. Load the manifest and the core layer Read manifest.yaml. It declares the backend axis, the allowed values, and the file paths each value maps to. Use it when developers need nature figure tasks covered in the upstream ECC or community skill documentation with concrete commands, prerequisites, and workflow steps rather than guessing APIs or conventions.7.9kinstalls2Nature PolishingNature-polishing restructures and polishes academic prose using writing-strategy principles, curated Nature article patterns, and Academic Phrasebank phrase families. Developers use it to improve research papers, methods papers, and abstracts before submission by diagnosing argument structure first (paper type, section logic, claim-evidence alignment) and then applying sentence-level refinements. Key workflows include identifying paper type, applying hourglass structure to Introductions and Discussions, ensuring Results stay descriptive and Discussion interpretive, checking title credibility, and enforcing reproducibility in Methods sections. The skill prioritizes logical architecture over cosmetic polish and explicitly protects core scientific arguments from AI invention while allowing human-controlled refinement of grammar, clarity, and tone. Diagnoses paper type (research, methods, hypothesis, algorithmic) before editing to apply correct narrative logic. Applies hourglass structure: Introduction narrows to specific gap; Discussion/Conclusion widens back to literature and implications. Separates Results (what happened) from Discussion (how it fits, when it may fail) with distinc.7.2kinstalls3Nature Paper2pptThe nature-paper2ppt skill Build a complete but efficient Nature-style Chinese PPTX presentation from a scientific paper, preprint, PDF, article text, abstract, figure legends, or reading notes. Use this skill whenever the user asks to make slides/PPT/PPTX for journal club, group meeting, paper sharing, thesis seminar, lab meeting, department report, or academic presentation from a research paper, not only medical papers. It identifies the paper type and argument, selects only the figures needed for the story, writes Chinese slide content and speaker notes, creates the actual .pptx deck, and runs an explicit self-review/corrective revision loop focused on figure quality, text overflow prevention, and non-template visual design before delivery. Purpose Transform a scientific paper or paper-derived notes into a complete Chinese, figure-integrated PPTX presentation package with a Nature-style reporting logic. The skill must not stop at an outline or script. The expected end product is a real .pptx deck. Keep supporting files minimal unless the user asks for more traceability. Use this skill for papers across scientific fields, including: -6.6kinstalls4Nature CitationThe nature-citation skill adds strict Nature and CNS journal-family citations to manuscript text through a router that loads static core fragments and on-demand references. Developers and researchers use it when they need segmented text-to-reference correspondence, Nature-series or CNS support for statements, or reference-manager exports while writing academic manuscripts. The router always loads manifest.yaml, core principles, Chinese-user operating mode, and the seven-step workflow before searching. Scope is limited to accepted flagship and subjournal titles from Nature Portfolio, the AAAS Science family, and Cell Press, filtered by publication time range. The workflow segments passages, parses claims, searches conservatively, evaluates support, exports one reference-manager file by default, and reports with an HTML browser path first. It prefers scripts/nature_citation.py when internet access is available and forbids citing papers based on title similarity alone or inventing bibliographic fields. Long articles switch to a batched strategy documented in references/script-usage.md when more than about ten segments are present.6.4kinstalls5Nature Datanature-data prepares Data Availability statements and supporting metadata for Nature-family and general academic submissions through a static and dynamic two-layer router. Every invocation loads manifest.yaml plus always_load fragments: stance, chinese-mode, and an eight-step workflow from static/core without applying logic from memory. Runtime variation covers journal or article type conflicts, per-dataset access routes such as public repository, controlled access, within paper, reused public, third-party restricted, justified request, or not applicable, and Chinese-to-English drafting with a verification block when users write in Chinese. The workflow inventories supporting datasets, classifies each access route, chooses repository and identifier strategy before drafting, maps datasets to locations explicitly, adds formal dataset citations, runs a FAIR metadata audit, and returns ready-to-paste text plus unresolved fields. Deep references under references/ load on demand for policy principles, repository identifiers, statement patterns, FAIR checklists, and Chinese author alignment. The skill forbids inventing DOIs, accession numbers, repository names, licences, embargo dates, o.6.3kinstalls6Nature Responsenature-response is a router skill for drafting, auditing, or revising point-by-point reviewer response letters for Nature-family manuscript revisions and broader peer-review replies. It splits work into a versioned static layer under static/ and a dynamic router that loads fragments from disk instead of improvising from memory. Every invocation starts by reading manifest.yaml and always_load files stance.md and workflow.md, then identifies task mode such as draft, audit, revise, triage-only, or appeal-like, decision type including minor or major revision, and user language for optional Chinese alignment blocks. The ten-step workflow extracts editor instructions and reviewer comments with stable IDs, classifies each item, builds a strategy summary, drafts point-by-point responses from preserved comments, maps claimed changes to manuscript locations or explicit placeholders, flags missing author input as AUTHOR_INPUT_NEEDED, and returns a package with readiness state. Deep references under references/ load on demand for comment taxonomy, tone, difficult cases, and QA. The skill never invents experiments, citations, line numbers, figure panels, or manuscript changes, and routes appea.5.9kinstalls7Nature WritingThe nature-writing skill drafts and restructures Nature-style scientific manuscript sections from author-provided claims, results, figures, notes, or Chinese drafts. It uses a router plus manifest.yaml to load only relevant static fragments for paper type, section, language, and journal axes instead of applying drafting logic from memory. The five-step protocol loads manifest and core stance, detects axis values, reads matching fragment files, drafts using priority order from core intake through journal and language rules, and opens deep references on demand. The eight-step workflow in core/workflow.md requires planning steps before prose, surfacing missing evidence as Assumptions or missing inputs placeholders rather than inventing content. Triggers include abstract, introduction, methods, discussion, conclusion, and Chinese academic writing phrases. It pairs structurally with nature-polishing for shared static layer evolution across both skills.5.6kinstalls8Nature Academic SearchNature Academic Search is a router skill for coordinated literature research beyond a single MCP call. A static layer under static/ holds MCP tool inventory, source routing, and operational rules, while this router plus manifest.yaml selects one of five workflows: multi-source search, citation verification, MeSH strategy, citation file management, or reference management. Each invocation loads manifest.yaml, always reads core/tools.md and core/routing-and-ops.md, detects the workflow axis, then loads only the matching references/workflows fragment. Execution follows the T1 to T2 to T3 source fallback chain, reports tool failures, broadens terms on empty results, and falls back to manual metadata generation if scripts fail twice. Triggers include literature review, citation verification, .nbib or .ris or .bib conversion, BibTeX work, and Chinese phrases like 文献检索 or 引文核对. The router stays short on purpose; scope changes belong in fragments and references, not the top-level file.5.4kinstalls9Nature ReaderNature Reader builds full-paper Chinese-English side-by-side Markdown readers for journal and conference papers from PDF, DOI, arXiv, publisher HTML, or pasted text. A static layer under static/ holds versioned fragments for core principles, workflow, output contract, and per-source extraction guidance, while this router and manifest.yaml load only what each job needs. Routing detects source_format as pdf-text, scanned-pdf, html, doi-arxiv, or pasted-text, states the choice to the user, then loads matching fragments without reading every file. Applied priority is core principles first, then source-format extraction, six-step source-map-first workflow, and output contract verification. The skill must not degrade to summary-only output unless explicitly requested. It produces paper.md, source_map.json, translation_notes.md, and a Terminology Ledger glossary. References for figure extraction, output schema, and grounding rules load on demand. The split mirrors nature-writing and nature-polishing so shared content lives in _shared/.5.3kinstalls10Nature Reviewernature-reviewer is an agent skill from yuan1z0825/nature-skills that apply nature-reviewer agent skill workflows from documented skill.md guidance. # Nature Reviewer Assessment Skill Use this skill to simulate a `Nature`-style reviewer assessment package from the referee side. This skill is for reviewer-style manuscript evaluation, not for drafting the authors' response. If the user wants rebuttal writing, route to `nature-response`. ## Default stance - Ground the review only in the local Developers invoke nature-reviewer during build/integrations work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments. Category AI & Agent Building with ai vertical focus supports repeatable agent-guided delivery.4.5kinstalls11Nature Paper To Patentnature-paper-to-patent is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted development.3.6kinstalls