Skill-hub

Windows C盘、项目垃圾、Docker/WSL、Git worktree、CodeGraph、Agent 会话或内存异常需要安全审计、清理和验证时使用。

Manage Agent Skills across Codex, Claude Code, local/global/project roots, Skill Hubs, GitHub, skills.sh, and managed plugin caches. Use when searching, installing, updating, upgrading, syncing, merging, auditing, deduplicating, cleaning, packaging, publishing, validating, resolving upstream sources, choosing runtime or scope, or working with skill-upgrade, check_upstreams.ps1, Claude Code plugins, skills-lock.json, or skill-upstreams.json. Do not use for ordinary software-package upgrades or non-skill file cleanup.

Create or refresh screenshot-led product hero images, localized interface captures, and image-led README sections for GitHub repositories. Use whenever a user asks to make a GitHub README show a GUI product visually, add software screenshots, design a README hero image, reproduce a screenshot-collage reference, or prepare repository-bound visual documentation. For GUI applications, default to the deterministic product-proof split hero with real software screenshots unless the user explicitly requests another visual direction.

Use when the user wants to design, redesign, shape, critique, audit, polish, clarify, distill, harden, optimize, adapt, animate, colorize, extract, or otherwise improve a frontend interface. Covers websites, landing pages, dashboards, product UI, app shells, components, forms, settings, onboarding, and empty states. Handles UX review, visual hierarchy, information architecture, cognitive load, accessibility, performance, responsive behavior, theming, anti-patterns, typography, fonts, spacing, layout, alignment, color, motion, micro-interactions, UX copy, error states, edge cases, i18n, and reusable design systems or tokens. Also use for bland designs that need to become bolder or more delightful, loud designs that should become quieter, live browser iteration on UI elements, or ambitious visual effects that should feel technically extraordinary. Not for backend-only or non-UI tasks.

Design engineering principles for making interfaces feel polished. Use when building UI components, reviewing frontend code, implementing animations, hover states, shadows, borders, typography, micro-interactions, enter/exit animations, or any visual detail work. Triggers on UI polish, design details, "make it feel better", "feels off", stagger animations, border radius, optical alignment, font smoothing, tabular numbers, image outlines, box shadows.

UI/UX design intelligence for web and mobile. Searchable local database with 84 styles, 192 color palettes, 74 font pairings, 192 product types, 98 UX guidelines, 104 icon entries, 16 GSAP motion presets, and 25 chart types across 22 stacks (React, Next.js, Vue, Nuxt, Svelte, Astro, SwiftUI, React Native, Flutter, Tailwind, shadcn/ui, Jetpack Compose, Angular, Laravel, JavaFX, WPF, WinUI, Avalonia, Uno Platform, UWP, Three.js, and HTML/CSS). Use when designing, building, or reviewing UI: pages, components, color schemes, typography, layout, accessibility, animation, or data visualization.

使用 Netlify CLI 完成安装、认证、站点关联、本地开发、环境变量管理与部署。

网页需要产品清晰度、信息层级、响应式或视觉系统美化、重设计或设计审计时使用。

可变体的 warm editorial 演示设计系统，用于 presentation、case sharing、项目汇报、战略叙事和产品发布。支持横向全屏 HTML deck 与程序化 PPTX，通过 style profile、语义 slide manifest、构图新颖度规划和反重复验证，在保持编辑设计 DNA 的同时避免固定配色、固定卡片和固定 11 页模板。用户要求 Pfizer/Anthropic/warm editorial/编辑设计风格、类似血液科那套，或需要高规格且非模板化的 HTML/PPT 演示时使用。

Use when work produces a deliverable, factual claim, dataset, code change, decision, publication, deployment, automation, or irreversible action whose failure would matter; when the user asks for a quality gate, acceptance criteria, QA, completeness, validation, audit, evidence, preflight, release readiness, or a definition of done; or before claiming completion on medium- or high-risk work. Automatically decide whether a formal gate is warranted, derive task-specific pass/fail criteria, gather evidence, and block unsupported completion. Make sure to use this skill even when the user does not say "quality gate" if consequential work needs acceptance criteria or completion evidence. Skip formal gating for trivial, reversible, low-impact requests.

Project-private fixture that must not be flattened into the local shared root.

Shared fixture from the repository.

Locally modified fixture.

Use when auditing, planning, measuring, or implementing GitHub repository discoverability work, including README/README_zh SEO review, Description or Topics proposals, release copy review, search ranking tracking, competitor or alternative term handling, Baidu/Bing/AI search channel assessment, site asset readiness, or portfolio rollout.

Use when running recurring GitHub repository SEO evaluation, comparing broad Chinese search visibility over time, producing metrics evidence, weekly summaries, manual Baidu/Bing/AI search tasks, or deciding whether an SEO experiment should trigger another optimization round.

Use when asked to study any writer or author, imitate article style, directly mimic a living or historical writer's prose style when requested, revise an imitation draft, or adapt Chinese articles for WeChat or Xiaohongshu.

Run evidence-disciplined Xiaohongshu social listening for foreign pharma employee pain points. Use when the user asks to collect and analyze notes/comments about pharma companies, aliases, roles, compliance, medical affairs, or AI topics, then produce a cited and scored sampled-insight report. Includes independent-evidence counting, MediaCrawler batch collection, and CAPTCHA/461 fallback to Agent Reach/OpenCLI.

Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.

剪映 (JianYing) AI自动化剪辑的高级封装 API (JyWrapper)，提供开箱即用的 Python 接口，支持录屏、素材导入、字幕生成、Web 动效合成及项目导出。全面适配 MacOS (Apple Silicon/Intel) 与 Windows，支持 v5.9+ (draft_info.json) 架构、工程自修复、智能配音字幕及录屏变焦。

Use this skill any time a .pptx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx file (even if the extracted content will be used elsewhere, like in an email or summary); editing, modifying, or updating existing presentations; combining or splitting slide files; working with templates, layouts, speaker notes, or comments. Trigger whenever the user mentions "deck," "slides," "presentation," or references a .pptx filename, regardless of what they plan to do with the content afterward. If a .pptx file needs to be opened, created, or touched, use this skill.

Use this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.

从 PubMed/PMC 论文取结构化全文并按章节切分为 RAG 友好的 chunk。当用户需要批量抽取论文全文、按 Methods/Results 等章节分段、提取表格上下文、构建文献向量库时，使用本 skill。支持 PMID（标题+摘要）和 PMCID（全文，开放获取子集）。token 级 RAG chunking 不跨章节，支持缓存与限流。

Use this skill whenever the user asks about clinical trial status, recruitment, trial phases (1-4), outcomes, trial registries, NCT IDs, or any query on ClinicalTrials.gov data. Real-time lookup of trial protocols, results, enrollment status, sponsors, and interventions via ClinicalTrials.gov API v2. Handles: 'What is the latest status of NCT04368728?', 'Find Phase 3 BTK inhibitor trials currently recruiting', 'Show primary outcomes for this trial', 'Which trials use Pembrolizumab?', 'Recruiting trials for Acute Myeloid Leukemia'. Real-time, single-query ≤1000 records. Delegates to: aact-bulk-trials (bulk >10k), pubmed-eutils (PubMed literature), medical-evidence-grading (GRADE rating), evidence-appendix-sync (reference formatting).

抓中国权威医学指南正文（CSCO / NCCN 中文版 / NMPA / CDE / NHC）并返回结构化"治疗推荐表"。当用户做中国市场的疾病调研、需要 I 级证据来源、或需要交叉验证治疗方案推荐是否符合 CSCO 指南时使用。v2 (2026-04-26)：Cite-or-Block 升级 — 加 sources_dir 参数完整存档原文 + locate_section_in_guideline 函数，与 citation-anchor-resolver 联动。Iron Law：locale=zh-CN+geo=CN 时本 skill 在 PubMed 召回之前必跑，所有治疗方案推荐必须命中指南 I 级 ≥ 1 个。Fetches Chinese authoritative clinical guidelines (CSCO, NCCN-zh, NMPA, CDE, NHC) and returns structured "treatment recommendation tables". MUST run before PubMed recall when locale=zh-CN+geo=CN; all proposed treatment recs must hit at least one guideline Level-I drug.

多源文献聚合检索：PubMed + PMC + 预印本（bioRxiv / medRxiv / ChemRxiv / Research Square）+ NCBI Bookshelf + 专利 一次调用，自动 PMID/PMCID/DOI 三级去重。用于医学 RAG 知识库广召回、全文 OA 链接发现、资助信息检索。返回完整元数据（OA 链接、citations、grants）。定位：广召回 · 多源 · RAG 友好。与 pubmed-eutils 互补 —— 本 skill 做"撒网"式多源聚合，后者做单源精准 + Clinical Queries。不做：单源精准检索（→ pubmed-eutils）、全文 XML 抽取（→ bioc-fulltext-fetch）、实体识别（→ pubtator-entity-search）、证据分级（→ medical-evidence-grading）。

Use when finalizing market research report appendix C (参考文献清单) before delivery; Use when verifying evidence/ folder PMID coverage matches the report's reference list; Use when generating 附录 C / Appendix C from a populated evidence/ directory; Use when reviewing a delivered report for PMID 漏列 / 幻觉 / 等级缺失. Trigger keywords: 市场调研报告交付前 / 附录 C 生成 / 参考文献清单 / evidence 同步 / PMID 双向校验 / 文献附录核对. Library-agnostic — works for hematology IFI, respiratory ICU 真菌, oncology, cardiology, ICU sepsis, any evidence-driven medical/pharmaceutical market research report where evidence/*.md and 报告附录 C must stay bidirectionally synchronized. Inherits L0 market-sizing-mece-foundation §evidence-driven pattern. v1.1 (2026-04-25):基于血液科 v2.5.2 复盘,新增 Iron Law 5(每条 PMID 必含 5 项元数据, 零裸数字)与 Iron Law 6(PMID 主题相关性核查,防关键词污染),配套 SOP Step 6/7/8 与失败模式 #9/#10。 v1.2 (2026-04-25):基于血液科 v2.5.3 复盘,新增 Iron Law 7(evidence/ 内部标注 vs PubMed 真实文献冲突时,信 PubMed 删 evidence/ 错误,禁止"保留 + 警告"妥协), 确立数据源信任优先级链(PubMed > 期刊官网 > NMPA > evidence 笔记 > LLM 推断), 配套 SOP Step 9(

Generate 50+ page professional market research reports with consulting-firm quality (McKinsey, BCG, Gartner style). Includes Porter's Five Forces, PESTLE, SWOT, TAM/SAM/SOM analysis, extensive visualizations, and LaTeX formatting. Use for competitive landscape analysis, market sizing, investment due diligence, and strategic planning. Integrates with research-lookup for data gathering and scientific-schematics for visualization generation.

L0 foundation for all medical market-sizing tasks. Use this skill whenever a task involves MECE department/age/region/indication slicing, evidence grading, drug-target three-state activity matrix, or cross-department patient-flow sizing. Trigger phrases: "市场多大", "market sizing", "用药占比", "竞品格局", "患者流". Disease-specific L1 skills (ifi-market-sizing-skill, etc.) must reference this first.

Medical evidence grading orchestrator. 当用户需要评估医学文献的证据强度、比较多篇论文的质量、查找某疾病的高等级证据(指南或 RCT)、或验证临床声明的证据支持度时,调用本 skill。整合 6 个原子 skill 的检索结果,按证据金字塔自动分级为 A/B/C/D。优先使用 BioMCP fast-path(~3-5s),无 BioMCP 时回退 slow-path(~15-25s)。

中国药品权威数据库查询 — NMPA(国家药监局) + CDE(药品审评中心) + DrugBank(英文名/ATC) + PubChem(化学结构) 多源融合,提供(通用名 ↔ 商品名 ↔ 英文名 ↔ 适应症 ↔ NMPA 批文号 ↔ 上市日期)精确映射。Make sure to use this skill whenever the user mentions 药品名 / 通用名 / 商品名 / NMPA / 国药准字 / 药监局 / CDE / drug brand name / generic name / drug registry / pharmaceutical lookup / drug verification / 药物核验 / 商品名对应 / drug cross-check / drug authority lookup. 在任何 medical / market-sizing / clinical / disease-research 报告生成阶段,**所有药品提及必须先调 lookup_drug() 验证**,严禁 LLM 凭训练记忆拼凑通用名↔商品名↔英文名对应关系。LLM 训练截止后药企会更换商品名、新药持续上市,记忆完全不可信。Iron Law:lookup 返回 None → 调用方必须 raise UnverifiedDrugError,不允许 fallback 到记忆。免费公开数据源,跨平台:Claude Code / Codex / Gemini CLI 三端均可 auto-trigger,实现仅依赖 httpx + beautifulsoup4 标准生态,不绑定任何厂商 SDK。

Search PubMed and retrieve structured biomedical literature with PMID lists, abstracts, MeSH terms, publication types, and PMC IDs for full-text access. Use this skill whenever the user mentions PubMed, PMID, MeSH terms, Clinical Queries, NCBI, E-utilities, or needs evidence for systematic reviews, RCT searches, drug-disease research, mechanism mapping, market research, or clinical evidence gathering. Make sure to use at the START of any medical evidence pipeline (diagnostic queries, prognosis, therapy effectiveness, observational studies, real-world evidence, clinical guidelines). Outputs structured records (title, abstract, journal, year, MeSH, publication type, PMID↔PMCID mapping)—does NOT perform semantic reranking (delegated to medical-evidence-grading), full-text extraction (delegated to bioc-fulltext-fetch), or entity annotation (delegated to pubtator-entity-search). Free official NCBI API covering 37M+ biomedical records. Cross-platform (Claude Code / Codex / Gemini CLI) with zero external SDK depende

NLM PubTator3 实体级关系挖掘原子 skill — 在 PubMed 文献里检索"疾病-药物-基因-化学品-突变-物种-细胞系"的标注与关系三元组。当用户问"BTK 与肺曲霉病的关联文献"、"BTK 抑制剂(MeSH D000077180)在哪些研究被讨论"、"标注这些 PMID 中提到的所有疾病/药物/基因实体"、"ibrutinib 的 MeSH/DrugBank ID"、"voriconazole 与 CYP2C19 药物-基因相互作用"、"BRAF V600E 突变 / rs113488022 检索"、"实体共现 / co-mention / 关系挖掘 / 实体规范化 / NER / annotation / entity normalization / disease-gene association / drug-gene interaction / chemical-disease relation" 时使用。无需 API key,跨平台稳定。**实体级 + 关系级**定位 — 不做综合文献检索(走 pubmed-eutils / europepmc-search)、不做全文段落抽取(走 bioc-fulltext-fetch)、不做证据等级排序(走 medical-evidence-grading)。

把多页 HTML 报告(主页 + N 子页 + PNG 资产)合并为单一独立 HTML / PDF / XLSX 三件交付物。 Use when finalizing a disease market sizing report and need shareable deliverables — single-file HTML for email/IM(零依赖,双击浏览器开),PDF for print/archive, XLSX for data export. 领域无关、跨疾病通用。 下游 of disease-market-sizing-html-template;上游 of email/IM/print 分发。 Use proactively whenever the user says "做交付物" / "生成 PDF" / "导 Excel" / "合并单文件 HTML"。

白板板书风格 PPT 生成。用户说「做PPT」「生成幻灯片」「白板风格演示」「板书」等时使用。支持 5 步标准工作流：确认风格→整理页码→检查依赖→逐页生成→合成下载。输出高保真白板照片，手写感 shufa 笔触，黑/蓝/红三色马克笔，适合教学、汇报、头脑风暴。每页需参考图和提示词前缀，封面必用专用前缀防止保留原文字。

WeChat article creator with Fu Sheng's writing style - conversational, story-driven, opinionated. Use this skill whenever the user mentions writing a WeChat article, drafting public account content, crafting a viral headline, creating article covers, writing in Fu Sheng's style, or any request involving 公众号, 微信文章, 爆款标题, or 傅盛风格 — even if they don't explicitly say "skill".

Use when generating Mermaid clinical decision trees that must include Leverage Point (LP) annotations at specific stage nodes (not just text-only in summary box). Trigger on tasks like "决策树标 LP 切入点", "在流程图节点上 标注杠杆点", "决策树嵌入 LP", or any flowchart that has accompanying LP strategy. Library-agnostic — works for hematology IFI, respiratory ICU, oncology, cardiology, any disease where decision tree + leverage points must be visually integrated. Inherits from L0 market-sizing-mece-foundation §10 LP decision diagram spec. LP annotations sourced from `pubtator-entity-search.entity_search()` (drug-disease relations).

Cite-or-Block 架构的基石原子 skill。把报告里的 citation 锚点 (例如 [guideline:CSCO-2024-NSCLC:§5.5.2] / [pmid:12345678:abstract]) 解析为源文件片段,并核对事实声明里的关键词是否在引用源原文里出现。 Foundation atomic skill for the Cite-or-Block architecture. Resolves citation anchors (e.g. [guideline:CSCO-2024-NSCLC:§5.5.2] / [pmid:12345678:abstract]) to raw source text fragments, and verifies whether claim keywords actually appear in the cited source. 使用场景 / Use when: 1. 报告生成后要做 fact-check (resolve_citation + verify_claim_against_source) 2. content-verification-layer 要扫全报告锚点 (parse_citations_in_text) 3. drug-citation-verifier / 任何 cross-check skill 需要"锚点 → 源原文" P0 守护:本 skill **绝不维护任何已知字典**。源里没有 = verified=False, 无例外、无 fallback、无"已知答案兜底"。 P0 guard: this skill **never maintains any built-in dictionary**. Not in source = verified=False. No fallback, no built-in answers.

Phase 4 Turn 1 of disease-market-sizing-orchestration. Generates report HTML with mandatory citation anchors per fact claim. Reads contract + staging + sources/. Every drug/treatment/stat/recommendation MUST carry [pmid:.../guideline:.../nct:...] anchor that resolves to an actual file in sources/. NEVER hardcode drug names or numerical claims. Use AFTER evidence recall (phase 3), BEFORE content-verification-layer (Turn 2).

Phase 4 Turn 2 of disease-market-sizing-orchestration (A6' upgrade). Audits compose output via Cite-or-Block strict — every anchor must reverse-resolve to an actual file in sources/, every fact claim must have an anchor. Returns verdict (ok/blocked) + violations list. Use AFTER cite-bound-content-generator (Turn 1), BEFORE report-bundle-builder (phase 5).

Use FIRST when user requests any disease market sizing report (e.g. "做 X 市场调研" / "Run market sizing for X"). Elicits 4-6 startup questions (evidence window, geography, international comparison, report depth, user role) and writes .cache/<slug>/contract.json. ALWAYS invoke before disease-market-sizing-orchestration. Validates against schemas/contract.schema.json.