写作
内容写作与编辑
deft-directive-glossary
deftai
Extract a DDD-style ubiquitous language glossary from the current conversation, flagging ambiguities and proposing canonical terms. Saves to UBIQUITOUS_LANGUAGE.md. Use when defining domain terms, building a glossary, hardening terminology, or when "domain model" or "DDD" is mentioned.
app-store-changelog
Dimillian
Create user-facing App Store release notes by collecting and summarizing all user-impacting changes since the last git tag (or a specified ref). Use when asked to generate a comprehensive release changelog, App Store "What's New" text, or release notes based on git history or tags.
room-edit
rolandcanyon-cmd
Edit a previously-catalogued room — rename, add/remove aliases, move devices between rooms, fix typos, mark devices inoperable, remove dead entries. Same review-before-commit pattern as /room-walk.
bazi-pro
Minervaowl7
专业八字命理解读 Skill v5.2 — 可审计、可交互、可视化的命理分析引擎（Replayable Evidence Engine），配合 Bazi MCP 排盘数据使用。 古籍检索：6部经典（子平真诠/滴天髓/渊海子平/神峰通考/三命通会/穷通宝鉴） 共2964条原文，BM25+jieba实时匹配。总计29.8万中文字。 排盘计算由 Bazi MCP（TypeScript 硬编码）完成，本 Skill 专注解读环节。 核心能力：喜用神判断（格局主导+扶抑修正+调候调节+病药验证 四层架构，带量化评分）、 格局分析（月令透干定格+暗格+建禄月劫+化气格+从格 全覆盖，含从强格/假从强格）、 大运流年喜忌引动（含大运上限原则）、空亡/神煞/胎元命宫整合解读、 六维度分述（性格/事业/财运/感情/健康/近运）+ 历史事件闭环校准。 多流派分析（v5.2 典籍对齐版）： - 盲派（段建业体系）：宾主/体用/6种做功（含墓用复合）/贼神捕神/五党成势/功神废神 - 新派（李涵辰体系）：百神论(动态查找)/空亡论(含出空)/反断论(同宗对)/格局分类(扶抑从强从弱) v5.2: 盲派新派典籍对齐——盲派7项修正(体用定义/墓用复合/贼神捕神/五党成势) + 新派6项修正(反断同宗对/百神动态/身旺身弱综合/格局分类/出空机制) + Golden Cases 507个。 v5.1: 古籍校准——调候为急例外(格局不可绝对压制调候)、运可成格变格(大运非硬件上限)、会方检测(寅卯辰/巳午未/申酉戌/亥子丑)、假从强格(印比≥80%未达极旺)、财破印通关检查(官杀通关化解)。 v4.5-4.8: 动态SVG命盘+ 命运河流时间轴+ 推理图谱DAG+ FastAPI服务+ Hybrid Search INT8量化+ ViewModel统一化。 v4.2: Replayable Evidence Engine——Golden Cases 边界回归测试(83例)+ GitHub Actions CI + Evidence Pipeline(分析→证据JSON→仪表盘)+ Web Demo 首页(dist/index.html)+ bazi_pro 引擎包重构。 v4.1: 版本号统一(4.1.0)+ version 单一源 + smoke tests(7项)+ bazi doctor 环境诊断 + README 引擎化 + dashboard v3.0(Evidence Inspector+刑冲合害图谱)+ Hybrid Search breakdown(matched_terms+why)。 v4.0: 引擎化升级——Evidence Object 证据链结构化输出、BM25 索引缓存(冷启动1.8s→热查询0.5s)、批量检索 API、Hybrid Search(BM25+向量+权威权重)、仪表盘 v2.0(五行着色+评分色环+系统主题)、examples/样例文件、pyproject.toml+requirements.txt 依赖管理。 v3.5: 层3暗格比劫拦截——月令本气为比肩/劫财时禁止生造"暗劫财格"等非经典术语，强制使用《子平真诠》"建禄月劫"框架（另寻财官煞食透干取用）。v3.4.1: 报告生成器跨平台字体修复（Linux/WSL 环境下 emoji→文本降级、中文字体栈修正 Noto Serif CJK SC、等宽字体 DejaVu Sans Mono、@font-face emoji 声明、PDF 渲染错误日志）。 v3.3: 算析分离·线性执行流（废除暂定/回补）、两轮对话协议（详细版Turn1数据校准+Turn2全盘解读）、计算外包原则（LLM 只能解释 full_analysis 结果，复杂计算外包MCP）、燥湿寒暖偏枯预检、病药突破大运上限、动态负面清单比例。 v3.2: 负面清单强制规则（Step8每维≥1风险）、评分基准锚防通胀、连续性忌神大运专项分析、回补完整性校验、jieba缺失硬阻断。 v3.1.2: 修复0.0阈值不一致（增设60-74%灰区→建议双通道）；0.2措辞统一（强制/建议分级）；Step4 0-A数据依赖+Step5格局复核闭环；Step6大运缺失降级；Step5.1/8.1补充交叉引用。 v3.1.1: 标准版输出增强（十神表+神煞表+六层筛查+ASCII图）、每步≥1古籍引用、每维度≥3依据。 v3.1: 第〇步双通道检索+反事实裁决；从强格/假从强格纳入层0；层0筛查先全局后局部。 v2.3: 第〇步古籍检索（BM25+jieba，2964条6经典）。v2.2: 三层用神架构；大运上限；反泛化红线。v2.1: 十神根气虚实检查；病药法虚实预检。 Use this skill when user provides Bazi MCP output (JSON) or asks for 八字解读、 命理分析、喜用神、格局、大运 after getting Bazi data from MCP. Triggers: "解读八字" "分析命盘" "看喜用神" "什么格局" "大运分析" "bazi analysis" "多角度" "深度分析" "详细"。
jay-brain-and-skill-index
joecastelino
Operate, diagnose, and repair Jay's GBrain (the single Jay-only knowledge brain) plus the self-maintaining skill index. Use when the nightly brain refresh fails or looks frozen, when gbrain is "command not found", when GBrain search returns nothing/stale, when session knowledge needs to be captured into the brain so it's retrievable later, or when the skill index needs descriptions/metadata/usage-counts. Covers the bun-PATH break, the multi-brain split, the embedding-key gotcha, and the capture-session / skill-usage tracking scripts.
ai-daily-frontier
SeeyeeTech
Generate a flagship-grade daily "all-web AI frontier" digest benchmarked against MorningAI — covering papers, model & product launches, benchmarks, funding/acquisitions, and engineering practice across labs, OSS, and media. Fan-out collection (targeted WebFetch + bilingual WebSearch) feeds two-stage scoring and adversarial fact verification, then assembles a tiered Chinese digest. Use when the user wants the all-web AI news daily ("AI 前沿日报", "AI 资讯日报", "daily AI news", "frontier digest"). Sibling skill ai-daily-ecosystem covers this project's open-source ecosystem signals.
ai-daily-ecosystem
SeeyeeTech
Generate a daily "open-source AI ecosystem" digest centered on this project's own unique data — today's GitHub Trending, what whitelisted AI leaders are starring, cross-window heating repos — with deep links back to 380+ existing analyses. Use when the user wants the open-source AI daily ("开源 AI 日报", "生成今日开源日报", "ecosystem digest"). Sibling skill ai-daily-frontier covers general all-web AI news.
scientific-writing
trimcrae
Write and revise this repository's submission texts so a working scientist can read them once and follow them. Load BEFORE drafting or revising any manuscript, SI, cover letter or abstract that will go to a preprint server, a journal or aiXiv, and before applying a hardening round's findings to prose. Covers the four moves that carry almost all the gain (topic and stress positions, characters as subjects, actions as verbs, one sentence one idea), the revision pass that finds the damage, and the two failure modes that make a "clearer" paper worse — a hedge quietly dropped to shorten a sentence, and a readability score chased as a target. Metrics are a SCREEN that says where to look, never a measure of whether the prose is good; lint_readability.py reports, this file decides.
sd-research-repo
platypeeps
Use when writing, rendering, reviewing, or publishing a document in a research repo that follows the shared research-repo standard — the numbered layout, provenance blocks, Status sections, and the outward mirrors.
sd-author
platypeeps
Use when the user wants to develop an original evidence-backed technical article through a one-question interview, approved editorial brief, staged drafting, review, and publication handoff.
sd-profile
platypeeps
Use when the user wants to create, inspect, correct, review, import, export, or forget a consent-driven personal operating profile with traceable assertions.
sd-paper
platypeeps
Use when the user wants to develop a credible research paper through question refinement, an approved research brief, explicit literature and methodology protocols, traceable evidence, reproducibility, and venue-aware review.
edge-esmeralda-2026
aromeoes
Connect to Edge Esmeralda 2026 data — event schedule, attendee directory, wiki, newsletters, and organization info.
hmasd-peer-collaboration
CartmanFatass
Coordination between the Claude DM session and the Codex Root as peer researchers - channels, message shape, inbox convention, shared-working-tree hygiene and disagreement routing under constitution section 2. A method, not a role or an authority.
reflect
Mike-Byrne-AI
Cross-artifact gap surfacing after a substantial implementation pass — looks across the files just changed for structural drift, missing connections, and inconsistencies the file-by-file review pass cannot see. Use when the user asks to reflect, after generating 3+ interconnected files, when the user senses something was missed, or before /handoff on significant sessions. Distinct from /review (single-file correctness lens) and /adversarial (failure-mode discovery). For non-trivial sessions, delegates the cross-artifact review to the code-reviewer agent for orthogonal context.
mao-zedong-perspective
arraycervicalartery576
毛泽东思维框架：以《毛泽东选集》五卷为核心，提炼毛泽东分析问题、制定战略、组织行动的认知操作系统。 核心来源：《矛盾论》《实践论》《论持久战》《中国社会各阶级的分析》《星星之火可以燎原》《论联合政府》《关于正确处理人民内部矛盾的问题》等。 核心模型：7个。决策启发式：10条。 触发词：「毛泽东」「毛选」「教员」「用毛泽东的方式分析」「从毛选的角度」「教员怎么看」 局限：本Skill聚焦于毛泽东的分析方法论和战略思维框架，适用于战略分析、组织管理、竞争策略、问题诊断等场景。不涉及具体政治立场的评判。 素材来源：《毛泽东选集》一至五卷、公开演讲与书信、诗词作品。
scriptwriter
simplebusiness26
Write or refine short animated episode scripts that are visual, funny, production-feasible, age-appropriate, and faithful to the show bible.
anlin-writing
github-yjc
"Generate, review, or evaluate Anlin/日寄-style anonymous blind-evaluation prose against the Anlin corpus. Use ONLY when the user explicitly mentions anlin-writing, anlinwriting, Anlin, 日寄, Anlin-style, 像Anlin那样写, 模拟日寄, 盲评, or asks for Anlin corpus evaluation. In any workspace containing .anlin-clean-eval-mode, check that marker before drafting and use clean_run_checker.py, not the normal checker."
ib-write-tools
wavespestcontrolfl
Use when adding or changing an Intelligence Bar tool that writes anything (create, update, send, schedule) — every write goes through the confirm trust boundary and must register in write-gates.js.
hmasd-peer-collaboration
CartmanFatass
Coordination between the Claude DM session and the Codex Root as peer researchers - channels, message shape, inbox convention, shared-working-tree hygiene and disagreement routing under constitution section 2. A method, not a role or an authority.
blueprint-authoring
Mike-Byrne-AI
Task-pack authoring conventions. Use when composing or refining a TP-N task pack, scoping a pack to a defect class, drafting code-fix instructions, structuring scope-in vs scope-out, or naming tasks within a pack — also when the user mentions task pack format, blueprint structure, RUNBOOK, or pack execution. Encodes Espalier-Harness's blueprint-authoring discipline.
write-docs
browseros-ai
Write BrowserOS feature documentation. Use when the user wants to create or update documentation for a BrowserOS feature. This skill explores the codebase to understand features and writes concise Mintlify MDX docs.
visual-explainer
nicobailon
Generate beautiful, self-contained HTML pages that visually explain systems, code changes, plans, and data. Use when the user asks for a diagram, architecture overview, diff review, plan review, project recap, comparison table, or any visual explanation of technical concepts. Also use proactively when you are about to render a complex ASCII table (4+ rows or 3+ columns) — present it as a styled HTML page instead.
create-scratch
documenso
Create a new scratch file in .agents/scratches/ with a unique three-word ID, frontmatter, and formatted title