bahayonghang

Create or review publication-ready academic figures that meet journal submission specs (IEEE / Elsevier / Nature, plus Springer and chinese-thesis presets) using matplotlib(+seaborn) or plotly; in industrytslib projects it drives that library's built-in visualization system. Use for 论文配图, 期刊图, 科研绘图, 中文学位论文配图, "publication-ready figure". Do NOT use to reproduce a specific paper's figure or style — that is paper-plot.

Audit and improve Codex AGENTS.md guidance files and companion code_map.md navigation maps. Use when the user asks to check, audit, optimize, or fix AGENTS.md; mentions nested AGENTS.md conflicts, code maps, stale commands, or scoped instructions; or says 优化 AGENTS.md, 审计 AGENTS.md, 更新 Codex 项目指导, 生成 code_map.md.

Archive root-level planning files (`task_plan.md`, `findings.md`, and `progress.md`) into a timestamped `.plannings/` directory for the current feature. Use when Codex should run `$archive-planning [feature-name]`, close out an active plan, or preserve completed planning context.

Write, debug, and validate ast-grep structural code search rules. Use for syntax-aware code search, AST pattern matching, structural refactor discovery, or searches plain text tools like rg can miss — functions with particular descendants, calls inside specific contexts, decorators, or other Tree-sitter-backed structures.

Audit and improve Claude Code CLAUDE.md guidance files, .claude/rules/ path-scoped rules, and companion code_map.md navigation maps. Use when the user asks to check, audit, optimize, or restructure CLAUDE.md, mentions nested CLAUDE.md or code_map.md, or says 优化 CLAUDE.md, 审计 CLAUDE.md, 生成 code_map (Claude). Not for trivial single-line edits the user has already fully specified.

Analyze a repository and current Codex environment, then recommend Codex CLI/App, AGENTS.md, skills, subagents, plugins, MCP servers, config/hooks, and OMX workflow improvements without modifying files. Use when the user asks to optimize Codex workflows or setup, recommend Codex automation, configure MCP/plugins/subagents, or says 优化 Codex 流程, Codex 工作流推荐, 给 Codex 配 MCP.

Turn vague or complex agent tasks into strong `/goal` commands for Claude Code and Codex — outcome, verification, constraints, boundaries, iteration policy, and stop conditions. Use for Codex/Claude Code goal instructions, Goal 指令, 目标指令, `/goal` prompts, 中文 Goal 模板, plan-to-goal interviews, or bounded agent work definitions.

Use when the user needs text or regex content search with ripgrep: composing rg commands, choosing flags, glob/type filtering, multiline or PCRE2 searches, pipeline output, grep-to-rg migration, or diagnosing why rg missed a file (gitignore, hidden, binary defaults). Not for syntax-aware structural queries (ast-grep) or semantic renames/references (language tooling).

Plan-first brainstorming workflow that turns an idea into an approved implementation plan before coding. Use when the user wants to brainstorm, design, scope, stress-test, or plan a feature/spec. Spark adapts to the active planning surface: in Codex native Plan mode it stays read-only and returns a chat-only final plan; in Claude Code Plan mode it uses Claude's plan approval/exit flow and writes files only after approval exits Plan mode; in writable/default mode it saves Markdown plans under .plannings/YYYY-MM-DD-feature-slug.md. HTML or visual artifacts are explicit and writable-mode-only.

Write, debug, and validate ast-grep structural code search rules. Use this skill when the user needs syntax-aware code search, AST pattern matching, structural refactor discovery, language-construct queries, or searches that plain text tools like rg can miss, such as finding functions with particular descendants, calls inside specific contexts, missing error handling, React hook shapes, decorators, or other Tree-sitter-backed code structures.

Audit and improve Codex AGENTS.md guidance files and companion code_map.md navigation maps in repositories. Use when the user asks to check, audit, update, optimize, or fix AGENTS.md files; asks for Codex project guidance maintenance; mentions nested AGENTS.md conflicts, code maps, stale commands, scoped instructions, sandbox or approval boundaries; or says "优化 AGENTS.md", "审计 AGENTS.md", "检查 nested AGENTS.md", "更新 Codex 项目指导", or "生成 code_map.md".

Archive root-level planning files (`task_plan.md`, `findings.md`, and `progress.md`) into a timestamped `.plannings/` directory for the current feature. Use when Codex should run `$archive-planning [feature-name]`, close out an active plan, preserve completed planning context, or replace the deprecated Codex prompt workflow with a deterministic skill script.

Audit and improve Claude Code CLAUDE.md guidance files, .claude/rules/ path-scoped rules, and companion code_map.md navigation maps. Use when the user asks to check, audit, update, optimize, or fix CLAUDE.md files; mentions nested CLAUDE.md, additive loading, ancestor/descendant loading, path-scoped rules, @import chains, CLAUDE.local.md, claudeMdExcludes, or code_map.md; or says "优化 CLAUDE.md", "审计 CLAUDE.md", "检查嵌套 CLAUDE.md", "Claude Code 项目指导", "生成 code_map (Claude)". Make sure to use this skill whenever the user wants to manage Claude Code memory files at any layer, even if they only mention CLAUDE.md without explicit audit wording.

Analyze a repository and current Codex environment, then recommend Codex CLI, Codex App, AGENTS.md, skills, native subagents, plugins, MCP servers, config/hooks, and optional OMX workflow improvements without modifying files. Use when the user asks to optimize Codex workflows, improve Codex setup, recommend Codex automation, configure MCP/plugins/subagents, analyze AGENTS.md scope needs, or says "优化 Codex 流程", "Codex 工作流推荐", "给 Codex 配 MCP", "配置 Codex plugins", or "配置 Codex subagents".

Create self-contained HTML artifacts, single-file by default and split into offline per-page bundles when oversized, for complex, visual, comparison-heavy, reviewable, or shareable work outputs. Use this skill when the user asks for an HTML artifact, browser-viewable report, implementation plan, PR/code review report, architecture explainer, design comparison, research briefing, incident report, status dashboard, temporary structured-data editor, mini deck, or design-system specimen. Prefer this skill for long Markdown-like outputs that benefit from navigation, cards, diagrams, filters, copy/export buttons, or annotated diffs. Do not use it for short answers, simple commands, commit messages, tiny patch summaries, or production UI implementation unless the user explicitly asks for a review artifact.

Modify the Beautiful Mermaid live editor itself rather than writing ordinary Mermaid diagrams. Use when the task mentions the Beautiful Mermaid repo, `editor.ts`, generated `editor.html`, config panel/options, themes or dark mode, zoom, PNG/SVG export, clipboard behavior, sample presets, or renderer wiring for the editor. Do not use for generic Mermaid syntax help or normal Markdown Mermaid authoring.

Incrementally compile tracked raw captures into reviewable draft knowledge. Use this skill whenever the user says "compile wiki", "compile kb", "sync drafts", "digest these captures", "turn my clips into drafts", "编译wiki", "更新草稿层", "同步草稿", or wants tracked material under `raw/human/**`, `raw/agents/{role}/**`, bootstrap `raw/*.md`, `raw/**/assets/*`, or `raw/**/data/*` turned into reviewable summaries, topics, concepts, entities, and draft indices. This is also the right skill when Chinese-LLM-Wiki wording such as `来源页`, `主题页`, `实体页`, `综合页草稿`, `整理成来源页`, `补主题页`, or `补实体页草稿` clearly means “build draft pages from raw evidence first.” Do not treat `raw/**/papers/*.pdf` as a normal compile trigger: those paper ingests still belong to `paper-workbench`, and `kb-compile` should only surface or defer them.

Register and normalize raw source intake for an Obsidian knowledge base. Use this skill whenever the user says "kb ingest", "ingest sources", "refresh manifest", "register new raw files", "scan raw library", "同步素材清单", "登记新来源", "更新 manifest", or wants `raw/**` captures recorded into `raw/_manifest.yaml` before compile runs. Treat browser-side collection as an upstream companion lane: if the source is still on the web and not in `raw/` yet, use `web-access` or Web Clipper first, then come here. Do not use it for draft generation, review, query, or maintenance once the source registry is already current.

Initialize, migrate, or repair a review-gated Obsidian knowledge base. Use this skill whenever the user says "kb init", "initialize knowledge base", "repair vault", "migrate an old layout", "add draft/live review gate", "set up briefings", "LLM Wiki", "Karpathy wiki", "Obsidian IDE", "knowledge compiler", "personal knowledge base", "second brain", "初始化知识库", "迁移知识库", "修复知识库结构", "中文优先 wiki", "raw/wiki/output 脚手架", "把这个 vault 搭成中文优先 wiki", or wants a fresh vault or legacy-layout vault brought onto the review-gated contract. Do not use it for normal compile, review, query, or maintenance work once the support layer is already healthy.

Query, search, and generate grounded outputs from the approved live layer of an Obsidian vault. Use this skill whenever the user asks what approved notes already say, wants grounded answers, ranked local candidates, archived Q&A reuse, publishable artifacts, or static web exports from `wiki/live/`; says "query kb", "kb-search", "archive this answer", "归档这个回答", "复用旧 Q&A", "写成公众号/thread", "对外内容", "一鱼多吃", "问知识库", or "搜索批准层并回答"; or uses Chinese-LLM-Wiki terms such as `综合页`, `output/analyses`, `中文优先分析`, `原文证据摘录`, or `先读 wiki/index.md 再回答`. This is the canonical read-side skill for approved retrieval, candidate ranking, answer archiving/reuse, creator-facing publishing, and static export. Do not use for generic writing, open web research, deterministic rendering, or governance/maintenance work handled by `kb-render` or `kb-review`.

Render deterministic outward-facing artifacts from approved knowledge. Use this skill whenever the user says "kb render", "render slides", "render report", "make a canvas", "generate chart brief", "生成幻灯片", "生成报告", "输出 canvas", `把已批准主题做成报告`, or clearly wants a deterministic derivative from approved live pages or already-grounded archived answers. This is the separate shipped skill for slides, reports, charts, and canvas outputs. Do not use it for retrieval, grounded Q&A, static web export, governance reports, or maintenance work.

Run the canonical governance lane for an Obsidian knowledge base. Use this skill whenever the user says "kb review", "kb health", "review gate", "health check", "approve drafts", "reject draft knowledge", "promote to live", "rebuild briefings", "lint live wiki", "clear archive backlog", "audit archived outputs", "检查归档内容", "检查风格约束是否冲突", "检查账号 briefing", "审校草稿", "批准草稿", "知识库体检", or uses Chinese-LLM-Wiki maintenance wording such as `lint`, `孤儿页`, `断链`, `旧结论被覆盖`, `output/reports`, `治理报告`, `先做 lint`, and `原文证据摘录是否充分`. This skill owns both the immediate gate and the longer-horizon maintenance lane through internal `gate` and `maintenance` modes. Do not use it for normal approved-layer retrieval or deterministic derivative rendering.

Diagnose and route ambiguous, workflow-level review-gated Obsidian vault requests. Use this skill when the user is talking about an Obsidian vault workflow as a whole, asks which lifecycle step should run next, says things like "what should I run first", "which stage am I in", "LLM Wiki", "Karpathy wiki", "Obsidian IDE", "knowledge compiler", "creator knowledge compiler", "personal knowledge base", "second brain", "archive this answer", "归档这个回答", "存回知识库", "复用归档内容", "清理 archive backlog", "先读 wiki/index.md", "这个知识库下一步该做什么", "现在应该初始化/摄取/编译/审校/检索/发布/渲染/体检哪个步骤", or uses Chinese-LLM-Wiki wording such as `中文优先`, `来源页`, `主题页`, `实体页`, `综合页`, `output/analyses`, `output/reports`, `原文证据摘录`, `先读 wiki/index.md 再判断`, or `先判断该做来源页还是综合页` without making the operation explicit. Prefer the operation-specific skills when the user already clearly means init, ingest, compile, review, query, publish, render, or archive maintenance, and only route through this package entry skill when the workflow step is genuinely ambiguous.

Fixture companion skill for tests.

中文 LaTeX 学位论文助手，面向已有 .tex 硕博论文工程：编译诊断、GB/T 7714、模板识别、结构/格式/公式断行、术语一致性、逻辑与文献综述、标题优化、去 AI 味、盲审隐匿、对照学校规范逐项终检。触发词：学位论文/毕业论文/硕士/博士论文。英文论文用 latex-paper-en，审稿总评用 paper-audit。

Search and cite from local BibTeX/BibLaTeX .bib libraries, including Zotero exports. Use to find, filter, preview, export, or generate LaTeX/Typst citation snippets by topic, author, year, venue, DOI, arXiv ID, keywords, abstract, fields, recency, or claim support. Do not use for manuscript writing or polishing.

Submission cover-letter assistant for existing LaTeX manuscripts. Use to generate, optimize, align-check, preflight, and journal-fit-check cover letters against paper evidence, novelty claims, and target venue expectations. Also handles Chinese requests (写投稿信 / 致编辑信). Do not use for editing main.tex, full manuscript audit, bibliography search, or a job-application 求职信.

English LaTeX assistant for existing .tex journal or conference papers. Use for compile repair, venue formatting, bibliography/citation checks, section writing, logic, related work, tables, pseudocode, de-AI polish, translation, adaptation, and submission readiness; use latex-thesis-zh for Chinese theses and paper-audit for critique.

Reviewer-style audit and submission gate for academic papers in .tex, .typ, or .pdf. Use for peer-review critique, readiness/gate decisions, blocker triage, revision roadmaps, journal-style reports, and re-audits. Do not use for source editing, sentence polishing, bibliography search, or compile repair.

Typst paper assistant for existing .typ manuscripts in English or Chinese. Use for compile/export diagnosis, venue formatting, BibTeX/Hayagriva checks, grammar, logic, abstract/title, tables, pseudocode, related work, research-gap framing, adaptation, de-AI polish, translation, and submission readiness; use LaTeX skills for .tex.

Contract-first workflow skill for building a brand-new deck from topic, brief, notes, website, PDF, white paper, or brand assets. Use this whenever the user wants a full new PPT, slide deck, pitch deck, board pack, teaching deck, policy briefing, thesis defense, or structured presentation from scratch and expects research, argument structure, page planning, style direction, and delivery artifacts instead of ad-hoc slide drafting. Trigger even on requests like "帮我做一套 PPT", "make a presentation", or "create slides" when the intent is a brand-new deck. Also trigger when the user wants to stop early at `outline_only` or `spec_only` as a staged checkpoint inside a full new-deck project. Do not use this for existing deck edits, critique-only requests, template tweaks, single-slide asks, export-only work, or lightweight outline / story-direction requests that explicitly do not want AIPPT contracts, research, page planning, or delivery artifacts.

Route new board briefing, board deck, board update, executive operating review, management update, management review, or operating review requests here before generic AIPPT. Use this whenever the user needs a net-new board-facing presentation with dense evidence, decision points, and dated operating metrics. Do not use this scene pack for all-hands updates, team sync decks, or edits to an existing board template.

Route new company introduction, company overview, corporate overview, corporate profile, business overview, enterprise product intro, or customer-facing "about us / who we are / what we do" deck requests here before falling back to generic AIPPT. Use this whenever the user wants a brand-new company or product-led intro deck for prospects, partners, procurement, or customer conversations. Do not use this scene pack for hiring decks, single customer case-study decks, product launch keynotes, or edits to existing files.

Route new investor pitch, fundraising deck, pitch deck, VC deck, startup deck, seed/Series A deck, or venture-facing narrative requests here before generic AIPPT. Use this whenever the user wants a new fundraising story from scratch and expects market, traction, team, and ask to be structured for investors. Do not use this scene pack for product launch decks, internal sales pitch materials, existing-deck reviews, or non-fundraising presentation work.

Route new teaching deck, lecture slides, classroom lesson deck, training deck, training slides, course slides, or workshop deck requests here before generic AIPPT. Use this whenever the user wants a new instructional deck with learning goals, examples, practice, and comprehension-oriented pacing. Do not use this scene pack for manuals, worksheet-only outputs, research talks, or single-slide classroom assets.

Route new thesis defense, dissertation defense, oral defense, viva, academic defense, or research defense deck requests here before generic AIPPT. Use this whenever the user wants a new defense narrative with research question, gap, method, results, contribution, and limitations. Do not use this scene pack for paper proofreading, research posters, conference talks, or non-defense writing tasks.

Create, edit, replicate, import, and export draw.io diagrams with an offline YAML-first workflow: architecture, network topologies, flowcharts, UML/ER, org charts, Mermaid/CSV conversion, existing .drawio bundles, style presets, themes, and non-publication formula diagrams. For publication figures (paper, thesis, IEEE, camera-ready) use drawio-academic-skills instead.

Publication-figure overlay for draw.io. Use instead of drawio whenever the diagram is for a paper, thesis, dissertation, journal, conference, IEEE/ACM submission, manuscript, camera-ready, Word/LaTeX figure, or other publication. Applies venue, figure-type, color, caption/legend, formula, and paper-readability gates for architecture, workflow, roadmap, network-topology, and replicated paper figures.

当用户在 Claude Code、Codex 或类似 agent 里，想直接查询、提取、整理或安全更新本机已有的 Zotero 内容时，必须使用这个 skill。重点是 Zotero 里的 metadata、notes、tags、attachments、PDF fulltext、outline、annotations、collections、saved searches、feeds 和 reading workspace，而不是教人背 CLI。Rust `zot` CLI 只是执行层。适用于库内搜索、citation key 查询、批注与 PDF 提取、workspace 建立与检索、saved search 保存、附件下载、semantic index/search、Scite 检查、配置排障、desktop bridge setup/pair/status，以及无 Web API key 的本机 duplicate merge/dedupe 或明确授权的 Zotero Web API 写操作。不要把它用于泛化找论文、普通总结、引用格式教学、或不落到 Zotero / workspace 的 PDF 处理。

Use when the user wants Zotero-backed research brainstorming, gap analysis, limitation analysis, innovation directions, or next-step research ideas from existing Zotero collections, workspaces, or explicit item keys. This skill turns real Zotero references into traceable Markdown and HTML reports. Do not use for ordinary Zotero lookup, library maintenance, generic web literature search, citation-format teaching, or paper summaries that are not grounded in Zotero sources.