20:"$Sreact.view_transition" 21:I[9754,["/_next/static/chunks/2t36f5p4pjzlb.js","/_next/static/chunks/1gyqeidjm5eoa.js","/_next/static/chunks/0_fjnxc4tt_l_.js","/_next/static/chunks/1ymt1shyhmu-4.js","/_next/static/chunks/3kv2zf0i7s2wc.js","/_next/static/chunks/1vs_3g_g75zmk.js","/_next/static/chunks/2g4dh802s8oz3.js","/_next/static/chunks/1bc3f0shukf1j.js","/_next/static/chunks/192kgs9v_t14j.js","/_next/static/chunks/0z6pw0v5c_lx5.js","/_next/static/chunks/08r60e6icmtkc.js","/_next/static/chunks/3trx6c0i7_x-3.js","/_next/static/chunks/40sro5bzp9bi_.js","/_next/static/chunks/1wgf3tk38y6cf.js"],"ProjectDetail"] 22:T49bc, Awesome AI Pedia · 一站坐标式 AI 知识库

I/OS · 坐标导航 · 参考图

一站坐标式
AI 知识库·

一个把Awesome AI Pedia当作编码坐标导航图来读的项目：每个知识领域都是有框节点，钉在同一张正交笛卡尔格上；领域之间的交叉引用画成带关系标签的直角引线。以 VitePress 构建，推送到 master 即自动发布到 GitHub Pages。

纵轴 · 生命周期阶段 横轴 · 主题深度

浏览知识库 GitHub 仓库

栈 · TypeScript / VitePress 授权 · MIT 维护者 · awesome-ai-pedia 自动发布 · GitHub Pages

坐标键：怎么读这张 Pedia

非滚动营销页，而是一张编码坐标导航图。横轴按主题深度把内容摊开，纵轴按成长阶段（从想法到落地）分卷，所有正文落在发丝网格的交叉点上。

纵轴 · 生命周期阶段 横轴 · 主题深度

网格线贯穿各节；只有「活跃引用」状态使用信号绿，其余一切只有可读的参考线。

领域网格：五个坐标单元

五个知识领域作为有框坐标单元铺满整张格，当前字段由十字热点点亮。每个单元仅来自 README 的目录标题，不发明新分类。

主题深度 → 生命周期 →

A · 01

AI 编码助手

Claude Code、Cursor 等编码助手的入门与场景。

Claude CodeCursor

A · 02

生态与协议

MCP 与各类连接协议的索引，连接 Agent、Skills 的现实链路。

MCP协议

A · 03

提示词与规则

Prompt 工程与书写技艺的集中整理。

Prompt规则

A · 04

技能与工具

Agent 的 Skills 与配套工具的收录。

SkillsAgent

A · 05

学习与实战

零基础路线与实战经验，从入门走向应用。

路线实战

节点交叉引用：字段 × 主题索引

直角引线只在有真实归属时画出——每条线都对应一条可检索的目录关系，不添加任何材料外的连接。

MCP连接 AgentAgent

Agent承载 SkillsSkills

Prompt驱动对话Claude Code

编码助手↔ 生态协议MCP

生态与协议A · 02

MCP协议连接

技能与工具A · 04

AgentSkills工具

提示词与规则A · 03

Prompt规则技艺

生态与协议连线

MCP、Agent、Skills 是同一张协议网的三个节点，提示词与规则承上启下。每条直角引线都有对应索引条目背书。

MCP↔ AgentSkills

Agent依赖 Prompt规则

编码助手启用 SkillsSkills

检索与学习路径

从内置全文搜索定位一个字段，沿直角引线走向相邻领域，再从学习路线落到实战——一条可读的发现路径。

全文检索定位问题所在的字段，视作读取该单元坐标。
沿直角引线顺藤摸到相邻领域，补全上下文。
进入学习与实战，沿零基础路线逐级推进。
返回主题索引，形成你自己的可复用路径。

版权、引用与自动发布

内容由社区持续整理，深度与准确性需自行甄别；结构经全文检索与暗黑主题支撑，推送到 master 即自动重建并发布到 GitHub Pages。

授权

MIT

仓库级许可

语言

TypeScript

项目主语言

构建

VitePress

srcDir 指向仓库根

发布

GitHub Pages

push → 自动部署

Awesome AI Pedia · 坐标导航 · MIT License

23:T4b82, Awesome AI Pedia · A Cartesian AI Knowledge Atlas

I/OS · coordinate navigation · reference sheet

A Cartesian ·
AI knowledge atlas

Treat the Awesome AI Pedia as a coded coordinate navigator: every knowledge field is a framed node pinned to one orthogonal Cartesian lattice, and cross-references between fields are drawn as right-angle wires with relationship labels. Built with VitePress, it auto-publishes to GitHub Pages on every push to master.

vertical · life-stage axis horizontal · topic depth

Browse the pedia GitHub repo

Stack · TypeScript / VitePress License · MIT Maintainer · awesome-ai-pedia Auto-publish · GitHub Pages

Coordinate key: how to read the pedia

Not a scroll of marketing pages but a coded coordinate navigator. The horizontal axis lays out content by topic depth, the vertical axis volumes it by life stage — from idea through production — and every entry sits on a hairline grid intersection.

vertical · life-stage horizontal · topic depth

Grid rules persist across every section; signal green is reserved for live/reference states only, everything else is a legible reference line.

Category lattice: five coordinate cells

Five knowledge fields render as framed coordinate cells across the lattice; a single crosshair hot cell marks the current field. Each cell comes only from the README's own directory headings — no invented categories.

topic depth → life stage →

A · 01

AI coding assistants

Getting started with coding assistants like Claude Code and Cursor.

Claude CodeCursor

A · 02

Ecosystem & protocols

An index of MCP and connection protocols, linking the real chain that runs to agents and skills.

MCPprotocols

A · 03

Prompts & rules

Prompt engineering and the craft of writing rules, gathered in one place.

Promptrules

A · 04

Skills & tools

Agent skills and the tools that carry them, cataloged.

SkillsAgent

A · 05

Learning & practice

Beginner routes and hands-on experience, moving from first principles to applied use.

routespractice

Node cross-reference: field × topic index

Right-angle wires are drawn only where a real relation exists — each edit corresponds to a searchable directory relationship, never an invented link.

MCPlinks AgentAgent

Agentcarries SkillsSkills

Promptdrives dialogueClaude Code

Assistants↔ ecosystemMCP

Ecosystem & protocolsA · 02

MCPprotocolsconnectivity

Skills & toolsA · 04

AgentSkillstools

Prompts & rulesA · 03

Promptrulescraft

Ecosystem & protocol wires

MCP, Agent, and Skills are three nodes on the same protocol network, with prompts and rules bridging the flow. Every orthogonal wire is backed by a matching index entry.

MCP↔ AgentSkills

Agentdepends on Promptrules

Assistantsopt-in SkillsSkills

The search & learn path

Locate a field with the built-in full-text search, follow the right-angle wires into neighboring fields, then land on a route — a readable discovery path.

Full-text search pins the problem to a field, like reading that cell's coordinate.
Follow the right-angle wires to adjacent fields and close the context.
Step into learning & practice and climb the beginner route.
Return to the topic index and build your own reusable path.

Colophon: license, citations & live auto-publish

Content is community-curated, so verify its depth and accuracy yourself; the structure runs on full-text search and a dark theme, and rebuilds plus over a push to master onto GitHub Pages.

License

MIT

repo-level

Language

TypeScript

primary language

Build

VitePress

srcDir → repo root

Publish

GitHub Pages

push → auto-deploy

Awesome AI Pedia · coordinate atlas · MIT License

a:[["$","$20",null,{"enter":{"secondary-forward":"secondary-forward","secondary-back":"secondary-back-enter","default":"none"},"exit":{"secondary-forward":"secondary-forward","secondary-back":"secondary-back-exit","default":"none"},"update":{"secondary-forward":"secondary-forward","secondary-back":"secondary-back-update","default":"none"},"default":"none","children":["$","$L21","cmt2zgyqn005l6271flgw4uyb:zh-CN:beautified",{"project":{"id":"cmt2zgyqn005l6271flgw4uyb","slug":"awesome-ai-pedia-awesome-ai-pedia","name":"awesome-ai-pedia","description":"全面的 AI 知识库与博客平台","authorName":"awesome-ai-pedia","authorVerification":{"verified":false,"verifiedAt":null},"category":"development_tools","subcategory":"data_ai_tools","experienceUrl":"https://awesome-ai-pedia.github.io/Awesome-AI-Pedia/","qualityScore":92,"featured":false,"officialSite":null,"views":{"original":{"available":true},"official":{"available":false,"reason":"disabled_by_policy"},"beautified":{"available":true,"reason":null},"defaultView":"beautified"},"beautifiedView":{"revisionId":"cmt2zgyr5005x6271fth3xeiq","kind":"artifact","artifact":{"schemaVersion":"localized-static-html-v4","kind":"static_html","locales":["zh-CN","en"],"defaultLocale":"zh-CN","documents":{"zh-CN":"$22","en":"$23"},"designDirections":{"schemaVersion":"dodot-design-directions-v1","directions":[{"id":"pedia-grid","name":"Cross-Reference Coordinate Atlas","projectMetaphor":"The '-pedia' is not a scroll of marketing pages but a coded coordinate navigator: every knowledge field is a framed cell pinned to an orthogonal Cartesian lattice, and the cross-references between fields (AI coding assistants ↔ MCP/ecosystem ↔ prompts & rules ↔ skills & tools ↔ learning & practice) are drawn as explicit right-angle wires with relationship labels.","signatureVisual":"A full-bleed, edge-to-edge transparent drafting lattice. Category nodes are thin-walled coordinate cells anchored on grid lines; cross-reference edges are drawn as orthogonal (Manhattan) wire routes with inline relationship tags. Nothing glows and nothing is decorative — every stroke is a legible reference line.","heroGeometry":"One big Cartesian quadrant occupying the first viewport: two labeled axes (topic-depth vs. life-stage), grid-line rules that persist across sections, and the framed category cells distributed on-lattice with a single cursor-style crosshair hot cell for the current field.","sectionSequence":["Coordinate key & axes legend (how to read the pedia)","Category lattice overview (5 fields as coordinate cells)","Node cross-reference: field × topic index","Ecosystem & protocol wires (MCP, agents, skills)","Search & learn path (full-text discovery)","Colophon: license, citations & live auto-publish"],"surfaceModel":"Technical drawing film with a drafting-rule visual system: hairline grid, frameless edge-to-edge bleed, numbered ticks, and one signal annotation color used only for live/reference states. No shadows, no rounded chrome, no glossy cards.","palette":"Blueprint blue ground with near-white stroke, ink-blue ink, and a single signal green reserved for active reference lines; a warm parchment swatch for human-annotated field marks.","typography":"Compact geometric sans for headings, dense technical body set tight, and tabular monospace numerals for all coordinates, call numbers, and version stamps.","desktopComposition":"Full-bleed Cartesian lattice as the persistent spatial frame; a right-side detail rail shows the currently active field's coordinates and its inbound/outbound cross-reference wires.","mobileTransformation":"The two-axis lattice folds into a single-column list of 'coordinates' (each field becomes a position row), with cross-reference labels inlined as small tagged wires that precede the description; the lattice flattens to a compact header key.","mediaStrategy":"No local asset manifest and no qualified local media: drop all media regions. The signature visual is the evidence-driven cartesian reference system rendered entirely in HTML/CSS/SVG-like strokes, so the coordinate lattice substitutes for any product screenshot and never fakes one.","forbiddenPatterns":["Radial ring or hub-and-spoke graph (already used in a recent fingerprint)","Decorative glow, gradient, particle, or shimmer replacing a legible reference line","Voronoi/tiled partition map of neighboring cells","Feature-card wall or three metric stat cards in the first viewport","Browser/editor/terminal/dashboard chrome as an unsupported shell","Fake search input, toggles, or floating controls"],"compositionFingerprint":"full-bleed Cartesian reference lattice with orthogonal wire cross-reference edges and a right-side coordinate detail rail"},{"id":"reference-shelf","name":"Reading-Room Ledger","projectMetaphor":"The knowledge as a lending reference library worked by a careful reader: thick tomes spine-on-spine on an abstracted shelf, each spine a knowledge field, and a call-number rail for locating any entry in the stacks.","signatureVisual":"A stepped shelf-front of receding reference tomes: tall vertical spine bars of decreasing length, each spine embossed with a field name and a small field mark, paired with a shelf-index call-number rail running down the margin.","heroGeometry":"Left-block spine hierarchy occupying the first viewport — a stepped stack of vertical tome spines whose heights encode field prominence, with a call-number tag hanging from the active spine and a quiet ledger ruled line beneath.","sectionSequence":["Shelf-index call-number prompt (locate a field)","The stacks: category shelf-walk","Reading desk: one field in detail","Protocol shelf: ecosystem & MCP front-list","Push-to-press: auto-rebuild & publish run","Colophon: license, authors & shelf end"],"surfaceModel":"Matte bound-paper system: parchment ground, deckled edges on call-number tags, tight ruled baseline grid, and dense letterpress-ink text; one ember field-mark accent only on the active spine.","palette":"Deep ink-plum ground, parchment cream, bone, and one ember accent restricted to the active field-mark; hairline oxblood ruled lines.","typography":"Readable serif display for shelf-Field spines, regular serif body, and tabular monospace for call numbers and shelf indices.","desktopComposition":"Two-part composition: the receding spine-shelf on the left and a ruled ledger of field entries on the right, sharing one ruled baseline and one call-number rail.","mobileTransformation":"The stepped spine-shelf compresses into a vertical accordion of spine-bars (each spine a tappable segment), the ledger follows as stacked ruled entries, and the call-number rail becomes a sticky right margin index.","mediaStrategy":"No qualified local asset: remove media regions entirely. The signature is the typographic spine-and-ledger system, so the evidence lives in composition and type, not in any placeholder image frame.","forbiddenPatterns":["Vertically indented category-stem leaf index (a recent fingerprint)","Library-card edges that restage the same 'index card' grammar as a recent design","Metric stat cards or feature-card walls","Glow, grid, or particle fields as decoration","Fake search/switches/inputs","Desktop columns merely stacked without spatial recomposition"],"compositionFingerprint":"receding reference-tome spine-ledger with shelf-index call-number rail and shared ruled baseline"},{"id":"press-run","name":"Auto-Published Journal Broadsheet","projectMetaphor":"A living newspaper issue whose contents are re-set and re-publiched on every merge to master (auto-build → GitHub Pages). The knowledge fields are columns on a broadsheet front page, and the publish cycle is a running press mark.","signatureVisual":"An editorial broadsheet front page: masthead banner atop ruled newsprint columns, drop caps, hairline column rules, a vertical folio/margin index rail down one edge, and a small press-run ticker showing the fetch→build→deploy cycle.","heroGeometry":"Stacked broadsheet columns framed by a masthead banner in the first viewport, with a folio rail (page/category index) running down the right margin and the press ticker pinned to the base.","sectionSequence":["Masthead & lede (what the pedia is)","Newsbrief columns: category digest","Feature story: one field in depth","Ecosystem pull-quote: MCP & protocols","Next issue / contribute & auto-publish call","Colophon: license, version & press credits"],"surfaceModel":"Newsprint system: warm stock, coarse leading, hairline column rules, ink-bleed serif type, and a single vermilion masthead accent; nothing glossy or floating.","palette":"Ink black on warm newsprint, vermilion reserved for the masthead and press ticker, and faded grey for marginal folio indexes.","typography":"Sharp serif display masthead, dense body serif with justified columns, and letter-spaced small-caps folio numerals.","desktopComposition":"Three-to-four ruled editorial columns under one masthead, with the folio/margin index rail and a persistent press-run ticker at the foot; column heads act as section navigation.","mobileTransformation":"Columns reflow into a single reading column; the masthead shrinks to a compact banner, the folio rail becomes a sticky top index tab, and the press ticker stays as a thin baseline animating the publish cycle.","mediaStrategy":"No local asset present: media regions are removed. The signature visual is the typographic press grammar itself — masthead, columns, and folio — so evidence is expressed through editorial composition rather than any image placeholders.","forbiddenPatterns":["Radial routing disc (a recent fingerprint) as any kind of circular diagram","Feature-card walls or three-metric first screen","Glow or particle decoration","Fake tabs, search, or interactive controls","Desktop columns merely vertically stacked without editorial recomposition"],"compositionFingerprint":"editorial broadsheet column grid with masthead banner, folio margin index rail, and press-run publish ticker"}],"selectedDirectionId":"pedia-grid","selectionRationale":"pedia-grid best expresses the project's actual mechanism: a '-pedia' is an indexed, cross-referenced encyclopedia navigated by a searchable coordinate/taxonomy structure, and the material's own catalog (AI coding assistants, MCP/ecosystem, prompts & rules, skills & tools, learning & practice) is a set of fields whose relationships are the product. Its Cartesian-lattice + explicit right-angle wire reference system is evidence-driven (each wire is a real cataloged cross-reference, not decoration), suits the full-text-search and discovery audience intent, and avoids every recent fingerprint (voronoi partition, radial routing disc, vertical stem index) while requiring no fabricated product screenshot given an empty asset manifest. reference-shelf and press-run are fully distinct in metaphor, signature visual, geometry, and surface, but pedia-grid's relationship-drawing best matches a continuously cross-linked, auto-published knowledge base."},"designBrief":{"schemaVersion":"dodot-design-brief-v1","selectedDirectionId":"pedia-grid","projectMetaphor":"The '-pedia' is not a scroll of marketing pages but a coded coordinate navigator: every knowledge field is a framed cell pinned to an orthogonal Cartesian lattice, and the cross-references between fields (AI coding assistants ↔ MCP/ecosystem ↔ prompts & rules ↔ skills & tools ↔ learning & practice) are drawn as explicit right-angle wires with relationship labels.","signatureVisual":"A full-bleed, edge-to-edge transparent drafting lattice. Category nodes are thin-walled coordinate cells anchored on grid lines; cross-reference edges are drawn as orthogonal (Manhattan) wire routes with inline relationship tags. Nothing glows and nothing is decorative — every stroke is a legible reference line.","compositionFingerprint":"full-bleed Cartesian reference lattice with orthogonal wire cross-reference edges and a right-side coordinate detail rail","heroGeometry":"One big Cartesian quadrant occupying the first viewport: two labeled axes (topic-depth vs. life-stage), grid-line rules that persist across sections, and the framed category cells distributed on-lattice with a single cursor-style crosshair hot cell for the current field.","sectionSequence":["Coordinate key & axes legend (how to read the pedia)","Category lattice overview (5 fields as coordinate cells)","Node cross-reference: field × topic index","Ecosystem & protocol wires (MCP, agents, skills)","Search & learn path (full-text discovery)","Colophon: license, citations & live auto-publish"],"surfaceModel":"Technical drawing film with a drafting-rule visual system: hairline grid, frameless edge-to-edge bleed, numbered ticks, and one signal annotation color used only for live/reference states. No shadows, no rounded chrome, no glossy cards.","palette":"Blueprint blue ground with near-white stroke, ink-blue ink, and a single signal green reserved for active reference lines; a warm parchment swatch for human-annotated field marks.","typography":"Compact geometric sans for headings, dense technical body set tight, and tabular monospace numerals for all coordinates, call numbers, and version stamps.","desktopComposition":"Full-bleed Cartesian lattice as the persistent spatial frame; a right-side detail rail shows the currently active field's coordinates and its inbound/outbound cross-reference wires.","mobileTransformation":"The two-axis lattice folds into a single-column list of 'coordinates' (each field becomes a position row), with cross-reference labels inlined as small tagged wires that precede the description; the lattice flattens to a compact header key.","mediaStrategy":"No local asset manifest and no qualified local media: drop all media regions. The signature visual is the evidence-driven cartesian reference system rendered entirely in HTML/CSS/SVG-like strokes, so the coordinate lattice substitutes for any product screenshot and never fakes one.","evidencePlan":["Render the five cataloged knowledge fields (AI coding assistants, ecosystem & protocols/MCP, prompts & rules, skills & tools, learning & practice) only from readme headings — never invent categories or products","Draw each cross-reference wire/passing edge strictly from a real readme relation (e.g. MCP ↔ agents ↔ skills ↔ prompts); add no relationship that lacks readme support","Use GitHub metadata only where evidenced (TypeScript, MIT license, auto-deploy to Pages on push to master, full-text search & dark theme) for the colophon and press/version stamps","Both zh-CN and en must adopt the same section sequence and coordinate grammar while keeping copy density natural per language"],"localeContract":{"zh-CN":"zh-CN 采用同一坐标格语法与章节顺序：以代码字段轴与生命周期轴作为坐标、五类知识领域作为有框节点、直角连线作为交叉引用。文案密度贴近中文阅读习惯：坐标键、领域网格、节点交叉索引、生态与协议连线、检索学习路径、版权与自动发布小样依次呈现，术语沿用知识库原文（MCP、Agent、Skills、Prompt 等）。","en":"en uses the same Cartesian-graph grammar and section order as zh-CN: topic-depth and life-stage axes, five cataloged knowledge fields as framed nodes, and orthogonal right-angle wires as cross references. Copy is set naturally for English reading density across the coordinate key, category lattice, node cross-index, ecosystem/protocol wires, search-and-learn path, and the colophon with live auto-publish stamp."},"forbiddenPatterns":["Radial ring or hub-and-spoke graph (already used in a recent fingerprint)","Decorative glow, gradient, particle, or shimmer replacing a legible reference line","Voronoi/tiled partition map of neighboring cells","Feature-card wall or three metric stat cards in the first viewport","Browser/editor/terminal/dashboard chrome as an unsupported shell","Fake search input, toggles, or floating controls"]},"visualReview":{"schemaVersion":"dodot-automated-design-review-v1","status":"approved","summary":"页面忠实实现 pedia-grid 坐标格导航方向：全出血发丝网格坐标系、右缘尺标轨、五个来自 README 目录的有框坐标单元、带关系标签的直角交叉引线与字段×主题索引，章节顺序完全贴合 Brief 的 6 段序列；手机断点将五列格折叠为单一坐标列表并隐藏尺标轨，形成真实空间重构而非简单堆叠；双语采用同一坐标语法与章节顺序且各语言文案自然。所有事实均有材料依据，signatureVisual 真实存在，未回退到通用 Hero、三卡卡片墙或无依据设备壳，亦无假控件。","checks":{"projectFit":{"passed":true,"evidence":"6 个 data-dodot-section 章节（coordinate-key / category-lattice / node-cross-index / ecosystem-protocol-wires / search-learn-path / colophon）与 design-brief.json 的 sectionSequence 一一对应；坐标键、领域网格、节点交叉索引、生态连线、检索路径、自动发布小样全部落地，贴合项目作为跨引用编辑式 AI 知识库的机理。"},"evidenceBound":{"passed":true,"evidence":"五个领域（AI 编码助手、生态与协议/MCP、提示词与规则、技能与工具、学习与实战）均直接取自 material.json extraction.readme.headings 及摘要；授权 MIT、语言 TypeScript、VitePress 构建、push→master 自动发布 GitHub Pages、全文检索与暗黑主题均有证据；交叉引线仅在 MCP↔Agent↔Skills↔Prompt 等 README 既有目录关系上画出，未发明分类或连接。"},"signatureVisual":{"passed":true,"evidence":"固定 .film 发丝格栅 + 右缘 .rail 尺标轨 + 双坐标轴标注 + 五个 .cell 有框坐标单元（.hot 十字热点）+ .wire 直角引线带关系标签（.lbl），中英两文件同构，完整呈现 full-bleed Cartesian reference lattice with orthogonal wire cross-reference edges and a right-side coordinate detail rail。"},"mobileTransformation":{"passed":true,"evidence":"@media (max-width:900px) 将 .grid-5 由 repeat(5,1fr) 折叠为 1fr 单一坐标列表、.index .row 折叠、.rail 隐藏，符合 Brief mobileTransformation「两轴格折叠为单列坐标，格子扁平化为紧凑头键」；属真实空间重构而非简单纵向堆叠（页面移动端 scrollHeight 亦相应增长）。"},"localeParity":{"passed":true,"evidence":"zh-CN.en.html 均含 6 章节、同一坐标语法、同一五单元目录与同一信号绿语法、同一组交叉引线；zh-CN 文案贴近中文密度（坐标键/领域格/节点交叉索引等），en 为自然英文（A Cartesian knowledge atlas），语义等价、无翻译腔。"},"antiTemplate":{"passed":true,"evidence":"首屏为坐标轴大象限 + 发丝格而非通用 Hero 或三指标卡；无径向环/轮盘、无 Voronoi 分区图、无特性卡片墙、无浏览器/终端/仪表盘设备壳；.film 上轻微 6% 透明度渐变仅作格底环境氛围，未取代可读参考线，未构成装饰性渐变网格，未命中任何 forbiddenPatterns。"},"noFakeUi":{"passed":true,"evidence":"无假搜索框、无假开关、无飘浮控件；CTA 均为指向真实仓库/页面 URL 的可达链接，qa-report.json 两语言两视口 fakeControlCount=0、emptyLinkCount=0、brokenImageCount=0。"}},"issues":[]},"repairEvidence":[],"provenance":{"executor":"opencode","model":"deepseek/deepseek-v4-flash","opencodeVersion":"1.1.58","promptVersion":"dodot-showcase-deepseek-v4","inputHash":"5cbec0a58355bd7d374fdf55dd34ecb088f908163bd47e7482341ff84596f3fe","artifactHash":"7b8c7cc737b3452ce6930ec94336624b39c2a00d74faf3927b16eb2a078a8278","designSkill":"dodot-showcase-design"},"qaEvidence":{"zh-CN":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":6,"sectionCount":6,"textLength":1535,"cjkCount":758,"latinLetterCount":430,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3332,"h1Top":132.59375,"bottomInset":80.296875,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"62a17db13530a8ff36716af83c447a570ea567b453e23cfa8ea3e1c3c04b5aa1","middle":"b296654f8d67911b2c981eb28dfee4ce8c8ff96895c329ebeb55bfd96ca1f0fc","bottom":"6959751c9305395ed3aeb346630b2661990faf5d0343092ba89dd49788988de5"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":6,"sectionCount":6,"textLength":1500,"cjkCount":758,"latinLetterCount":406,"scrollWidth":390,"clientWidth":390,"scrollHeight":4590,"h1Top":120.59375,"bottomInset":80.21875,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"5b7076dd61cf882cf60d3ff884eebb6c1d3bcb34e355e000fee2ef71b9cf4d43","middle":"f416d2293e75fa35c88438a79becf317afd4c772a6e72786875a2e8f3ea8dc22","bottom":"7d15a0d28959a0ca4fb80e75fe42cfa112abd7977dcdc117634a152b0011c840"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":6,"sectionCount":6,"textLength":3589,"cjkCount":0,"latinLetterCount":2841,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3557,"h1Top":132.59375,"bottomInset":80.265625,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"1b8c15990624641f57c50b02554926a158efc2b03ee3856758c9e98ed4d72d2f","middle":"dd0e4c7e31249b66a6b916d4764da5c9aebb475b0dbf0e23978465415acfc04f","bottom":"c9f887fa878d77f06a964c8d0e50480f02a27f05d165d17efc115273b7976e3c"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":6,"sectionCount":6,"textLength":3554,"cjkCount":0,"latinLetterCount":2817,"scrollWidth":390,"clientWidth":390,"scrollHeight":5147,"h1Top":139.1875,"bottomInset":80.46875,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"50efdd92653584d760089816cb0c02571763ca0cb2fc465b4b985ea75441cc0f","middle":"2832142580129eb2a571a5058df33ca5d77c9d5c22bb5d5bea280be4624a7609","bottom":"eeae383bf5cc9d4c485d7b9b79df810cc60057d155060cee2280b5805bede0f7"},"consoleErrors":[],"networkRequests":[]}]}},"generatorVersion":"opencode-1.1.58:deepseek/deepseek-v4-flash:dodot-showcase-deepseek-v4","publishedAt":"2026-08-20T15:38:11.863Z"},"rawExtraction":{"projectRevisionId":"cmt2zgyqw005r6271bemf1hdz","projectVersion":1,"sources":[{"canonicalKey":"github:awesome-ai-pedia/awesome-ai-pedia","role":"primary","revisionVersion":1,"url":"https://github.com/Awesome-AI-Pedia/Awesome-AI-Pedia"}],"facts":[{"key":"repository_owner","value":"awesome-ai-pedia","evidence":[{"path":"repository.owner","excerpt":"awesome-ai-pedia"}]},{"key":"repository_name","value":"awesome-ai-pedia","evidence":[{"path":"repository.name","excerpt":"awesome-ai-pedia"}]},{"key":"description","value":"AI skills大全 mcp ai知识库 Agent 全维度 AI 资源百科，DSH插件收录大模型、智能 Agent、RAG 检索增强、多模态、MLOps、AI 应用工具、AI面试集、Vibe coding 大全、零基础学习路线，持续更新前沿 AI 开源项目，开发者一站式 AI 导航库","evidence":[{"path":"repository.description","excerpt":"AI skills大全 mcp ai知识库 Agent 全维度 AI 资源百科，DSH插件收录大模型、智能 Agent、RAG 检索增强、多模态、MLOps、AI 应用工具、AI面试集、Vibe coding 大全、零基础学习路线，持续更新前沿 AI 开源项目，开发者一站式 AI 导航库"}]},{"key":"primary_language","value":"TypeScript","evidence":[{"path":"repository.language","excerpt":"TypeScript"}]},{"key":"license","value":"MIT","evidence":[{"path":"repository.license","excerpt":"MIT"}]},{"key":"topics","value":"dsh, dsh-bundle, dsh-plugin, dsh-plugin-market, dsh-plugin-verify, dsh-plugins, dsh-skill","evidence":[{"path":"repository.topics","excerpt":"dsh, dsh-bundle, dsh-plugin, dsh-plugin-market, dsh-plugin-verify, dsh-plugins, dsh-skill"}]},{"key":"source_count","value":"1","evidence":[{"path":"projectRevision.sources","excerpt":"github:awesome-ai-pedia/awesome-ai-pedia"}]}],"taxonomy":{"category":"development_tools","subcategory":"data_ai_tools","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI skills大全 mcp ai知识库 Agent 全维度 AI 资源百科，DSH插件收录大模型、智能 Agent、RAG 检索增强、多模态、MLOps、AI 应用工具、AI面试集、Vibe coding 大全、零基础学习路线，持续更新前沿 AI 开源项目，开发者一站式 AI 导航库"}]},"lifecycle":{"stage":"active","confidence":0.75,"evidence":[{"path":"repository.pushedAt","excerpt":"2026-08-17T03:38:28.000Z"}]},"userIntents":[{"slug":"build_software","label":"构建或改进软件","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI skills大全 mcp ai知识库 Agent 全维度 AI 资源百科，DSH插件收录大模型、智能 Agent、RAG 检索增强、多模态、MLOps、AI 应用工具、AI面试集、Vibe coding 大全、零基础学习路线，持续更新前沿 AI 开源项目，开发者一站式 AI 导航库"}]},{"slug":"automate_work","label":"自动化工作流程","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI skills大全 mcp ai知识库 Agent 全维度 AI 资源百科，DSH插件收录大模型、智能 Agent、RAG 检索增强、多模态、MLOps、AI 应用工具、AI面试集、Vibe coding 大全、零基础学习路线，持续更新前沿 AI 开源项目，开发者一站式 AI 导航库"}]},{"slug":"create_content","label":"创作内容或设计","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI skills大全 mcp ai知识库 Agent 全维度 AI 资源百科，DSH插件收录大模型、智能 Agent、RAG 检索增强、多模态、MLOps、AI 应用工具、AI面试集、Vibe coding 大全、零基础学习路线，持续更新前沿 AI 开源项目，开发者一站式 AI 导航库"}]},{"slug":"learn_explore","label":"学习与探索","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI skills大全 mcp ai知识库 Agent 全维度 AI 资源百科，DSH插件收录大模型、智能 Agent、RAG 检索增强、多模态、MLOps、AI 应用工具、AI面试集、Vibe coding 大全、零基础学习路线，持续更新前沿 AI 开源项目，开发者一站式 AI 导航库"}]}],"readme":{"headings":["🦞 Awesome AI Pedia","✨ 项目简介","🎯 核心特性","📂 知识库分类","🤖 AI 编码助手","🔌 生态与协议","💬 提示词与规则","🛠️ 技能与工具","📚 学习与实战","🚀 快速开始","方法一：使用启动脚本（推荐）","首次运行需要授予执行权限","一键启动：自动检查 Node 版本、安装依赖、启动 dev server","🗂️ 项目结构","📞 联系方式"],"commands":[]},"changeSet":[{"field":"description","beforeHash":"466ad2ff237af071c75d24d65d806ca6f8e79f486556afc616bab72b31ace85a","afterHash":"8d348a6c11755deb4595c2b7105f1194620d60266c8ca0ae76c850facadde6e2"},{"field":"repository","beforeHash":"c2212612ac8a367d75882447cbb195cbf8ce39b6b4f1680ebcd898a063f70e22","afterHash":"8bd861dd6c9d69e5398cdfb0f84a497a92ed786f04e706a31f1025b8b8753c0b"}]},"publishedAt":"2026-08-16T12:34:55.055Z","updatedAt":"2026-08-21T13:26:48.325Z","summary":{"id":"cmt2zgyr0005t6271c3l7a2yw","projectType":"other","oneLine":"全面的 AI 知识库与博客平台","overview":"Awesome AI Pedia 是一个现代化的 AI 知识库，为开发者提供全面的 AI 工具使用指南、最佳实践和实战经验。内容覆盖 Claude Code、Cursor、MCP、Agent、Skills、Prompt 工程等主流方向，按目录分类整理，内置全文搜索和暗黑主题，支持自动部署到 GitHub Pages。","highlights":["覆盖主流 AI 方向","实战导向内容","本地全文搜索","自动部署"],"bestFor":["想系统学习 AI 工具的开发者","需要 AI 面试准备的学习者","寻找 Vibe Coding 实践指南的用户"],"caveats":["内容为社区整理，深度和准确性需自行甄别","部分目录可能更新不及时","依赖 GitHub Pages 部署，访问速度可能受限"],"categorySuggestion":"education","subcategorySuggestion":"documentation_reference","lifecycleStages":["idea_requirements","product_design","frontend","backend_data"],"audienceIntents":["learn","discover","build"],"confidence":0.85,"evidence":[{"claim":"Awesome AI Pedia 是一个全面的 AI 知识库与博客平台，覆盖 Claude Code、Cursor、MCP、Agent、Skills、Prompt 工程等方向。","sourceType":"readme","locator":"https://github.com/Awesome-AI-Pedia/Awesome-AI-Pedia"},{"claim":"知识库按目录分类，包括 AI 编码助手、生态与协议、提示词与规则、技能与工具、学习与实战等。","sourceType":"readme","locator":"https://github.com/Awesome-AI-Pedia/Awesome-AI-Pedia"},{"claim":"内置全文搜索、暗黑主题，并支持推送到 master 自动构建部署到 GitHub Pages。","sourceType":"readme","locator":"https://github.com/Awesome-AI-Pedia/Awesome-AI-Pedia"},{"claim":"项目使用 VitePress 构建，通过 srcDir 将仓库根目录作为文档源，侧边栏自动扫描生成。","sourceType":"readme","locator":"https://github.com/Awesome-AI-Pedia/Awesome-AI-Pedia"}],"publishedAt":"2026-08-17T13:14:54.246Z"},"github":{"url":"https://github.com/Awesome-AI-Pedia/Awesome-AI-Pedia","ownerAvatarUrl":"https://avatars.githubusercontent.com/u/238374711?v=4","starsCount":167,"forksCount":21,"language":"TypeScript","syncedAt":"2026-08-17T13:07:28.427Z","license":"MIT","defaultBranch":"master","archived":false,"disabled":false,"lastChangedAt":"2026-08-17T03:38:28.000Z","missing":false},"tags":[],"updates":[]},"entryPoint":"discovery","claim":{"state":"unclaimed","verified":false},"initialView":"beautified","community":{"viewer":null,"state":{"likesCount":0,"creativeCount":0,"followersCount":0,"commentsCount":0,"liked":false,"creative":false,"followed":false},"comments":{"comments":[],"nextCursor":null}}}]}],"$L24"]