20:"$Sreact.view_transition" 21:I[9754,["/_next/static/chunks/2qcf_o2xotudn.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/1ucasf31l8j2v.js","/_next/static/chunks/0yj6_8vip2oqy.js","/_next/static/chunks/2l5xdnifczg1b.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:T421f, MonkeyCode — 模型路由矩阵

模型调度矩阵 / Model Routing Map

MonkeyCode chaitin / monkeycode — TypeScript · AGPL-3.0

同一个需求，交给可插拔的模型矩阵去调度。

MonkeyCode 是一个开源的企业级 AI 开发平台：你把需求交给一台可切换的模型路由台，平台在真实的云端服务端环境里完成构建、测试与预览，最终回填到同一份团队任务面。

CLOUD RUNTIME MULTI-MODEL MOBILE SYNC PRIVATE DEPLOY

进入 MonkeyCode ↗ GitHub 源码 ↗

ROUTING BOARD 图 01

需求 ▼ 路由干道

GLM

Kimi

MiniMax

Qwen

DeepSeek

B 构建 BUILD

T 测试 TEST

P 预览 PREVIEW

多模型，一个入口，按需切换。

平台内置了 GLM、Kimi、MiniMax、Qwen、DeepSeek 等主流模型。你可以按任务的类型来自动匹配合适的模型，也可以在需要时手动切换——模型只是可插拔的节点，需求始终走同一条路由干道。

GLM Kimi MiniMax Qwen DeepSeek

任务驱动

按任务类型自动匹配

也可手动

随时切换所选节点

一条干道

需求始终单一入口

构建、测试、预览，都在云端运行时里完成。

每一个任务都在真实的服务端环境里运行，而不是在本地模拟。构建、测试和预览作为同一份运行时里的三段推进轨道，全部在云端落地并回填结果。

B构建 · Build

节点计算出产物，服务端环境完成编译与打包。

T测试 · Test

同一运行时内执行校验，验证任务是否符合预期。

P预览 · Preview

在云端实时预览运行效果，无需本地安装任何环境。

每一条分支，最终回填到同一份团队任务面。

需求进入、模型计算、产物生成，这些环节不是彼此孤立的块，而是一条连续的流程。所有结果都回填到同一个任务面上，让团队能围绕一个可追踪的作业进行协作与跟进。

需求从干道持续进入，任务状态一以贯之
模型节点切换时，同一任务面保持连续
构建 / 测试 / 预览复用同一份任务上下文
团队围绕单一可追踪作业协作

打开浏览器即可开始，无需本地配置。

不需要下载客户端，也不用手动搭本地环境。开发环境、模型、任务与需求都被收敛到浏览器这一个极简入口里——在线即用，随时开始你的 AI 开发任务。

在线即用

打开浏览器直接开始

零配置

无需下载客户端

无需本地环境

一切在云端执行

核心代码开源，支持私有离线部署。

MonkeyCode 的核心代码在 GitHub 上公开，采用 AGPL-3.0 许可证，同时支持私有离线部署。作为面向团队的开源 AI 编码平台，它让有数据管控诉求的团队既能掌握代码，也能按需进行隐私化部署。

开源

核心代码公开于 GitHub

许可证

AGPL-3.0

部署

支持私有离线部署

在 GitHub 上查看 MonkeyCode 源码 ↗

23:T40be, MonkeyCode — Model Routing Map

Model Routing Map

MonkeyCode chaitin / monkeycode — TypeScript · AGPL-3.0

One requirement, routed through a hot-swappable model matrix.

MonkeyCode is an open-source, enterprise-grade AI development platform. You hand a requirement to a switchable routing desk; the platform builds, tests, and previews it inside a real server-side cloud runtime, then folds the result back into a shared team task surface.

CLOUD RUNTIME MULTI-MODEL MOBILE SYNC PRIVATE DEPLOY

Try MonkeyCode ↗ Source on GitHub ↗

ROUTING BOARD FIG 01

Requirement ▾ Trunk

GLM

Kimi

MiniMax

Qwen

DeepSeek

B Build

T Test

P Preview

Many models, one entry point, switched as needed.

The platform ships with GLM, Kimi, MiniMax, Qwen, and DeepSeek among its supported models. It can match a model to the task at hand automatically, or you can switch manually — models are just pluggable nodes while the requirement always flows down the same trunk.

GLM Kimi MiniMax Qwen DeepSeek

Task-driven

Auto-matched by task type

Manual too

Switch the active node anytime

Single trunk

One constant entry for requirements

Build, test, and preview — all inside the cloud runtime.

Every task runs in a real server-side environment rather than a local simulation. Build, test, and preview are three advancing tracks within that same runtime, executed in the cloud and folded back as results.

BBuild

The node produces an artifact, compiled and packaged on the server side.

TTest

The same runtime validates the result against expectations.

PPreview

Run the output live in the cloud, with no local environment to install.

Every branch folds back into one shared team task surface.

A requirement entering, a model computing, an artifact being produced — these are not isolated blocks but one continuous flow. Every result lands back on the same task surface, so the team can follow and collaborate around a single, traceable job.

Requirements keep entering through the trunk, status stays coherent
The task surface holds steady while model nodes switch
Build, test, and preview reuse the same task context
The team collaborates around one traceable job

Open a browser and start — no local setup required.

No client to download, no local stack to assemble. Your environments, models, tasks, and requirements are gathered into a single minimal browser entry — online, immediate, ready whenever you are.

Online-ready

Start straight from the browser

Zero config

No client download

No local env

Everything executes in the cloud

Core code is open source, deployable privately and offline.

The core of MonkeyCode is public on GitHub under the AGPL-3.0 license, while private offline deployment is supported. As an open-source AI coding platform for teams, it lets teams with data-control needs keep the code in hand and deploy privately as required.

Open source

Core code public on GitHub

License

AGPL-3.0

Deployment

Supports private offline setup

View MonkeyCode on GitHub ↗

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","cmsndouy8003ncz475h0ch7nh:zh-CN:beautified",{"project":{"id":"cmsndouy8003ncz475h0ch7nh","slug":"chaitin-monkeycode","name":"monkeycode","description":"面向团队的开源 AI 编码平台","authorName":"chaitin","authorVerification":{"verified":false,"verifiedAt":null},"category":"development_tools","subcategory":"data_ai_tools","experienceUrl":"https://monkeycode-ai.net/","qualityScore":87,"featured":false,"officialSite":null,"views":{"original":{"available":true},"official":{"available":false,"reason":"disabled_by_policy"},"beautified":{"available":true,"reason":null},"defaultView":"beautified"},"beautifiedView":{"revisionId":"cmt2zgz4z00as6271zjnzg09e","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":"task-runtime-stack","name":"任务即运行时 / 云端工作台叠层","projectMetaphor":"每一个 AI 任务都是一个在云端服务器端真实运行的作业；平台像一座多层的云端作业塔，构建、测试、预览在同一运行时里逐层完成。浏览器的极简入口只是塔顶的一扇小窗。","signatureVisual":"一组纵向层叠的『任务作业片』，每片是一块代表该任务的画布，边缘标有一条『云端运行时』的状态刻度，构建→测试→预览作为该片内纵向推进的三段轨道；多片从后向前错层堆叠，像待命中的作业队。","heroGeometry":"纵向前景一张主任务作业片，其内呈三段式垂直轨道（轨道/构建/测试/预览）向上推进；其后数张更小、更灰的作业片以纵深错位叠在背景，形成一座微微前倾的作业塔。","sectionSequence":["runtime-tower-hero","cloud-environment","model-switching","task-flow","mobile-sync","private-deploy"],"surfaceModel":"粗糙分层玻璃：前景作业片为磨砂高亮面（可读工程内容），后景作业片为压暗的拉丝面；每张片带一条独立侧面厚度线，表达『塔的层数』而非卡片。面与面之间靠厚度和明度分层，不用圆角卡片堆叠。","palette":"近黑墨底 #0e1620，运转中的作业片用暖工程绯?#a more precise blue) 主体 #2f6bff，构建/测试/预览分别用青、黄、绯作三段状态色；状态刻度线用低饱和灰蓝。","typography":"粗体大标题用无衬线的工程黑体（650–700 weight，紧字距），作业片内的小标签用等宽字体，模拟运行时日志/灰级标签；数字编号（01/02/03）强调塔的层级。","desktopComposition":"左侧固定叙述纵列（标题+行动），右侧为作业塔：前景大片占主，背景用纵深错位弱化；段与段之间以『塔高回落』的负空间过渡，云环境与任务流用同一套作业片语言延续。","mobileTransformation":"塔的三维纵深在窄屏压平成单列纵向流：每张作业片按构建→测试→预览自下而上堆叠为一条推进条；状态刻度改作屏幕左侧的竖向进度轨，01/02/03 编号变成行内小标。","mediaStrategy":"未采用外部媒体区，无合格本地素材；视觉由作业片/刻度组成的证据驱动图表承担，不放置任何远程或占位图。","forbiddenPatterns":["不出现浏览器/终端/编辑器外壳","不出现三卡指标或功能卡片墙","不出现发光网格或粒子背景","不伪造模型下拉框/开关等控件"],"compositionFingerprint":"云端作业塔：多层作业片纵深错位 + 每片三段式垂直推进轨道 + 左侧工程叙事纵列"},{"id":"model-routing-map","name":"模型调度台 / 路由矩阵","projectMetaphor":"平台把同一个编码需求交给多维可切换的模型，像一个可插拔的模型调度矩阵：输入经路由汇入某一节点，节点计算后再分支到构建、测试与预览，最终回填同一份团队任务面。","signatureVisual":"一组代表模型的离散节点（GLM、Kimi、MiniMax、Qwen、DeepSeek）以可动插槽排列在若干轨道上，相同需求经一条路由干道汇入当前选中节点，节点之后以三条分支散开对应构建/测试/预览——节点与分支都是真实的流程关系图，不冒充截图。","heroGeometry":"左入一条明亮的需求路由干道，向右扇形汇聚到一个居中高亮的模型节点，再从该节点向下/右展开构建、测试、预览三条分支；整图是一张扇形的调度分布图，非对称单焦点。","sectionSequence":["routing-hero","model-dial","cloud-runtime","team-task-surfacing","browser-entry","offline-deploy"],"surfaceModel":"图纸型浅底工程面：大面积的暖纸白规整网格作为路由底图，模型节点是高对比色块，路由干道与分支是实线路径；各章共享同一张『图纸』但滚动时视口跟随当前区域，边线、坐标小标强化制图语言。","palette":"暖纸白底 #f6f1e7，线条墨 #1d2733，需求干道为引擎绯红，模型节点用沉着靛蓝，构建/测试/预览分支分别用青/黄/绯；整体印染感、扁平、无渐变。","typography":"正文用正经的无衬线正文字（常规 400–500），节点名与路由标签用粗体工程字；章节标题用超大号但克制、少字距，让图面自己说话。","desktopComposition":"以一张持续存在的『路由图纸』为主景：Hero 用非对称扇形矩阵占满主区，右侧或下沿保留一列叙述；后续章节在图纸上平移，各章以不同区域高亮，而非换一套卡片语言。","mobileTransformation":"扇形矩阵在窄屏收拢成一条纵向干线：需求在顶部，模型节点横向平铺成可横滑的一行插槽，三条分支纵排为构建→测试→预览；图纸底纹降为细横线，保留制图身份。","mediaStrategy":"无外部媒体区；视觉完全由模型→构建/测试/预览的流程关系图承担，全部为派生的关系示意（有证据：模型集成与云端构建测试预览），不冒充产品截图。","forbiddenPatterns":["不使用浏览器壳或终端壳做容器","不使用三卡指标墙","不伪造模型选择框或开关控件","不使用发光网格/粒子作背景"],"compositionFingerprint":"模型路由矩阵：扇形单焦点汇聚 + 需求入→模型节→构建测试预览三支分流 + 整页一张暖底制图 + 章节在图纸上平移"},{"id":"cross-device-continuum","name":"跨端联合 / 同一帧状态","projectMetaphor":"MonkeyCode 让 PC 与移动端共享同一份任务状态、在线即开、无需本地环境；平台像一台把『浏览器入口』与『移动端掌持』接到同一根云状态的联控杆，两端看到的是同一帧正在运行的任务。","signatureVisual":"一条居中的『云端同步脊』将一个宽屏入口帧与一个竖屏掌持帧勾连；两帧内部渲染的是同一棵任务推进树（同一状态的不同视口），同步脊上的数据流环沿脊上下移动，表达实时一致。","heroGeometry":"以一条竖向中央同步脊为轴：左为宽屏入口景框、右为竖屏手机景框，二者镜像同高、沿同一棵树；树以根在上、叶在下的结构向下生长，数据环沿脊循环，整体是对称的长廊构图但两翼内容分开叙述。","sectionSequence":["sync-spine-hero","browser-entry","cloud-runtime","multi-model","mobile-pairing","team-state","open-source"],"surfaceModel":"极简的『帧内画布』：两翼作为内容帧共享同一层半透明面板，中央同步脊是唯一的实色元素；帧内播放同一棵树，帧外用大留白与细描边，表面模型是『两帧 + 一脊』而非卡片阵列。","palette":"冷灰深底 #10151d，两翼帧为半透明面板，中央同步脊用通体电蓝色，树节点与推进环在两端用相同的三个状态色（青/黄/绯）；两端同色以强化『同一状态』。","typography":"标题与正文用统一的无衬线体系，重点在轻巧与节奏；竖屏帧内的小文档标签与编号用细等宽字，显得像设备系统级文本；两端用镜像的行内标签。","desktopComposition":"Hero 为对称长廊：中轴同步脊、左宽右竖两翼展开同一棵任务树；后续章节转成页面主体上的宽栏叙述，移动端小节把竖屏帧单独挑出放大，接在同步脊之后。","mobileTransformation":"中轴长廊在窄屏塌缩为上下两段：先宽屏入口帧（指代浏览器入口），随后同步脊转为一条水平推进线，再展开竖屏帧占据全屏（表达掌持移动端）；树改为窄纵列，数据环变作进度点。","mediaStrategy":"无外部媒体区；视觉由『两端共享同一棵树』的同步关系承担——这是有证据的关系（PC 与移动端数据同步、在线即用），以派生的帧内树示意呈现，不伪造截图。","forbiddenPatterns":["不使用真正的浏览器/手机渲染壳冒充截图","不使用三卡指标或卡片墙","不使用发光网格/粒子背景","不伪造同步开关或设置控件"],"compositionFingerprint":"跨端同步脊：中央竖轴 + 左右宽/竖双帧渲染同一任务树 + 循环数据环 + 两翼镜像同色强调同一状态"}],"selectedDirectionId":"model-routing-map","selectionRationale":"在三个方向中，'model-routing-map' 最能表达 MonkeyCode 的项目机制：它有可验证的证据支撑——多模型集成（GLM/Kimi/MiniMax/Qwen/DeepSeek 可切换）与云端构建测试预览形成一条真实的需求→模型→产物流程，而浏览器即用、跨端同步、私有部署则分别作为该流程的入口与延伸。相较：'task-runtime-stack' 的作业塔偏重运行时单体，弱化了模型切换这一核心差异点；'cross-device-continuum' 强调双端同步，更适合作支持性叙事而非主机制。'model-routing-map' 的主视觉（扇形单焦点汇聚 + 分支分流）既是流程亦是数据结构，彻底规避营销 Hero、三卡指标与浏览器/终端壳等禁用默认，且其暖纸制图表面与扁平印染配色显著区别于历史流水线（左至右管段），满足 antiTemplate 与历史隔离要求。"},"designBrief":{"schemaVersion":"dodot-design-brief-v1","selectedDirectionId":"model-routing-map","projectMetaphor":"平台把同一个编码需求交给多维可切换的模型，像一个可插拔的模型调度矩阵：输入经路由汇入某一节点，节点计算后再分支到构建、测试与预览，最终回填同一份团队任务面。","signatureVisual":"一组代表模型的离散节点（GLM、Kimi、MiniMax、Qwen、DeepSeek）以可动插槽排列在若干轨道上，相同需求经一条路由干道汇入当前选中节点，节点之后以三条分支散开对应构建/测试/预览——节点与分支都是真实的流程关系图，不冒充截图。","compositionFingerprint":"模型路由矩阵：扇形单焦点汇聚 + 需求入→模型节→构建测试预览三支分流 + 整页一张暖底制图 + 章节在图纸上平移","heroGeometry":"左入一条明亮的需求路由干道，向右扇形汇聚到一个居中高亮的模型节点，再从该节点向下/右展开构建、测试、预览三条分支；整图是一张扇形的调度分布图，非对称单焦点。","sectionSequence":["routing-hero","model-dial","cloud-runtime","team-task-surfacing","browser-entry","offline-deploy"],"surfaceModel":"图纸型浅底工程面：大面积的暖纸白规整网格作为路由底图，模型节点是高对比色块，路由干道与分支是实线路径；各章共享同一张『图纸』但滚动时视口跟随当前区域，边线、坐标小标强化制图语言。","palette":"暖纸白底 #f6f1e7，线条墨 #1d2733，需求干道为引擎绯红，模型节点用沉着靛蓝，构建/测试/预览分支分别用青/黄/绯；整体印染感、扁平、无渐变。","typography":"正文用正经的无衬线正文字（常规 400–500），节点名与路由标签用粗体工程字；章节标题用超大号但克制、少字距，让图面自己说话。","desktopComposition":"以一张持续存在的『路由图纸』为主景：Hero 用非对称扇形矩阵占满主区，右侧或下沿保留一列叙述；后续章节在图纸上平移，各章以不同区域高亮，而非换一套卡片语言。","mobileTransformation":"扇形矩阵在窄屏收拢成一条纵向干线：需求在顶部，模型节点横向平铺成可横滑的一行插槽，三条分支纵排为构建→测试→预览；图纸底纹降为细横线，保留制图身份。","mediaStrategy":"无外部媒体区；视觉完全由模型→构建/测试/预览的流程关系图承担，全部为派生的关系示意（有证据：模型集成与云端构建测试预览），不冒充产品截图。","evidencePlan":["主视觉模型节点列表取自 verifiedClaims 集成的主流模型（GLM、Kimi、MiniMax、Qwen、DeepSeek），仅作列式示意，不伪造能力。","构建→测试→预览三支分流来自 readme 云端开发环境证据：每个任务在真实服务端环境运行、构建测试预览皆在云端。","browser-entry 章节基于 README 『打开浏览器即可开始、无需本地配置』；offline-deploy 基于 AGPL-3.0 与私有离线部署证据；均不引入额外数字。"],"localeContract":{"zh-CN":"正文用中文，品牌名 MonkeyCode 保留原文；模型名、AGPL-3.0、TypeScript 等技术专名不译。","en":"Body copy in English; brand term MonkeyCode kept as-is; model names (GLM, Kimi, MiniMax, Qwen, DeepSeek), AGPL-3.0, TypeScript kept untranslated."},"forbiddenPatterns":["不使用浏览器壳或终端壳做容器","不使用三卡指标或功能卡片墙","不伪造模型选择框或开关控件","不使用发光网格/粒子作背景"],"historicalReuse":{"sourceAssetId":"HEAD:deepseek-v2/chaitin-monkeycode","decision":"rebuild","rationale":"旧方向把项目描成『需求→理解→变更→审查』的横向流水线，其 sectionSequence、compositionFingerprint 与 signatureVisual 均为旧叙事；本方向选择 rebuild，改用可被当前证据直接支撑的『模型→构建/测试/预览』机制。","reusedElements":["暖纸浅底 #f4f6f8 → 暖纸白 #f6f1e7 的成片浅色调性（保留纸感亲近），但换成非蓝冷的印染配色。","等宽/灰级小标签用于工程标签的排印手法，用于模型节点与路由干道标签。","扁平、无浏览器壳的示意式专有视觉原则，延续『定制结构而非通用落地页』的姿态。"],"discardedElements":["横向四段流水线与管线/衔接阀的 compositionFingerprint（前向直通叙事不符合当前机制主轴）。","左至右管段 + 物料流动的签名视觉与纵深移动方式。","旧 sectionSequence（pipeline-hero/requirement/understanding/change/review/media）。","蓝绿主色架构（旧 #2563eb 主色）与 Geist 组合，改用印染纸面与制图语言。"]}},"visualReview":{"schemaVersion":"dodot-automated-design-review-v1","status":"approved","summary":"页面源码忠实实现了所选方向 model-routing-map 的专属 Brief：Hero 的 ROUTING BOARD 提供需求路由干道→GLM(active) 模型插槽→构建/测试/预览三分支的路由矩阵签名视觉，是真实流程关系图而非冒充截图；六个章节顺序与 Brief.sectionSequence 完全一致；暖纸白 #f6f1e7/墨线 #1d2733/绯红干道/靛蓝节点/青黄绯分支的印染制图配色落地；手机 620px 断点把扇形矩阵重构为横向可横滑插槽行 + 纵排三分支 + 细横线底纹，真正形成 brief 所述的移动端重排；双语章节逐节等价、译文自然，MonkeyCode/模型名/AGPL-3.0/TypeScript 保留原文；无通用 Hero/三卡墙/卡片墙/设备壳/发光网格/假控件。所有事实均有材料证据，未新增数字。","checks":{"projectFit":{"passed":true,"evidence":"sectionSequence=[routing-hero,model-dial,cloud-runtime,team-task-surfacing,browser-entry,offline-deploy] 与 design-brief.json.sectionSequence 一致（两端 HTML data-dodot-section 均存在）；compositionFingerprint（需求入→模型节→构建测试预览三支分流 + 整页暖底制图）在 Hero diagram 与 cloud-runtime flow-grid 中实现；hero 文案『把需求交给可插拔的模型矩阵去调度』贴合所选机制。"},"evidenceBound":{"passed":true,"evidence":"模型节点 GLM/Kimi/MiniMax/Qwen/DeepSeek 取自 material.json.summary.evidence 『集成了 GLM、Kimi、MiniMax、Qwen、DeepSeek 等主流模型』并仅作列式示意；构建→测试→预览三支来自『云端开发环境，每个任务在真实服务端环境运行，构建、测试和预览都在云端完成』；browser-entry 基于『无需下载客户端或配置本地环境，打开浏览器即可开始』；offline-deploy 基于『AGPL-3.0 许可证，支持私有离线部署』。未引入额外数字或伪能力。"},"signatureVisual":{"passed":true,"evidence":"zh-CN.html:279-303 与 en.html:216-241 的 Routing Board 实现完整签名视觉：5 个 model-slot 插槽（GLM 为 slot.active 高亮选中）、route-label『需求路由干道』为入口、SVG 箭头连线、.branches 提供构建(B/青)/测试(T/黄)/预览(P/绯) 三分支；节点与分支为流程关系示意，role=\"img\" aria-label，不冒充截图。"},"mobileTransformation":{"passed":true,"evidence":"两文件 @media(max-width:620px) 中：.model-slots 由 2 列网格改为 flex row + overflow-x:auto（横向可横滑的一行插槽）、.slot flex:0 0 auto 白不换行、.branches 改为 column 纵排 B→T→P、body::before 背景仅留 100% 54px 横向细线，符合 Brief.mobileTransformation 的『需求顶部、节点横向可横滑插槽一行、三分支纵排构建→测试→预览、底纹降为细横线』，构成真正重构。"},"localeParity":{"passed":true,"evidence":"zh-CN 与 en 结构逐节对应（6 section、sec-num 01-05、flow-step、meta-row、wide-checklist、dial 完全平行）；品牌 MonkeyCode、模型名 GLM/Kimi/MiniMax/Qwen/DeepSeek、AGPL-3.0、TypeScript 在两端保留原文，符合 localeContract；en 文本为 zh 的自然扩展而非错位；qa-report.json 显示两语言各视口 0 consoleErrors、0 假控件、无 broken images。"},"antiTemplate":{"passed":true,"evidence":"签名视觉为专属路由矩阵制图而非通用 Hero/三卡卡片墙；flow-grid 的 B/T/P 是证据驱动的 build/test/preview 流程段而非指标计数卡；整页延续同一暖纸制图语言；无浏览器/终端/手机设备外壳；背景为扁平工程网格 body::before 而非发光网格/粒子，未回退通用落地页模板。"},"noFakeUi":{"passed":true,"evidence":"slot.active 与 dial.hot 仅以边框/底色/投影表达选中态，.slot/.dial/.meta/.branch 均为静态标签且 diagram 为 role=img 示意，无模型下拉框/开关/同步等可交互假控件；qa-report.json 两语言两视口 fakeControlCount=0；链接仅 homepageUrl 与 GitHub 真实地址。"}},"issues":[]},"repairEvidence":[{"attempt":1,"before":{"qaPassed":true,"designReviewApproved":false,"qaEvidence":{"zh-CN":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1285,"cjkCount":727,"latinLetterCount":307,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3240,"h1Top":217.015625,"bottomInset":200.375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"ea5219f88475434fdee5aa712a21b05878108cab004ded18095c47b4eae17348","middle":"0b9ca52ab42a4c351d645a64066d9bab2a48f8df6408c92c32dd92bb7659131f","bottom":"ef6f70a3e94e8f7c8a33d30146746807e4166a2e92a72414444f6849b4508947"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1285,"cjkCount":727,"latinLetterCount":307,"scrollWidth":390,"clientWidth":390,"scrollHeight":4378,"h1Top":211.21875,"bottomInset":199.78125,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"6b21dd0aee2239e72c56c697bad7b00ee8df8dcfb1b3b1a35cd221708a182b0c","middle":"d576c801d57805fdc6b7eefcadfd34dda966d9734ad4fd080c936ca856409f40","bottom":"1b1236304afb0efa4108232b59c73ebda910a225adec1a99b8c2df804039f470"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":2942,"cjkCount":0,"latinLetterCount":2349,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3513,"h1Top":217.015625,"bottomInset":200.484375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"6f5b5e55a7804457904eb8512a59db988f6125e2e415136b25e3e4b52d19be0a","middle":"85aacf0a54d44df676198c97fee37dcd76443bb0a8cdfead9d7e7cd4dcc49bb7","bottom":"a4ebe0267a824d13f26522f1100b3c38aa2dc70e46af94fac55e27cdec387aa9"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":2942,"cjkCount":0,"latinLetterCount":2349,"scrollWidth":390,"clientWidth":390,"scrollHeight":4882,"h1Top":211.21875,"bottomInset":199.765625,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"93a45df5fea8ab46486eaa2fbfd48eb342a892fda45340e869451906ca388220","middle":"0140bc3a642b0a34417f22a04860acb096e01e93af3635ee9615410e2ae873f1","bottom":"b760e103ff159677247e1f6ffe4fe6009ad7d520833f5364605547951403e3d1"},"consoleErrors":[],"networkRequests":[]}]},"visualReview":{"schemaVersion":"dodot-automated-design-review-v1","status":"revise","summary":"页面源码忠实实现了 design-brief.json 所选定的 model-routing-map 方向：温热纸面制图底、需求干道→模型节点→构建/测试/预览三支分流的专属路由示意图真实存在且为流程关系图而非截图，所有机制与数据均有材料证据支撑，双语版本结构等价且译文自然，无回退到通用 Hero/三卡指标墙/设备壳/发光网格，也无假控件。唯一未达要求的是手机端 CSS：brief 明确要求将扇形矩阵收拢成纵向干线并把模型节点改作横向可横滑的一行插槽、底纹降为细横线，但实现仅在 620px 以下把栏目改为一列，签名示意图在窄屏仍保持原样（模型插槽维持 2 列网格），未形成 brief 描述的『重构』，故 mobileTransformation 未通过，status 定为 revise。","checks":{"projectFit":{"passed":true,"evidence":"Hero 标题『同一个需求，交给可插拔的模型矩阵去调度』、ROUTING BOARD 签名示意图直接落位所选 model-routing-map 方向；章节顺序 routing-hero→model-dial→cloud-runtime→team-task-surfacing→browser-entry→offline-deploy 与 design-brief.json sectionSequence 完全对应，6 节处于 designContract 的 4–7 区间内。"},"evidenceBound":{"passed":true,"evidence":"模型中 GLM/Kimi/MiniMax/Qwen/DeepSeek 均来自 material.json summary.localizations.highlights 与 verifiedClaims 的多模型集成证据并仅作列式示意；构建/测试/预览三支来自 readme『每个任务在真实服务端环境运行、构建测试预览在云端』证据；browser-entry 对应 readme 在线即用/无需本地配置，offline-deploy 对应 AGPL-3.0 与私有离线部署；TypeScript、AGPL-3.0 与品牌名均取自事实，未引入额外数字。"},"signatureVisual":{"passed":true,"evidence":"zh-CN.html:52-79 与 en.html:91-117 的 .diagram『ROUTING BOARD 图 01』含需求路由干道、GLM 高亮选中节点、Kimi/MiniMax/Qwen/DeepSeek 插槽及构建/测试/预览三条分支，附真实 SVG 连线和 role=img aria-label，是派生的流程关系图而非冒充产品截图，signatureVisual 真实存在。"},"mobileTransformation":{"passed":false,"evidence":"@media(max-width:620px) 仅将 flow-grid/meta-row/wide-checklist 折为一列，签名 .diagram 与模型插槽在窄屏未按 brief 收拢为纵向单干线、未将节点改造为横向可横滑的一行插槽，54px 规整网格底纹也未被降为细横线（见两文件 @media(max-width:620px) 块），未构成 brief 明确的移动端重构。"},"localeParity":{"passed":true,"evidence":"zh-CN.html 与 en.html 结构、章节数、区段编号、标签齐全等价；正文各自自然成文，品牌 MonkeyCode 及技术/模型专名（GLM、Kimi、MiniMax、Qwen、DeepSeek、TypeScript、AGPL-3.0）保留不译，符合 localeContract；qa-report.json 显示两语言均 0 consoleErrors、0 假控件、无 broken images。"},"antiTemplate":{"passed":true,"evidence":"签名视觉是专有的路由机制示意图而非通用 Hero；flow-grid 的 B/T/P 三步为 build/test/preview 证据驱动的流程段而非指标计数卡；整页延续同一暖纸制图语言，未出现浏览器/终端/手机渲染外壳，背景为普通细线条网格而非发光网格/粒子，未回退通用落地页模板。"},"noFakeUi":{"passed":true,"evidence":"slot.active 与 dial.hot 仅用边框/底色表达选中态，.dial/.slot/.meta/.branch 均为静态标签与 role=img 示意图，无模型下拉框、开关或同步设置等可交互假控件；qa-report.json 两语言两视口 fakeControlCount 均为 0。"}},"issues":[{"code":"mobile-transformation-not-implemented","severity":"warning","locale":"all","description":"手机端未按 design-brief.json 的 mobileTransformation 将扇形路由矩阵收拢为纵向单干线：要求的『模型节点横向平铺可横滑一行插槽』未实现（窄屏下模型插槽仍是 2 列网格，且无横向滑动），要求的『图纸底纹降为细横线』也未实现（54px 规整网格在移动端保持原样），签名示意图在窄屏仅被动重排，未形成 brief 明确的移动端重构。","repairInstruction":"在 @media(max-width:620px) 中针对 .diagram 做重构：把模型插槽所在容器改为横向（overflow-x:auto + flex 行）可横滑的一行，将需求标签保留顶部、三条分支纵排，并把 body::before 或其他承载者改为细横线底纹；保持制图身份与不伪装设备壳约束不变，修改后重新生成截图供视觉复核。"}]}},"after":{"qaPassed":true,"designReviewApproved":true,"qaEvidence":{"zh-CN":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1285,"cjkCount":727,"latinLetterCount":307,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3240,"h1Top":217.015625,"bottomInset":200.375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"6d94bf5de291c85ba2264c7e00fa1a7e77f38f2eb3f034632f5f5ec969e6fe51","middle":"0b9ca52ab42a4c351d645a64066d9bab2a48f8df6408c92c32dd92bb7659131f","bottom":"ef6f70a3e94e8f7c8a33d30146746807e4166a2e92a72414444f6849b4508947"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1285,"cjkCount":727,"latinLetterCount":307,"scrollWidth":390,"clientWidth":390,"scrollHeight":4353,"h1Top":211.21875,"bottomInset":200.5,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"cd37c71c00f6f422c0cababbd921217ab7c50dc8fbb4cdf8728d2100ff40ba4d","middle":"fb0e90ec44e536516f3f9e0338bc5dd4ee5db2a841f0ac7e5a3a6fd37c5779c0","bottom":"8bf9c6800a89ec77e6aa709dede1ed766568c962fb4e996481fdfb4b5a7611b9"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":2942,"cjkCount":0,"latinLetterCount":2349,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3513,"h1Top":217.015625,"bottomInset":200.484375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"38367d8b1f2310b35d1e7e08c0c8ac25be44eb2a94d120c84d5f4b93bd715756","middle":"85aacf0a54d44df676198c97fee37dcd76443bb0a8cdfead9d7e7cd4dcc49bb7","bottom":"a4ebe0267a824d13f26522f1100b3c38aa2dc70e46af94fac55e27cdec387aa9"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":2942,"cjkCount":0,"latinLetterCount":2349,"scrollWidth":390,"clientWidth":390,"scrollHeight":4832,"h1Top":211.21875,"bottomInset":200.265625,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"fa85df220098083f816b7d0769c44182ad55886d92656470e2caf112b4d59f37","middle":"f5ed0792c5e6ae1e6e29dab053b740fc940792eb80b8a69fbb9684169b511597","bottom":"6389c7cb5ad8c90e7cd825ef0475ea35c4c8e9d0daae7eb1e4242e30c8c3f76c"},"consoleErrors":[],"networkRequests":[]}]},"visualReview":{"schemaVersion":"dodot-automated-design-review-v1","status":"approved","summary":"页面源码忠实实现了所选方向 model-routing-map 的专属 Brief：Hero 的 ROUTING BOARD 提供需求路由干道→GLM(active) 模型插槽→构建/测试/预览三分支的路由矩阵签名视觉，是真实流程关系图而非冒充截图；六个章节顺序与 Brief.sectionSequence 完全一致；暖纸白 #f6f1e7/墨线 #1d2733/绯红干道/靛蓝节点/青黄绯分支的印染制图配色落地；手机 620px 断点把扇形矩阵重构为横向可横滑插槽行 + 纵排三分支 + 细横线底纹，真正形成 brief 所述的移动端重排；双语章节逐节等价、译文自然，MonkeyCode/模型名/AGPL-3.0/TypeScript 保留原文；无通用 Hero/三卡墙/卡片墙/设备壳/发光网格/假控件。所有事实均有材料证据，未新增数字。","checks":{"projectFit":{"passed":true,"evidence":"sectionSequence=[routing-hero,model-dial,cloud-runtime,team-task-surfacing,browser-entry,offline-deploy] 与 design-brief.json.sectionSequence 一致（两端 HTML data-dodot-section 均存在）；compositionFingerprint（需求入→模型节→构建测试预览三支分流 + 整页暖底制图）在 Hero diagram 与 cloud-runtime flow-grid 中实现；hero 文案『把需求交给可插拔的模型矩阵去调度』贴合所选机制。"},"evidenceBound":{"passed":true,"evidence":"模型节点 GLM/Kimi/MiniMax/Qwen/DeepSeek 取自 material.json.summary.evidence 『集成了 GLM、Kimi、MiniMax、Qwen、DeepSeek 等主流模型』并仅作列式示意；构建→测试→预览三支来自『云端开发环境，每个任务在真实服务端环境运行，构建、测试和预览都在云端完成』；browser-entry 基于『无需下载客户端或配置本地环境，打开浏览器即可开始』；offline-deploy 基于『AGPL-3.0 许可证，支持私有离线部署』。未引入额外数字或伪能力。"},"signatureVisual":{"passed":true,"evidence":"zh-CN.html:279-303 与 en.html:216-241 的 Routing Board 实现完整签名视觉：5 个 model-slot 插槽（GLM 为 slot.active 高亮选中）、route-label『需求路由干道』为入口、SVG 箭头连线、.branches 提供构建(B/青)/测试(T/黄)/预览(P/绯) 三分支；节点与分支为流程关系示意，role=\"img\" aria-label，不冒充截图。"},"mobileTransformation":{"passed":true,"evidence":"两文件 @media(max-width:620px) 中：.model-slots 由 2 列网格改为 flex row + overflow-x:auto（横向可横滑的一行插槽）、.slot flex:0 0 auto 白不换行、.branches 改为 column 纵排 B→T→P、body::before 背景仅留 100% 54px 横向细线，符合 Brief.mobileTransformation 的『需求顶部、节点横向可横滑插槽一行、三分支纵排构建→测试→预览、底纹降为细横线』，构成真正重构。"},"localeParity":{"passed":true,"evidence":"zh-CN 与 en 结构逐节对应（6 section、sec-num 01-05、flow-step、meta-row、wide-checklist、dial 完全平行）；品牌 MonkeyCode、模型名 GLM/Kimi/MiniMax/Qwen/DeepSeek、AGPL-3.0、TypeScript 在两端保留原文，符合 localeContract；en 文本为 zh 的自然扩展而非错位；qa-report.json 显示两语言各视口 0 consoleErrors、0 假控件、无 broken images。"},"antiTemplate":{"passed":true,"evidence":"签名视觉为专属路由矩阵制图而非通用 Hero/三卡卡片墙；flow-grid 的 B/T/P 是证据驱动的 build/test/preview 流程段而非指标计数卡；整页延续同一暖纸制图语言；无浏览器/终端/手机设备外壳；背景为扁平工程网格 body::before 而非发光网格/粒子，未回退通用落地页模板。"},"noFakeUi":{"passed":true,"evidence":"slot.active 与 dial.hot 仅以边框/底色/投影表达选中态，.slot/.dial/.meta/.branch 均为静态标签且 diagram 为 role=img 示意，无模型下拉框/开关/同步等可交互假控件；qa-report.json 两语言两视口 fakeControlCount=0；链接仅 homepageUrl 与 GitHub 真实地址。"}},"issues":[]}},"changedFiles":["artifact/zh-CN.html","artifact/en.html"]}],"provenance":{"executor":"opencode","model":"deepseek/deepseek-v4-flash","opencodeVersion":"1.1.58","promptVersion":"dodot-showcase-deepseek-v4","inputHash":"7389ac59ef3b723ed8bffcba5087d084b61aaf32d22c384fccb28e99fbcd92da","artifactHash":"ae7a8bd383213a0f8279479e0fd81104e8b47116c5df0489d0a67d8c843c4c11","designSkill":"dodot-showcase-design","historicalSeed":{"source":"git_head","sourceCanonicalKey":"github:chaitin/monkeycode","assetId":"HEAD:deepseek-v2/chaitin-monkeycode","sourceSchemaVersion":"v1","sourceGeneratorVersion":"anthropic:claude-fable-5","sourceArtifactHash":"1276e8c4b0c1cd31ade01faf89c4749aa7670276a3f3d330cd3e51c8157df578","disposition":"repair_or_regenerate","seedPayloadHash":"55f11ca84f9fe2460e74efe682faee5f7816ac25ac534f4f82b9c449d66ba720"}},"qaEvidence":{"zh-CN":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1285,"cjkCount":727,"latinLetterCount":307,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3240,"h1Top":217.015625,"bottomInset":200.375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"6d94bf5de291c85ba2264c7e00fa1a7e77f38f2eb3f034632f5f5ec969e6fe51","middle":"0b9ca52ab42a4c351d645a64066d9bab2a48f8df6408c92c32dd92bb7659131f","bottom":"ef6f70a3e94e8f7c8a33d30146746807e4166a2e92a72414444f6849b4508947"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1285,"cjkCount":727,"latinLetterCount":307,"scrollWidth":390,"clientWidth":390,"scrollHeight":4353,"h1Top":211.21875,"bottomInset":200.5,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"cd37c71c00f6f422c0cababbd921217ab7c50dc8fbb4cdf8728d2100ff40ba4d","middle":"fb0e90ec44e536516f3f9e0338bc5dd4ee5db2a841f0ac7e5a3a6fd37c5779c0","bottom":"8bf9c6800a89ec77e6aa709dede1ed766568c962fb4e996481fdfb4b5a7611b9"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":2942,"cjkCount":0,"latinLetterCount":2349,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3513,"h1Top":217.015625,"bottomInset":200.484375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"38367d8b1f2310b35d1e7e08c0c8ac25be44eb2a94d120c84d5f4b93bd715756","middle":"85aacf0a54d44df676198c97fee37dcd76443bb0a8cdfead9d7e7cd4dcc49bb7","bottom":"a4ebe0267a824d13f26522f1100b3c38aa2dc70e46af94fac55e27cdec387aa9"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":2942,"cjkCount":0,"latinLetterCount":2349,"scrollWidth":390,"clientWidth":390,"scrollHeight":4832,"h1Top":211.21875,"bottomInset":200.265625,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"fa85df220098083f816b7d0769c44182ad55886d92656470e2caf112b4d59f37","middle":"f5ed0792c5e6ae1e6e29dab053b740fc940792eb80b8a69fbb9684169b511597","bottom":"6389c7cb5ad8c90e7cd825ef0475ea35c4c8e9d0daae7eb1e4242e30c8c3f76c"},"consoleErrors":[],"networkRequests":[]}]}},"generatorVersion":"opencode-1.1.58:deepseek/deepseek-v4-flash:dodot-showcase-deepseek-v4","publishedAt":"2026-08-20T12:28:43.587Z"},"rawExtraction":{"projectRevisionId":"cmt2zgz4k00am6271cot1zny7","projectVersion":1,"sources":[{"canonicalKey":"github:chaitin/monkeycode","role":"primary","revisionVersion":1,"url":"https://github.com/chaitin/MonkeyCode"}],"facts":[{"key":"repository_owner","value":"chaitin","evidence":[{"path":"repository.owner","excerpt":"chaitin"}]},{"key":"repository_name","value":"monkeycode","evidence":[{"path":"repository.name","excerpt":"monkeycode"}]},{"key":"description","value":"AI coding platform for teams","evidence":[{"path":"repository.description","excerpt":"AI coding platform for teams"}]},{"key":"primary_language","value":"TypeScript","evidence":[{"path":"repository.language","excerpt":"TypeScript"}]},{"key":"license","value":"AGPL-3.0","evidence":[{"path":"repository.license","excerpt":"AGPL-3.0"}]},{"key":"topics","value":"ai, ai-agent, ai-coding, ai-coding-assistant, aicoding, claude, codex, coding, cursor, opencode, vibe-coding","evidence":[{"path":"repository.topics","excerpt":"ai, ai-agent, ai-coding, ai-coding-assistant, aicoding, claude, codex, coding, cursor, opencode, vibe-coding"}]},{"key":"source_count","value":"1","evidence":[{"path":"projectRevision.sources","excerpt":"github:chaitin/monkeycode"}]}],"taxonomy":{"category":"development_tools","subcategory":"data_ai_tools","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI coding platform for teams"}]},"lifecycle":{"stage":"active","confidence":0.75,"evidence":[{"path":"repository.pushedAt","excerpt":"2026-08-17T13:14:27.000Z"}]},"userIntents":[{"slug":"build_software","label":"构建或改进软件","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI coding platform for teams"}]},{"slug":"automate_work","label":"自动化工作流程","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI coding platform for teams"}]},{"slug":"create_content","label":"创作内容或设计","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI coding platform for teams"}]},{"slug":"analyze_data","label":"分析与理解数据","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI coding platform for teams"}]},{"slug":"learn_explore","label":"学习与探索","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI coding platform for teams"}]}],"readme":{"headings":["MonkeyCode","What Is MonkeyCode?","Screenshots","Features","Usage","Online","Self-Hosted Deployment","Comparison","Community and Support","Star History","License"],"commands":[]},"changeSet":[{"field":"authorName","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"fa797acc3e838ba106d51222f4cdeba19e911db0a9801804240db7c4841ec688"},{"field":"description","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"7486f28beac8dd078d3795918ec1b591837f0d36a0be40c2ab6677f7e1b520b2"},{"field":"experienceUrl","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"f6a470cc75f581b090d290ef9d6674ba691e6cd61dd51d99981b3026cd2ab946"},{"field":"name","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"757bc496456499aaca68d40e7d2c94833db7c5417cc0a9a48cd83f471d1354aa"},{"field":"repository","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"1ca4257f729f07f63c0abc6afa8703969551d5200b8e691a0faea1ab90c93dc4"},{"field":"sourceCount","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"6b86b273ff34fce19d6b804eff5a3f5747ada4eaa22f1d49c01e52ddb7875b4b"}]},"publishedAt":"2026-08-10T15:20:32.447Z","updatedAt":"2026-08-21T13:26:48.834Z","summary":{"id":"cmt2zgz4p00ao6271c3sjf7h6","projectType":"product","oneLine":"面向团队的开源 AI 编码平台","overview":"MonkeyCode 是一个开源的企业级 AI 开发平台，提供云端开发环境、AI 模型管理、任务管理和需求管理。无需本地配置，浏览器即可使用，支持主流模型和移动端，适合团队协作与私有部署。。面向团队的开源 AI 编码平台。多模型支持。原生移动端","highlights":["云端开发环境","多模型支持","原生移动端","私有部署"],"bestFor":["需要团队协作的研发团队","希望快速体验 AI 编码的用户","有数据隐私要求的企业"],"caveats":["不提供本地 IDE 或 CLI","无代码补全功能","部署需要较高硬件配置"],"categorySuggestion":"development_tools","subcategorySuggestion":"coding_agents","lifecycleStages":["idea_requirements","frontend","backend_data","testing_quality"],"audienceIntents":["discover","evaluate","build","operate"],"confidence":0.8,"evidence":[{"claim":"MonkeyCode 是一个开源的企业级 AI 开发平台，内置开发环境管理、AI 模型管理、AI 任务管理和项目需求管理。","sourceType":"readme","locator":"https://github.com/chaitin/MonkeyCode"},{"claim":"支持在线使用，无需下载客户端或配置本地环境，打开浏览器即可开始 AI 开发任务。","sourceType":"readme","locator":"https://github.com/chaitin/MonkeyCode"},{"claim":"提供云端开发环境，每个任务在真实的服务端环境中运行，构建、测试和预览都在云端完成。","sourceType":"readme","locator":"https://github.com/chaitin/MonkeyCode"},{"claim":"集成了 GLM、Kimi、MiniMax、Qwen、DeepSeek 等主流模型，可按任务类型切换或手动选择。","sourceType":"readme","locator":"https://github.com/chaitin/MonkeyCode"},{"claim":"支持原生移动端，iOS 和 Android 深度支持，PC 与移动端数据同步。","sourceType":"readme","locator":"https://github.com/chaitin/MonkeyCode"},{"claim":"核心代码在 GitHub 上公开，采用 AGPL-3.0 许可证，支持私有离线部署。","sourceType":"readme","locator":"https://github.com/chaitin/MonkeyCode"}],"publishedAt":"2026-08-17T13:56:42.862Z"},"github":{"url":"https://github.com/chaitin/MonkeyCode","ownerAvatarUrl":"https://avatars.githubusercontent.com/u/7302766?v=4","starsCount":4435,"forksCount":665,"language":"TypeScript","syncedAt":"2026-08-17T13:53:14.187Z","license":"AGPL-3.0","defaultBranch":"main","archived":false,"disabled":false,"lastChangedAt":"2026-08-17T13:14:27.000Z","missing":false},"tags":[{"name":"Vibe coding","slug":"vibe-coding","verification":"verified","evidenceUrl":"https://github.com/chaitin/MonkeyCode","evidenceExcerpt":"GitHub topic: vibe-coding"}],"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"]