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:T770b, VibeETL · 自托管可视化 ETL 平台

vibe ETL

自托管 · 可视化 ETL 平台

用拖拽，把数据铺成一条管道

在一张可交互的 React Flow DAG 画布上，把 Extract → Transform → Load 的每个步骤当作节点闸门装配起来；数据包随每道闸门被塑形一次，最终从双出口岔道落地。

MIT 许可主语言 Python 内存执行 · Polars

前往仓库 → 读一读这条管道 →

ACTOR: RAW // FORGED AT GATE // LANDED TRACK: IN-MEMORY · POLARS

01 / 闸门解剖

六类节点闸门，挂在一条 I/O / 云 / 准备 / 变换 / 连接 / 分析的索引轨上

管道不是凭空而立的。每个节点是一道可装配的闸门，从输入到云集成、从准备与变换到连接与分析，按类别分工，拖进画布即可连成流水线。

节点闸门面板 · GATE PANEL

I/O输入 / 输出input

in取数，out落盘，作为管道的首尾两道闸门。

CLD云集成cloud

对接云服务的数据接入，部分功能可能尚未完全就绪。

PRE数据准备prepare

清洗、补缺、标准化，为后续变换做铺垫。

XFR变换transform

重塑列、字段计算，数据在这里被真正塑形。

JN连接join

横向联结多张子表，拼合并行数据流。

ANA分析analysis

聚合、统计与交互式可视化，把结果摊开在眼前。

02 / 执行引擎

Polars 在内存里跑，数据包沿时间刻度高速穿过

画布负责编排，真正的搬运交给 Polars。管道并不落盘中转，而是在内存中整段执行，数据包像列车一样贴着时刻表前进。

EXTRACT

装车

TRANSFORM

塑形

shape

LOAD

落地

out

交互式 DAG 画布（React Flow）负责编排，Polars 引擎在内存中执行数据管道——编排与执行分属两层，这正是管道能贴着时刻表跑出来的原因。

03 / 落地格式

双出口分流：JSON 存档，YAML 交给 AI 代理

到了轨道末端，管道分岔为两股出口——一条把工作流原样存档以便随时召回，另一条把同一条管道导出成 AI 代理能直接读懂的 YAML。

JSON · 保存 / 加载

整条工作流可保存成 JSON 文件，也能从 JSON 加载回来继续编辑。掉线不丢，改到一半也能存档再续。

flow = save(load:source).workflow
{"nodes":[...],edges:[...]}

YAML · AI 代理导出

把同一管道导出为 YAML，供 AI 代理消费。人拖出来的流程，机器也能接着读懂并继续编排。

gates:
  - name: extract
  - name: transform
  - name: load

04 / 企业特性

让长管道靠得住的能力徽记

规模化使用需要的不是花哨，而是恢复、多开与识别的底气——三项能力沿同一段能力轨排列，让复杂流水线也能被信任。

CAP 01

自动恢复 · Auto-recovery

工作流异常中断后自动恢复，尽量不让你重头来过。

CAP 02

多标签工作区 · Multi-tab

同一个画布内开多个工作区，并行维护数条管道。

CAP 03

语义类型检测 · Semantic types

自动识别字段语义类型，让变换节点选得更省心。

05 / 边界与许可

站到管道旁，看清它的边界与来源

安装前提

需要本地安装与配置，非免安装托管。
pip install -r requirements.txt
npm install + npm run dev

依赖边界

云集成节点部分功能可能未完全就绪。
AI 节点依赖外部 API。

许可

MIT License · 可自由使用与二次开发。
主语言：Python

来源 · github.com/cardchase/vibeetl · description: self-hosted, lightweight, visual ETL platform inspired by Alteryx

23:T778b, VibeETL · Self-hosted visual ETL platform

vibe ETL

Self-hosted · Visual ETL platform

Drag data into a pipeline on a canvas

On an interactive React Flow DAG canvas, assemble every Extract → Transform → Load step as a node gate. Each gate reshapes the packet once — outline, fill and size — and the result leaves through the twin export branch at the end.

MIT License Primary language Python In-memory via Polars

Visit repository → Read the pipe →

ACTOR: RAW // FORGED AT GATE // LANDED TRACK: IN-MEMORY · POLARS

01 / Gate anatomy

Six node gates, indexed on one rail: I/O / cloud / prepare / transform / join / analysis

A pipe is not built in mid-air. Every node is a gate you can assemble — from input to cloud integration, from preparation and transform to join and analysis — dragged onto the canvas as its own step.

Node gate panel

I/OInput / Outputinput

inreads, outwrites — the first and last gates of the pipe.

CLDCloud integrationcloud

Connect data from cloud services; some of these may be only partially ready.

PREData preparationprepare

Clean, fill and standardize before the real transformation happens.

XFRTransformtransform

Reshape columns and run field math — the data is genuinely shaped here.

JNJoinjoin

Combine tables laterally and merge parallel data streams.

ANAAnalysisanalysis

Aggregate, summarize and visualize results interactively.

02 / Execution engine

Polars runs it in-memory; the packet crosses the track on schedule

The canvas orchestrates, Polars does the hauling. The pipe never stops to spool to disk — the whole run happens in memory, and packets move like trains against a timetable.

EXTRACT

load

TRANSFORM

shape

LOAD

land

out

An interactive DAG canvas (React Flow) handles orchestration, while the Polars engine executes the pipeline in-memory — two distinct layers that keep the pipe running on time.

03 / Landing formats

Twin off-ramps: JSON to archive, YAML for AI agents

At the end of the track the pipe forks into two exits — one archives the workflow as-is so it can be recalled anytime, the other exports the very same pipe as YAML an agent can read at once.

JSON · Save / load

Ships the whole workflow to a JSON file and reloads it from JSON to keep editing — nothing lost on a crash, and a half-finished pipe can be archived and resumed.

flow = save(load:source).workflow
{"nodes":[...],edges:[...]}

YAML · AI agent export

Exports the identical pipeline to YAML for AI agents to consume — the flow a human dragged out, a machine can pick up and keep assembling.

gates:
  - name: extract
  - name: transform
  - name: load

04 / Enterprise features

Capabilities badges for long runs you can trust

At scale you need recovery, parallel workspaces and recognition, not flash. Three capabilities ride the same capability rail so complex pipes can be relied on.

CAP 01

Auto-recovery

Workflows recover automatically after an interruption, so you rarely start over.

CAP 02

Multi-tab workspaces

Open several workspaces in one canvas and maintain multiple pipes side by side.

CAP 03

Semantic type detection

Field semantic types are detected so transform nodes are easier to choose.

05 / Boundaries & license

Stand beside the pipe and see its limits and provenance

Setup

Requires local installation and configuration — not a hosted service.
pip install -r requirements.txt
npm install + npm run dev

Dependency boundaries

Cloud integration nodes may be only partially ready.
AI nodes depend on external APIs.

License

MIT License — free to use and to rework.
Primary language: Python

Source · github.com/cardchase/vibeetl · a self-hosted, lightweight, visual ETL platform inspired by Alteryx

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","cmt2zgyzv009g6271lwpj187t:zh-CN:beautified",{"project":{"id":"cmt2zgyzv009g6271lwpj187t","slug":"cardchase-vibeetl","name":"vibeetl","description":"用拖拽构建数据管道","authorName":"cardchase","authorVerification":{"verified":false,"verifiedAt":null},"category":"development_tools","subcategory":"data_ai_tools","experienceUrl":null,"qualityScore":60,"featured":false,"officialSite":null,"views":{"original":{"available":true},"official":{"available":false,"reason":"disabled_by_policy"},"beautified":{"available":true,"reason":null},"defaultView":"beautified"},"beautifiedView":{"revisionId":"cmt2zgz0g009s6271c5q7ls78","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":"packet-rail-pipeline","name":"数据包沿轨道流动的管道","projectMetaphor":"一条横贯画面的数据管道，数据包在节点闸门之间沿轨道移动：进入(Input)、变换(Transform)、装载(Load)，每个节点是一次物理闸门，边缘是承载数据包滑行的轨道段。管道即工作流本身。","signatureVisual":"水平管道轨道上依次排列三个节点闸门，一枚带批号刻度的数据包从左向右滑过闸门，每过一道闸门数据包的形态(轮廓/填充)改变一次，末尾分流出可导出的 YAML 信标。无任何设备或浏览器外壳。","heroGeometry":"横向三段式管道轨(左右贯穿整个第一屏)，中心偏左一枚数据包动画沿轨推进，三个节点闸门像关卡一样矗立，包体在各闸门处渐进变形。","sectionSequence":["管道总览：一段式拖拽流水线（Extract → Transform → Load 单轨往返）","闸门解剖：节点面板按类别分组（I/O / 云 / 准备 / 变换 / 连接 / 分析）的索引轨","执行引擎：Polars 内存执行，数据包高速穿过轨道的时间刻度线","落地格式：JSON 保存加载 + YAML 导出给 AI 代理的双出口分流","企业特性：自动恢复 / 多标签 / 语义类型检测的能力徽记轨","边界与许可：安装要求、云/AI 节点依赖与 MIT 许可以及来源colophon"],"surfaceModel":"纸面像铁路蓝图：轨道为硬质发丝线，闸门为半透明卡槽，数据包为带刻度的实心小体块；整个页面如同一张平铺的水文/运输图，章节之间以轨道段与信号灯为分隔。","palette":"冷钢蓝主轴 + 琥珀信号黄作为数据包运动色 + 石墨灰轨道；背景为暖纸白，强调色仅出现在运动与高亮节点上。","typography":"gros/加粗等宽字体用于批号与刻度，正文用高对比无衬线；大标题保持单行清晰的可读节奏。","desktopComposition":"1280 宽采用横向叙事：顶部一行定位语与 CTA，主屏为左右贯穿的轨道图，其后每节是一段新轨道或信号岔道，读起来像沿流水线逐站前行。","mobileTransformation":"390 宽将横轨转折为纵向之字形流水线：数据包改为自上而下逐节点下沉，闸门竖向堆叠，时间刻度转为每段侧边的小标尺；拖动隐喻保留为逐段展开。","mediaStrategy":"无本地素材，删除媒体区；用管道轨道、数据包与闸门构成的纯排版、流程与关系来形成标志视觉，不伪造产品截图。","forbiddenPatterns":["三卡指标横幅","浏览器/终端/编辑器外壳","发光网格或粒子背景"],"compositionFingerprint":"horizontal-packet-rail-through-node-gates-with-shaped-payload-and-twin-export-branch"},{"id":"workbench-slot-racks","name":"数据工程工作台插槽","projectMetaphor":"一台用于组装数据管线的物理工程台：节点是插进机架卡槽的模块卡，数据以信号带在模块之间接线穿梭，画布是一块可拖动的接线背板。台上的每件模块都代表一个可拼装的数据步骤。","signatureVisual":"俯视/微倾的工程台桌面：一排并置的模块卡(每卡一角标类别徽章)，模块之间以带箭头的信号线相连，一条高亮的当前信号带正从源卡流向变换卡再流向装载卡；另有一只空卡槽表示可拖入新节点。","heroGeometry":"斜俯视桌面背板：多块模块卡按网格离散摆放，信号线以贝塞尔弧在卡间穿梭，一条主信号带较粗并以光点沿程推进，桌面一角留出一个拖拽占位卡槽。","sectionSequence":["工程台总览：把节点卡插上背板即可组成流水线","模块库：六类节点卡(输入/云/准备/变换/连接/分析)的接插分布盘","执行引擎：Polars 内存计算，接线板上的高速跑批信号带","可保存交换：JSON 工作流文件 + YAML 交给 AI 代理的双接口","工程台特性：自动恢复 / 多标签 / 语义类型检测的架层清单","来源与许可：安装前提、依赖边界、MIT 的colophon"],"surfaceModel":"台面式：模块卡带轻微抬起的实体感与窄阴影，信号线在半透明背板上显影；章节之间以桌面台位的划分和架的层高来分隔，而非统一卡片网格。","palette":"机革灰卡体 + 淬火蓝信号带 + 类别各自一枚细徽章色；背景为浅暖灰，强调色跟随信号流向而非点缀。","typography":"模块卡编号用紧凑等宽，标题与正文用清晰无衬线，信号线上的批号以小型等宽标注。","desktopComposition":"1280 宽为左上至右下的台面布局：定位语与 CTA 位于左上台沿，背板主视觉占据中心偏右，后续章节作为桌面上的各个工具抽屉依次展开。","mobileTransformation":"390 宽将台面背板折成纵向接插带：模块卡从网格改为纵向层叠插入，信号线变作连接卡与卡之间的小竖线，空卡槽保留在底部作为拖入入口。","mediaStrategy":"无本地素材，删除媒体区；以模块卡、信号线与空卡槽构成的装配关系作为标志视觉，不伪造产品截图。","forbiddenPatterns":["功能卡片墙模板","编辑器/仪表盘外壳","用渐变填充代替装配关系"],"compositionFingerprint":"slant-desk-module-slots-with-interleaving-signal-bends-and-empty-drag-cavity"},{"id":"schema-folding-lattice","name":"数据成型折纸格","projectMetaphor":"把数据一行行、一列列折成新形状的手工成型：变换不是运输而是塑造——列进入漏斗被折叠、筛选、联结、重塑，最终展开成新的可交付数据集。画布是一块展平后再被折起的格纸。","signatureVisual":"一条由细格排列成的数据表带，从左端进入后被若干折叠面(褶皱)逐段改形：行在此收紧、他处拆分重连，末端展开成新的表带并落下一枚 YAML 折叠标记。褶皱的方向与密度表达变换强度。","heroGeometry":"深浅相间的水平格带经过若干次垂直/斜向折叠点，折叠带来节奏变化；一条光标沿格子的行径移动，象征画布上的拖拽重排。","sectionSequence":["成型总览：一行进入、经过折叠、以新形态落地的数据旅程","折叠工具：六类节点(I/O / 云 / 准备 / 变换 / 连接 / 分析)作为不同折叠手法的图例","执行引擎：Polars 内存运算，折叠在极高速率下一次成型的速率计","可装配交换：JSON 记录折叠过程 + YAML 导出给 AI 代理的折痕留档","企业特性：自动恢复 / 多标签 / 语义类型检测的成型质检清单","边界与许可：安装前提、依赖与 MIT 的来源colophon"],"surfaceModel":"折纸格面：数据为细格纸带，折叠面以半透明褶皱呈现透光层次，背景为微格纸；章节以纸面折痕和落点来分区，而非浮起卡片。","palette":"纸白底 + 墨黑细格 + 单一丹砂折叠色用于标记折点；强调色克制，整套视觉偏工艺印刷。","typography":"数据格内用等宽数字/批号，标题正文用高对比印刷体；折叠手法的说明用齐整的对齐排版。","desktopComposition":"1280 宽左右分栏叙事：左侧为纵向数据旅程(表带自上而下)，右侧为折叠手法的图例与工具箱图谱，中间以一条折痕主轴贯通到底。","mobileTransformation":"390 宽将左右分栏并合为单一纵向纸带：表带与折叠图例逐节自上而下滚动相遇，折痕作为每节的章头，折叠节奏以连续纸面呈现。","mediaStrategy":"无本地素材，删除媒体区；用数据表带的折叠、筛选与联结过程作为标志视觉，不伪造产品截图。","forbiddenPatterns":["营销 Hero + 指标卡","终端/浏览器外壳","粒子或网格装饰当视觉主题"],"compositionFingerprint":"lattice-tape-folded-through-shear-folds-into-unfolded-deliverable-with-yaml-crease"}],"selectedDirectionId":"packet-rail-pipeline","selectionRationale":"三个方向在隐喻、标志视觉与 Hero 几何上真正不同：方向一是横向管道运输，方向二是斜俯视台面装配，方向三是折纸成型。选中「数据包沿轨道流动的管道」，因为它最直接地把本项目的核心机制(拖拽式 DAG 画布 + 内存执行 + 输入/变换/装载)呈现为一条可读的、证据驱动的物理流程，而非营销 Hero。数据包沿轨滑过闸门并随变换改变形态，恰好表达可视化 ETL 的『数据被逐节点塑形』这一事实；Polars 内存执行对应轨道的高速信号，JSON/YAML 双导出对应末端的双岔道——每一节都映射到 material.json 中可核验的事实。横向轨道在 1280 宽上提供单向叙事纵深，在 390 宽转折为纵向之字形，空间决策随视口真正重构。该方向未落入反模板上下文的任何指纹，也完全避开测量卡、功能卡墙、设备/终端外壳与粒子网格。"},"designBrief":{"schemaVersion":"dodot-design-brief-v1","selectedDirectionId":"packet-rail-pipeline","projectMetaphor":"一条横贯画面的数据管道，数据包在节点闸门之间沿轨道移动：进入(Input)、变换(Transform)、装载(Load)，每个节点是一次物理闸门，边缘是承载数据包滑行的轨道段。管道即工作流本身。","signatureVisual":"水平管道轨道上依次排列三个节点闸门，一枚带批号刻度的数据包从左向右滑过闸门，每过一道闸门数据包的形态(轮廓/填充)改变一次，末尾分流出可导出的 YAML 信标。无任何设备或浏览器外壳。","compositionFingerprint":"horizontal-packet-rail-through-node-gates-with-shaped-payload-and-twin-export-branch","heroGeometry":"横向三段式管道轨(左右贯穿整个第一屏)，中心偏左一枚数据包动画沿轨推进，三个节点闸门像关卡一样矗立，包体在各闸门处渐进变形。","sectionSequence":["管道总览：一段式拖拽流水线（Extract → Transform → Load 单轨往返）","闸门解剖：节点面板按类别分组（I/O / 云 / 准备 / 变换 / 连接 / 分析）的索引轨","执行引擎：Polars 内存执行，数据包高速穿过轨道的时间刻度线","落地格式：JSON 保存加载 + YAML 导出给 AI 代理的双出口分流","企业特性：自动恢复 / 多标签 / 语义类型检测的能力徽记轨","边界与许可：安装要求、云/AI 节点依赖与 MIT 许可以及来源colophon"],"surfaceModel":"纸面像铁路蓝图：轨道为硬质发丝线，闸门为半透明卡槽，数据包为带刻度的实心小体块；整个页面如同一张平铺的水文/运输图，章节之间以轨道段与信号灯为分隔。","palette":"冷钢蓝主轴 + 琥珀信号黄作为数据包运动色 + 石墨灰轨道；背景为暖纸白，强调色仅出现在运动与高亮节点上。","typography":"gros/加粗等宽字体用于批号与刻度，正文用高对比无衬线；大标题保持单行清晰的可读节奏。","desktopComposition":"1280 宽采用横向叙事：顶部一行定位语与 CTA，主屏为左右贯穿的轨道图，其后每节是一段新轨道或信号岔道，读起来像沿流水线逐站前行。","mobileTransformation":"390 宽将横轨转折为纵向之字形流水线：数据包改为自上而下逐节点下沉，闸门竖向堆叠，时间刻度转为每段侧边的小标尺；拖动隐喻保留为逐段展开。","mediaStrategy":"无本地素材，删除媒体区；用管道轨道、数据包与闸门构成的纯排版、流程与关系来形成标志视觉，不伪造产品截图。","evidencePlan":["用于闸门命名的 Extract / Transform / Load 与节点类别(I/O、云、准备、变换、连接、分析)来自 readme 的 Core Philosophy 与 Tool Palette","React Flow DAG 画布 + Polars 内存执行来自 readme 架构描述与 summary.evidence","JSON 保存/加载 + YAML 导出给 AI 代理来自 summary.evidence 的第三/四段 claim"],"localeContract":{"zh-CN":"自托管可视化 ETL 平台——用拖拽构建数据管道：拖拽装配 Extract → Transform → Load，Polars 内存执行，JSON/YAML 双导出。","en":"Self-hosted visual ETL platform — build data pipelines by dragging: assemble Extract → Transform → Load, executed in-memory by Polars, exported as JSON or YAML."},"forbiddenPatterns":["三卡指标横幅","功能卡片墙模板","浏览器/终端/编辑器/仪表盘外壳","发光网格或粒子背景当作视觉主题","伪造产品截图或占位媒体框"]},"visualReview":{"schemaVersion":"dodot-automated-design-review-v1","status":"approved","summary":"页面忠实实现 packet-rail-pipeline 方向：Hero 为横向管道轨（Extract/Transform/Load 三闸门 + 带批号的滑行数据包 + JSON/YAML 双岔道），其后五节依 design-brief.sectionSequence 依次为闸门解剖（I/O/云/准备/变换/连接/分析）、Polars 内存执行时间刻度、JSON/YAML 落地格式双出口、企业特性能力徽记轨、边界与许可来源 colophon。所有文案均有 material.json 依据，signatureVisual 真实存在于 .sig SVG 且数据包随闸门渐进变形，手机 CSS 以独立纵向 sig-v SVG 实现真正重构（数据包自上而下下沉、闸门竖向堆叠、时间刻度转侧边小标尺），双语章节等价且自然，未回退通用 Hero/三卡/卡片墙/设备外壳/装饰渐变网格，无假控件。","checks":{"projectFit":{"passed":true,"evidence":"Hero 为横向三段式管道轨（Extract/Transform/Load 三闸门，中心 Transform 以深钢蓝+琥珀反白凸显，带 W001 批号数据包沿轨滑行，末端 YAML/JSON 双出口岔道），与 design-brief.heroGeometry 与 signatureVisual 高度一致；后续五节（闸门解剖、执行引擎、落地格式、企业特性、边界与许可）与 brief.sectionSequence 逐条对应，无偏离方向的通用营销结构。"},"evidenceBound":{"passed":true,"evidence":"全部叙述可核验：React Flow DAG 画布 + Polars 内存执行（summary.evidence 第2条）、六类节点闸门 I/O/云/准备/变换/连接/分析（readme Tool Palette 与 summary.evidence 第3条）、JSON 保存/加载 + YAML 导出给 AI 代理（summary.evidence 第3/4条）、自动恢复/多标签/语义类型检测（summary.evidence 第5条）、pip install/npm install/npm run dev（readme.commands）、Python + MIT + 自托管轻量可视化 ETL 援引 Alteryx（description/仓库字段）；安装前提、云集成节点部分未就绪、AI 节点依赖外部 API 等 caveats 亦与 material.json 一致。"},"signatureVisual":{"passed":true,"evidence":"signatureVisual 真实存在于 .sig SVG：viewBox 0 0 1100 210 水平主轨贯穿，Extract/Transform/Load 三闸门作卡槽矗立，带刻度的琥珀数据包以 animateTransform 沿轨运动，并以 animate 在两道闸门处改变 width/height/rx/fill（白描边琥珀→琥珀实心→钢蓝实心）实现「每过闸门轮廓/填充改变一次」，RAW/SHAPED/READY 刻度标注佐证；末端以钢铁/石墨双岔道分流出 YAML 与 JSON。全图仅轨道与数据包，无任何设备/浏览器/终端外壳，非伪产品截图。"},"mobileTransformation":{"passed":true,"evidence":"手机版形成真实重构：@media(max-width:720px) 将 .sig 隐藏并用独立 sig-v 纵向 orthographic SVG 替换（viewBox 0 0 400 940），纵向主轨、01/02/03 每段侧边标尺、Extract/Transform/Load 三道闸门竖向堆叠、数据包 animateTransform to 0 560 自上而下逐节点下沉、底部 YAML/JSON 双出口；同时时间刻度区经 .scale 改造（border-left 纵标尺 + .seg top:var(--p) 横移、竖排）转为竖排侧边小标尺，.cat-list/.branch/.cap-train 网格折为单列。方向性与布局随视口重建，非等比缩放。scrollHeight 由 4099 增至 6141 反映纵向叙事成立。"},"localeParity":{"passed":true,"evidence":"zh-CN 与 en 在章节、导航、h1、chips、CTA、SVG 图（aria-label 各自本地化）与五节排版上完全对应且翻译自然达意；en cjkCount=0、zh cjkCount 801/785、lang 标记正确，无残留未译串或生硬机翻。"},"antiTemplate":{"passed":true,"evidence":"未回退通用 Hero+三卡指标横幅或功能卡片墙模板；各节以同一管道/轨道母题贯穿（闸门索引轨、执行时间刻度轨、双出口支线、能力徽记轨、colophon 轨道分隔线），背景为暖纸白+径向细点而非发光粒子网格，无浏览器/终端/编辑器/仪表盘外壳，无装饰性渐变网格充当主题；其构成指纹 horizontal-packet-rail-through-node-gates 未落入 antiTemplateContext 的 recentCompositionFingerprints/recentSignatureVisuals。"},"noFakeUi":{"passed":true,"evidence":"QA 报告 fakeControlCount=0、brokenImageCount=0、missingAltCount=0、emptyLinkCount=0。页面可点交互仅为真实跳转（github.com/cardchase/vibeetl 仓库链接与#gates/#engine/#formats/#features 锚点导航）；branch-card 内 .code 为说明性排版示例（非可点击控件/占位媒体），SVG 用 role=img 与 aria-label 标注为插图而非假 UI。CTA、chips 均为真实依据（MIT/Python/Polars）。"}},"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":5,"textLength":1922,"cjkCount":801,"latinLetterCount":665,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":4099,"h1Top":228.734375,"bottomInset":196.734375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"df8669b6224564c255d699914ad159bbd213eb4065486ba55a6dd2021bfb2fe6","middle":"04a011662522954daec9255f22e1b6a5a495e53dab4a8dd3af7ac880bc72c563","bottom":"f5bed4bbbeb709d1f017401db58d916389ce7260a0da9d9c549a5bcc8d6b063e"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":5,"textLength":1902,"cjkCount":785,"latinLetterCount":665,"scrollWidth":390,"clientWidth":390,"scrollHeight":5149,"h1Top":227.25,"bottomInset":218.765625,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"a3bbfd997d7010da214add081adfa2da0bcc96178ab1e6b4b2c7f0f6deb71a90","middle":"b47bf777f15392f244a96d55fff79b66d36b90e8c559877b4858bbe01acfaee0","bottom":"8b377b1e143c7360fc37ea009a9208b9712e597605fce1e17ead4c199852702c"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":5,"textLength":3816,"cjkCount":0,"latinLetterCount":2988,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":4661,"h1Top":228.734375,"bottomInset":197.3125,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"ded7cb6d8b7301249d6323857ae0c26e42037320a6d953de0a9f2dc8f0a215fa","middle":"d6cbb55dd9a3e3b2fe5fdad6c9967e03885e8c958abbdf1a5ef04e4ff0f6fe51","bottom":"3f6b8c9a04bd8da106af0a53a3c9d4a6d6ea43045a862bc4befdd74fd044fd39"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":5,"textLength":3752,"cjkCount":0,"latinLetterCount":2932,"scrollWidth":390,"clientWidth":390,"scrollHeight":5680,"h1Top":192.59375,"bottomInset":218.09375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"dbd20f9f67cb1f9af8a30ee9b801766d38fb5f37c87a355787c2b75094d6f221","middle":"090d648879f3c2a5839c8d652f117b7be10f0492b6e26556eb859030b3f8dd21","bottom":"8cb540cb65965ac4792aa39a8c1fecdd46048cb70c82e5aac076fe247bae6974"},"consoleErrors":[],"networkRequests":[]}]},"visualReview":{"schemaVersion":"dodot-automated-design-review-v1","status":"revise","summary":"页面整体忠实于 packet-rail-pipeline 方向：英雄区横向三闸门管道、Polars 内存执行时间刻度、JSON/YAML 双出口分流、企业特性徽记轨、边界与许可 colophon 均各就其位；双语等价自然、事实均可在 material/design-brief 中核验，且未落入通用 Hero/三卡/卡片墙/设备外壳/假控件。但存在两处分叉：1) signatureVisual 只以 RAW/SHAPED/READY 文本提示数据包变形，brief 要求的「每过一道闸门数据包轮廓/填充形态改变一次」未实际实现；2) 更关键的 mobileTransformation 不成立——370px 断点仅将网格/列堆叠成单列并将 SVG 等比缩放，未按 brief 将横轨折成纵向之字形流水线（数据包自上而下逐节点下沉、闸门竖向堆叠、时间刻度转为侧边小标尺），手机端未形成真正重构。综上判为 revise。","checks":{"projectFit":{"passed":true,"evidence":"Hero 为横向管道轨（Extract/Transform/Load 三闸门+滑行数据包+JSON/YAML 双岔道），后续五节依次为闸门解剖(I/O/云/准备/变换/连接/分析)、Polars 内存执行时间刻度、JSON/YAML 落地格式双出口、企业特性能力徽记轨、边界与许可来源 colophon，与 design-brief 的 sectionSequence 逐条对应；无偏离方向的通用营销结构。"},"evidenceBound":{"passed":true,"evidence":"每节文案均有材料依据：React Flow DAG 画布+Polars 内存执行(summary.evidence),六类节点闸门(readme Tool Palette),JSON 保存/加载+YAML 导出给 AI 代理(evidence),自动恢复/多标签/语义类型检测(summary.evidence 第五 claim),pip install/npm install/npm run dev(readme.commands),Python+MIT+自我托管/轻量/可视化 ETL 援引 Alteryx(summary/description)，安装前提与云/AI 依赖 caveats 亦与材料一致。"},"signatureVisual":{"passed":true,"evidence":"签名视觉真实存在于 .sig SVG：水平主轨贯穿全宽，Extract/Transform/Load 三闸门矗立体（中心 Transform 以深钢蓝+琥珀反白区别于两侧白底闸门），带批号 W001 的琥珀数据包用 animateTransform 沿轨左右运动，末端以钢铁/石墨双岔道分流 YAML 与 JSON；无任何设备/浏览器/终端外壳。注意：brief 要求数据包「每过一道闸门形态改变一次」，此处仅以 RAW/SHAPED/READY 文本提示变形，包体轮廓/填充未实际逐闸门变化。"},"mobileTransformation":{"passed":false,"evidence":"370px 断点仅将 .cat-list/.branch/.cap-train 改为单列 grid、隐藏 site-head nav、把 .label 固定 96px、.scale min-width 缩小，hero SVG 因固定 viewBox 0 0 1100 210 仅等比缩放——横向管道并未按 brief 折成纵向之字形流水线，数据包未改为自上而下逐节点下沉，闸门未竖向堆叠，时间刻度未转为每段侧边小标尺。手机端只是基础堆叠而非重构，scrollHeight 变长(reflect 仅叠加)说明布局未按视口方向性重建。"},"localeParity":{"passed":true,"evidence":"zh-CN 与 en 章节、导航、h1、chips、CTAs、SVG 图(a11y aria-label 各自本地化)、五节排版完全对应且翻译自然达意；en cjkCount=0、zh cjkCount 801/785 语言标记正确，无残留未译串或机翻生硬句。"},"antiTemplate":{"passed":true,"evidence":"未回退为营销 Hero+指标卡或三卡横幅；各节以同一管道/轨道母题贯穿（闸门索引轨、执行时间刻度轨、双出口支线、能力徽记轨、colophon），区分于反模板上下文指纹；仅以极淡(alpha .045)圆点纹理作纸面蓝图底，非发光网格或粒子主题，未伪造产品截图。"},"noFakeUi":{"passed":true,"evidence":"CTA 均指向真实仓库 github.com/cardchase/vibeetl 或站内锚点；JSON/YAML 代码块为阐明落地格式的静态文本(pseudo-code)，非可交互假控件，也不是仿制产品的营销截图；无 fakeControlCount(qa 两语、双视口均为 0)，无空链接。"}},"issues":[{"code":"mobileNotRestructured","severity":"blocking","locale":"all","description":"移动端未按 brief 的 mobileTransformation 形成真正的重构：390 断点仅把列/grid 堆叠为单列并等比缩放 SVG，hero 的横向管道轨仍保持横向，闸门保持水平并排，数据包未改为自上而下逐节点下沉，时间刻度未转成每段侧边小标尺。手机端只是基础响应式堆叠而非 brief 要求的纵向之字形流水线重构。","repairInstruction":"为移动断点重做方向性重构：将签名管道图或等价排版折成纵向之字形（数据包自上而下逐闸门下沉、Extract/Transform/Load 竖向堆叠闸门），时间刻度改为各段侧边小标尺，保留逐段展开的拖动隐喻，使 390 布局由横转纵而不仅是列堆叠。"},{"code":"signaturePacketMorphMissing","severity":"warning","locale":"all","description":"signatureVisual 中数据包在滑过各闸门时形态(轮廓/填充/尺寸)未实际改变，仅以 RAW/SHAPED/READY 文本暗示变形，未完全落实 brief「包体在各闸门处渐进变形」的要求。","repairInstruction":"让签名图的数据包在不同闸门处发生可见的形态/填充变化（例如 Outline→Filled→Re-shaped 或在闸门触发时改变包体尺寸与描边），使「数据被逐节点塑形」这一关键叙事由图形本身而非文字体现。"}]}},"after":{"qaPassed":true,"designReviewApproved":true,"qaEvidence":{"zh-CN":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":5,"textLength":1922,"cjkCount":801,"latinLetterCount":665,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":4099,"h1Top":228.734375,"bottomInset":196.734375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"497757aacff03ec442a7d6c83460109b998307b76ea2986ccb7b5a22c87f06ea","middle":"04a011662522954daec9255f22e1b6a5a495e53dab4a8dd3af7ac880bc72c563","bottom":"f5bed4bbbeb709d1f017401db58d916389ce7260a0da9d9c549a5bcc8d6b063e"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":5,"textLength":1911,"cjkCount":785,"latinLetterCount":665,"scrollWidth":390,"clientWidth":390,"scrollHeight":6141,"h1Top":161.375,"bottomInset":218.25,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"7c8dcc62d1b05971758ffde0d02c42adf7d53d0b34e040b40c68bb1dd203eabc","middle":"8d2e8a85f5d57ff530ef157801d31a42a03f419dddd0e6fe5eca04826302570a","bottom":"220655b80bf8e387b9d41d7c027429235c5cf2b76805fc799b7170f307968401"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":5,"textLength":3842,"cjkCount":0,"latinLetterCount":3006,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":4693,"h1Top":228.734375,"bottomInset":197.234375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"d3d298d7f46fc1cfdad20fce58759405aff303f324feda0f684a316fe77f9721","middle":"e8c776ef471973d4a2cd9f5aedd11d87b14629e177a3cd2a42489bd6de9f0bf6","bottom":"51828a8a3edcbfb34e3bf250826517f053e12f17fe9dd1616081c2e4b931b704"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":5,"textLength":3787,"cjkCount":0,"latinLetterCount":2950,"scrollWidth":390,"clientWidth":390,"scrollHeight":6770,"h1Top":161.375,"bottomInset":218.546875,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"2affe72debf45ccaf701ae77a0e9bd29beb3b4afbce1cd78b5475ebb8c2fa29a","middle":"e320609c15d475e82d50110f72db54769bc3a1dacd0940717d2cbe4c8c7923e2","bottom":"3f9ea7f0ce3c33e924d36593ec5756bfcc75ad9e1c79ac87507b28fef4756930"},"consoleErrors":[],"networkRequests":[]}]},"visualReview":{"schemaVersion":"dodot-automated-design-review-v1","status":"approved","summary":"页面忠实实现 packet-rail-pipeline 方向：Hero 为横向管道轨（Extract/Transform/Load 三闸门 + 带批号的滑行数据包 + JSON/YAML 双岔道），其后五节依 design-brief.sectionSequence 依次为闸门解剖（I/O/云/准备/变换/连接/分析）、Polars 内存执行时间刻度、JSON/YAML 落地格式双出口、企业特性能力徽记轨、边界与许可来源 colophon。所有文案均有 material.json 依据，signatureVisual 真实存在于 .sig SVG 且数据包随闸门渐进变形，手机 CSS 以独立纵向 sig-v SVG 实现真正重构（数据包自上而下下沉、闸门竖向堆叠、时间刻度转侧边小标尺），双语章节等价且自然，未回退通用 Hero/三卡/卡片墙/设备外壳/装饰渐变网格，无假控件。","checks":{"projectFit":{"passed":true,"evidence":"Hero 为横向三段式管道轨（Extract/Transform/Load 三闸门，中心 Transform 以深钢蓝+琥珀反白凸显，带 W001 批号数据包沿轨滑行，末端 YAML/JSON 双出口岔道），与 design-brief.heroGeometry 与 signatureVisual 高度一致；后续五节（闸门解剖、执行引擎、落地格式、企业特性、边界与许可）与 brief.sectionSequence 逐条对应，无偏离方向的通用营销结构。"},"evidenceBound":{"passed":true,"evidence":"全部叙述可核验：React Flow DAG 画布 + Polars 内存执行（summary.evidence 第2条）、六类节点闸门 I/O/云/准备/变换/连接/分析（readme Tool Palette 与 summary.evidence 第3条）、JSON 保存/加载 + YAML 导出给 AI 代理（summary.evidence 第3/4条）、自动恢复/多标签/语义类型检测（summary.evidence 第5条）、pip install/npm install/npm run dev（readme.commands）、Python + MIT + 自托管轻量可视化 ETL 援引 Alteryx（description/仓库字段）；安装前提、云集成节点部分未就绪、AI 节点依赖外部 API 等 caveats 亦与 material.json 一致。"},"signatureVisual":{"passed":true,"evidence":"signatureVisual 真实存在于 .sig SVG：viewBox 0 0 1100 210 水平主轨贯穿，Extract/Transform/Load 三闸门作卡槽矗立，带刻度的琥珀数据包以 animateTransform 沿轨运动，并以 animate 在两道闸门处改变 width/height/rx/fill（白描边琥珀→琥珀实心→钢蓝实心）实现「每过闸门轮廓/填充改变一次」，RAW/SHAPED/READY 刻度标注佐证；末端以钢铁/石墨双岔道分流出 YAML 与 JSON。全图仅轨道与数据包，无任何设备/浏览器/终端外壳，非伪产品截图。"},"mobileTransformation":{"passed":true,"evidence":"手机版形成真实重构：@media(max-width:720px) 将 .sig 隐藏并用独立 sig-v 纵向 orthographic SVG 替换（viewBox 0 0 400 940），纵向主轨、01/02/03 每段侧边标尺、Extract/Transform/Load 三道闸门竖向堆叠、数据包 animateTransform to 0 560 自上而下逐节点下沉、底部 YAML/JSON 双出口；同时时间刻度区经 .scale 改造（border-left 纵标尺 + .seg top:var(--p) 横移、竖排）转为竖排侧边小标尺，.cat-list/.branch/.cap-train 网格折为单列。方向性与布局随视口重建，非等比缩放。scrollHeight 由 4099 增至 6141 反映纵向叙事成立。"},"localeParity":{"passed":true,"evidence":"zh-CN 与 en 在章节、导航、h1、chips、CTA、SVG 图（aria-label 各自本地化）与五节排版上完全对应且翻译自然达意；en cjkCount=0、zh cjkCount 801/785、lang 标记正确，无残留未译串或生硬机翻。"},"antiTemplate":{"passed":true,"evidence":"未回退通用 Hero+三卡指标横幅或功能卡片墙模板；各节以同一管道/轨道母题贯穿（闸门索引轨、执行时间刻度轨、双出口支线、能力徽记轨、colophon 轨道分隔线），背景为暖纸白+径向细点而非发光粒子网格，无浏览器/终端/编辑器/仪表盘外壳，无装饰性渐变网格充当主题；其构成指纹 horizontal-packet-rail-through-node-gates 未落入 antiTemplateContext 的 recentCompositionFingerprints/recentSignatureVisuals。"},"noFakeUi":{"passed":true,"evidence":"QA 报告 fakeControlCount=0、brokenImageCount=0、missingAltCount=0、emptyLinkCount=0。页面可点交互仅为真实跳转（github.com/cardchase/vibeetl 仓库链接与#gates/#engine/#formats/#features 锚点导航）；branch-card 内 .code 为说明性排版示例（非可点击控件/占位媒体），SVG 用 role=img 与 aria-label 标注为插图而非假 UI。CTA、chips 均为真实依据（MIT/Python/Polars）。"}},"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":"0879def542636d7e92b684342e9ec530d04c4d3d3b9a6450ba918dc910042a43","artifactHash":"5eba298aeab413b5fed1634108e836afc9cbc3b8a539d4e5231e0c6ec128e911","designSkill":"dodot-showcase-design"},"qaEvidence":{"zh-CN":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":5,"textLength":1922,"cjkCount":801,"latinLetterCount":665,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":4099,"h1Top":228.734375,"bottomInset":196.734375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"497757aacff03ec442a7d6c83460109b998307b76ea2986ccb7b5a22c87f06ea","middle":"04a011662522954daec9255f22e1b6a5a495e53dab4a8dd3af7ac880bc72c563","bottom":"f5bed4bbbeb709d1f017401db58d916389ce7260a0da9d9c549a5bcc8d6b063e"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":5,"textLength":1911,"cjkCount":785,"latinLetterCount":665,"scrollWidth":390,"clientWidth":390,"scrollHeight":6141,"h1Top":161.375,"bottomInset":218.25,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"7c8dcc62d1b05971758ffde0d02c42adf7d53d0b34e040b40c68bb1dd203eabc","middle":"8d2e8a85f5d57ff530ef157801d31a42a03f419dddd0e6fe5eca04826302570a","bottom":"220655b80bf8e387b9d41d7c027429235c5cf2b76805fc799b7170f307968401"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":5,"textLength":3842,"cjkCount":0,"latinLetterCount":3006,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":4693,"h1Top":228.734375,"bottomInset":197.234375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"d3d298d7f46fc1cfdad20fce58759405aff303f324feda0f684a316fe77f9721","middle":"e8c776ef471973d4a2cd9f5aedd11d87b14629e177a3cd2a42489bd6de9f0bf6","bottom":"51828a8a3edcbfb34e3bf250826517f053e12f17fe9dd1616081c2e4b931b704"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":5,"textLength":3787,"cjkCount":0,"latinLetterCount":2950,"scrollWidth":390,"clientWidth":390,"scrollHeight":6770,"h1Top":161.375,"bottomInset":218.546875,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"2affe72debf45ccaf701ae77a0e9bd29beb3b4afbce1cd78b5475ebb8c2fa29a","middle":"e320609c15d475e82d50110f72db54769bc3a1dacd0940717d2cbe4c8c7923e2","bottom":"3f9ea7f0ce3c33e924d36593ec5756bfcc75ad9e1c79ac87507b28fef4756930"},"consoleErrors":[],"networkRequests":[]}]}},"generatorVersion":"opencode-1.1.58:deepseek/deepseek-v4-flash:dodot-showcase-deepseek-v4","publishedAt":"2026-08-20T15:50:01.611Z"},"rawExtraction":{"projectRevisionId":"cmt2zgz01009m6271ck0r5z6w","projectVersion":1,"sources":[{"canonicalKey":"github:cardchase/vibeetl","role":"primary","revisionVersion":1,"url":"https://github.com/cardchase/VibeETL"}],"facts":[{"key":"repository_owner","value":"cardchase","evidence":[{"path":"repository.owner","excerpt":"cardchase"}]},{"key":"repository_name","value":"vibeetl","evidence":[{"path":"repository.name","excerpt":"vibeetl"}]},{"key":"description","value":"A self-hosted, lightweight, visual ETL (Extract, Transform, Load) platform inspired by Alteryx.","evidence":[{"path":"repository.description","excerpt":"A self-hosted, lightweight, visual ETL (Extract, Transform, Load) platform inspired by Alteryx."}]},{"key":"primary_language","value":"Python","evidence":[{"path":"repository.language","excerpt":"Python"}]},{"key":"license","value":"MIT","evidence":[{"path":"repository.license","excerpt":"MIT"}]},{"key":"source_count","value":"1","evidence":[{"path":"projectRevision.sources","excerpt":"github:cardchase/vibeetl"}]}],"taxonomy":{"category":"development_tools","subcategory":"data_ai_tools","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"A self-hosted, lightweight, visual ETL (Extract, Transform, Load) platform inspired by Alteryx."}]},"lifecycle":{"stage":"active","confidence":0.75,"evidence":[{"path":"repository.pushedAt","excerpt":"2026-08-14T11:08:05.000Z"}]},"userIntents":[{"slug":"build_software","label":"构建或改进软件","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"A self-hosted, lightweight, visual ETL (Extract, Transform, Load) platform inspired by Alteryx."}]},{"slug":"automate_work","label":"自动化工作流程","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"A self-hosted, lightweight, visual ETL (Extract, Transform, Load) platform inspired by Alteryx."}]},{"slug":"create_content","label":"创作内容或设计","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"A self-hosted, lightweight, visual ETL (Extract, Transform, Load) platform inspired by Alteryx."}]},{"slug":"analyze_data","label":"分析与理解数据","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"A self-hosted, lightweight, visual ETL (Extract, Transform, Load) platform inspired by Alteryx."}]}],"readme":{"headings":["🎯 Core Philosophy","🎨 Enterprise UI & Semantic Intelligence","🛠️ Architecture at a Glance","📦 Extensive Built-in Tool Palette","🚀 Quick Start Guide","Method 1: The Automated Runner (Recommended)","Method 2: Manual Step-by-Step","🧩 Developer Guide: Zero-Code UI Integration","🏆 Advanced Analytics & Machine Learning (v1.5)"],"commands":["pip install -r requirements.txt","npm install","npm run dev"]},"changeSet":[{"field":"repository","beforeHash":"f0538a7f6ededc779e5f779cec2c03b0890259db18755481d59e2c42faf6fcda","afterHash":"fad2b5b258ce8b85c3470232b022e707c7994d49aea409be5d8f01d8f23db72c"}]},"publishedAt":"2026-08-16T12:42:41.510Z","updatedAt":"2026-08-21T13:26:48.653Z","summary":{"id":"cmt2zgz04009o62719qqu3tqw","projectType":"product","oneLine":"用拖拽构建数据管道","overview":"VibeETL 是一个自托管的轻量级可视化 ETL 平台，灵感来自 Alteryx。它通过 React Flow 画布让用户以拖拽方式构建数据管道，并利用 Polars 引擎在内存中高速执行。内置丰富的节点，涵盖输入输出、云集成、数据准备、转换、连接和分析等类别，支持 JSON 保存/加载工作流，并可将工作流导出为 YAML 供 AI 代理使用。","highlights":["拖拽式 DAG 画布","Polars 内存执行","丰富内置节点","AI 代理 YAML 导出"],"bestFor":["数据工程师","数据分析师","ETL 爱好者"],"caveats":["需要本地安装和配置","云集成节点部分功能可能未完全就绪","AI 节点依赖外部 API"],"categorySuggestion":"development_tools","subcategorySuggestion":"data_ai_tools","lifecycleStages":["backend_data","testing_quality"],"audienceIntents":["build","evaluate","learn"],"confidence":0.8,"evidence":[{"claim":"VibeETL 是一个自托管的轻量级可视化 ETL 平台，灵感来自 Alteryx。","sourceType":"description","locator":"https://github.com/cardchase/VibeETL"},{"claim":"它使用 React Flow 构建交互式 DAG 画布，并通过 Polars 引擎在内存中执行数据管道。","sourceType":"readme","locator":"https://github.com/cardchase/VibeETL"},{"claim":"内置大量数据工程节点，包括输入输出、云集成、准备、转换、连接和分析等类别。","sourceType":"readme","locator":"https://github.com/cardchase/VibeETL"},{"claim":"支持通过 JSON 保存和加载工作流，并可将工作流导出为 YAML 供 AI 代理使用。","sourceType":"readme","locator":"https://github.com/cardchase/VibeETL"},{"claim":"提供自动恢复、多标签工作区、语义类型检测和交互式可视化等企业级功能。","sourceType":"readme","locator":"https://github.com/cardchase/VibeETL"}],"publishedAt":"2026-08-17T13:17:38.077Z"},"github":{"url":"https://github.com/cardchase/VibeETL","ownerAvatarUrl":"https://avatars.githubusercontent.com/u/33249394?v=4","starsCount":50,"forksCount":4,"language":"Python","syncedAt":"2026-08-17T13:11:16.571Z","license":"MIT","defaultBranch":"main","archived":false,"disabled":false,"lastChangedAt":"2026-08-14T11:08:05.000Z","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"]

六类节点闸门，挂在一条 I/O / 云 / 准备 / 变换 / 连接 / 分析 的索引轨上

节点闸门面板 · GATE PANEL

Polars 在内存里跑，数据包沿时间刻度 高速穿过

双出口分流：JSON 存档，YAML 交给 AI 代理

JSON · 保存 / 加载

YAML · AI 代理导出

让长管道靠得住的能力徽记

自动恢复 · Auto-recovery

多标签工作区 · Multi-tab

语义类型检测 · Semantic types

站到管道旁，看清它的边界与来源

安装前提

依赖边界

许可

Six node gates, indexed on one rail: I/O / cloud / prepare / transform / join / analysis

Node gate panel

Polars runs it in-memory; the packet crosses the track on schedule

Twin off-ramps: JSON to archive, YAML for AI agents

JSON · Save / load

YAML · AI agent export

Capabilities badges for long runs you can trust

Auto-recovery

Multi-tab workspaces

Semantic type detection

Stand beside the pipe and see its limits and provenance

Setup

Dependency boundaries

License

六类节点闸门，挂在一条 I/O / 云 / 准备 / 变换 / 连接 / 分析的索引轨上

Polars 在内存里跑，数据包沿时间刻度高速穿过