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:T4eb4, 研究之流 — awesome-vibe-research

github.com/modelscope/awesome-vibe-research · Apache-2.0 · #0–8

研究之流.

AI 辅助科研的开放共建知识库

一个开放、协同共建的科研资源仓库，将科研全生命周期梳理为九道被社区持续淘洗的水段，每一段都沉淀出可复用的 agents、skills、workflows、tools 与最佳实践——从提出一个问题，到它被传播为影响。

397 Stars 43 Forks 9 个流程阶段 Python 主语言 Apache-2.0 许可

FLOW MAP · 科研流程河道 · 编号 0–8

0全流程

端到端自动科研：从问题到发布的一条连续动线。

端到端 · 全链路

1方向扫描

如何找到值得研究的问题并定义边界。

选题 · 问题定义

2文献研究

检索、精读、综述与引用网络。

检索 · 精读 · 综述

3方法设计

为问题设计可重复的方法与实验方案。

实验方案 · 假设

4实验执行

执行实验并分析结果，形成判断依据。

运行 · 分析

5科学可视化

论文插图、科学绘图与可视化表达。

插图 · 表达

6论文写作

写作、投稿与同行评审的推进。

写作 · 投稿 · 评审

7复现发布

让结果可复现、可发布并归档。

复现 · 发布 · 归档

8传播影响

传播、教学与影响力分析。

传播 · 教学 · 影响

流动主色文献色实验色断面厚度 ≈ 阶段沉淀密度

Stage Detail

每一帘水，都沉淀出什么

仓库把九大阶段逐段拆开：先点出每个阶段的典型问题，再列出可沉淀的 AI 组件类型。

全流程 · 端到端自动科研

一个问题如何一路自动跑到发布？

Q. 将多阶段串联为一条连续动线，跨阶段的工作流与编排。

方向扫描与问题定义

怎么找到值得研究的问题，并把它定义清楚？

Q. 选题挖掘、前沿追踪与问题表述的 agents。

文献研究

如何在大量文献里检索、精读并形成综述？

Q. 检索、精读、综述与引用网络工具。

方法设计

怎样构建设计可重复的方法与实验？

Q. 方法建模、实验方案与假设设计。

实验执行与分析

如何跑实验并得出可信的分析？

Q. 实验运行、跟踪与分析 workflows。

科学可视化

怎么把结果画成能被读懂的论文插图？

Q. 论文插图、科学绘图与可视化表达工具。

论文写作、投稿与评审

如何把结论写清楚并推进审稿？

Q. 写作、润色、投稿与评审辅助。

复现、发布与归档

怎样让研究结果可复现、可发布、可归档？

Q. 复现检查、发布载体与归档流程。

传播、教学与影响力

如何让研究被更多人看到、学会并产生影响？

Q. 传播、教学与影响力分析组件。

Curation Index

社区淘洗出的索引

每条河流的截面，都是不同类型的可复用资产被实测、筛选后归档。

agents承担某一阶段任务的智能体，自动化科研的特定环节。
skills可被复用的单项能力，能并入更大流程。
workflows把多个环节串联起来的完整动线。
tools支撑某一任务的工具，如检索、绘图与反馈。
best practices来自实践的沉淀经验，记录「怎么做更对」。

项目与技能按阶段归档，随社区持续补录。每种资产都尽可能给出项目入口、Stars 与论文，方便研究者判断是否值得接进自己的工作流。

这是一张活的清单，而非封闭的榜单——任何经验都有机会被下一段流程重新淘洗。

Contribution

把你自己涌进来的那一段

仓库向所有人开放补录：它正是靠一次次贡献成形的。

如何贡献

找到你想补充的空缺阶段或遗漏条目。

在对应的收录表格中添加缺失条目，附上项目信息与论文。

提交改动，让社区一起验证与收录。

前往仓库贡献 →

收录原则

仓库并非无差别堆砌，而是有筛选标准地把真正值得复用的资产沉淀下来，保持河流的清澈与方向。

Provenance & Community

它从哪里来，又流到哪里去

这条河流始于一群人的真实讨论，终点是更广阔的科研共同体。

沙龙起源

这面索引由「与 Agent 共事做研究」系列沙龙起步——沉淀 speaker 的实践经验、观众的反馈，以及被反复验证的开源项目与可复用流程。

Doing Research with Agents

社区的贡献

由 modelscope 组织，在 Apache-2.0 许可下开放共建。仓库仍在活跃更新，欢迎每一位研究者带着自己的问题与经验加入。

Apache-2.0 modelscope

23:T4d2f, Research Current — awesome-vibe-research

github.com/modelscope/awesome-vibe-research · Apache-2.0 · #0–8

Research Current.

Open knowledge base for AI-assisted research

An open, collaboratively-built repository for AI-assisted scientific research. It maps the whole research lifecycle into nine streams the community keeps sifting — each one distilling reusable agents, skills, workflows, tools, and best practices, from a single question to a widely felt result.

397 Stars 43 Forks 9 lifecycle stages Python primary language Apache-2.0 license

FLOW MAP · research-lifecycle river · #0–8

0End-to-end

One continuous track from question to release.

Full pipeline

1Direction scanning

Finding a question worth asking and framing it.

Ideation

2Literature

Search, deep reading, synthesis, citation.

Search · Reading

3Method design

Designing a repeatable method and protocol.

Design

4Experiments

Running experiments and reading the results.

Run · Analyze

5Visualization

Scientific figures and visual storytelling.

Figures

6Writing

Writing, submission, and peer review.

Writing · Review

7Reproduction

Making results reproducible, published, archived.

Release

8Impact

Sharing, teaching, and tracing influence.

Outreach

Flow Literature Experiments Section thickness ≈ density of distilled assets

Stage Detail

What each stream settles out

Across all nine stages, the repository names the typical question first, then the reusable AI components a stage can yield.

End-to-end autonomous research

How does a question run all the way to release?

Q. Multi-stage orchestration that threads the whole lifecycle into one flow.

Direction scanning & framing

How do you spot a good question and define it clearly?

Q. Agents for ideation, trend tracking, and problem framing.

Literature research

How do you search, read, and synthesize the literature?

Q. Tools for search, deep reading, reviews, and citation networks.

Method design

How do you design a repeatable method and experiment?

Q. Method modeling, experimental protocol, and hypothesis design.

Experiments & analysis

How do you run experiments and draw sound conclusions?

Q. Workflows for running, tracking, and analyzing experiments.

Scientific visualization

How do you turn results into readable figures?

Q. Tools for figures, scientific plotting, and visual expression.

Writing & review

How do you write it up and move a submission forward?

Q. Support for writing, polishing, submission, and peer review.

Reproduction & release

How do you make results reproducible, published, and archived?

Q. Reproducibility checks, publishing channels, and archiving.

Outreach & impact

How does the work get seen, learned, and felt?

Q. Components for sharing, teaching, and impact analysis.

Curation Index

An index the community keeps cleaning

Every cross-section of the river is a different kind of reusable asset, triaged and archived after real use.

agentsAgents that carry a stage of research, automating a specific part of the work.
skillsReusable single capabilities you can fold into a bigger flow.
workflowsFull tracks that string several stages together.
toolsUtilities underpinning a task — retrieval, figures, feedback.
best practicesLessons distilled from practice, recording how the work gets done right.

Projects and skills are filed by stage and keep being added by the community. Each entry points to the project, its star count, and supporting papers, so researchers can judge whether it belongs in their own workflow.

It is a living list, not a closed ranking — any experience can be re-sifted by the next stage of the flow.

Contribution

Add the stream you live in

The repository stays open for additions — it is shaped by contributions just like this one.

How to contribute

Find the stage or entry you want to fill in.

Add what is missing to the corresponding table, with project details and papers.

Propose the change and let the community verify and include it.

Visit the repository →

Selection principles

This is not an indiscriminate pile-up. Assets are kept only when they genuinely earn a place, keeping the river clear and on course.

Provenance & Community

Where it began, and where it flows to

The river begins in real conversations and empties into a wider research community.

Salon origins

This index grew out of the "Doing Research with Agents" salon series, distilling speakers' practice, audience feedback, and the open projects and workflows that held up.

Doing Research with Agents

The community that keeps it flowing

Organized by modelscope and openly co-built under the Apache-2.0 license. The repository is still actively updated — bring your own questions and experience.

Apache-2.0 modelscope

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","cmt2zh1iv013l6271tr4ivspp:zh-CN:beautified",{"project":{"id":"cmt2zh1iv013l6271tr4ivspp","slug":"modelscope-awesome-vibe-research","name":"awesome-vibe-research","description":"收集科研全流程的 AI 工具与最佳实践","authorName":"modelscope","authorVerification":{"verified":false,"verifiedAt":null},"category":"development_tools","subcategory":"data_ai_tools","experienceUrl":null,"qualityScore":63,"featured":false,"officialSite":null,"views":{"original":{"available":true},"official":{"available":false,"reason":"disabled_by_policy"},"beautified":{"available":true,"reason":null},"defaultView":"beautified"},"beautifiedView":{"revisionId":"cmt2zh1jh013x6271f4v6j3sv","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":"research-current","name":"研究之流 Research Current","projectMetaphor":"科研是有方向的流动：一个问题被提出后，流过方向扫描、文献、方法、实验、可视化、写作、复现、传播这九道被社区淘洗过的水系，每一帘都沉淀出可复用的组件。","signatureVisual":"一条横向蜿蜒的九段流程河道；每一段是一个阶段节点，断面厚度代表该阶段沉淀的组件密度，段与段之间用连续而非断点的箭头衔接，强调「流动而非堆叠」。","heroGeometry":"水平蛇形河道（serpentine track）：九段以 S 形从左到右蜿蜒，当前段高亮抬高，显示其上标着的阶段名与沉淀组件计数。","sectionSequence":["flow-map","stage-detail","curation-index","contribution","provenance-community"],"surfaceModel":"多层半透明「水帘」横带堆叠：每层次带承载一阶段，带间有看得见的线性流动轨迹，带面亚光、边缘收敛而非泛光，透气感来自稀疏留白而非光晕。","palette":"深青底 + 三段琥珀（流动主色）、水绿（文献）、石墨青（实验），白/冷灰文字，青色仅在进程轨上用。无渐变光晕。","typography":"衬线标题（科研叙事的沉稳）+ 等宽小字做阶段编号/组件名，正文用无衬线；编号系统（0–8）贯穿全页如水的刻度。","desktopComposition":"Hero 为九段蛇形河道，下方五段正文对齐河道节点成立轴基准；贡献与溯源区以细线列表呈现，让结构像地图上的图例而不像卡片墙。","mobileTransformation":"蛇形河道压平成纵向有序账本/检查清单：每阶段一行，前行有动线箭头，tap 展开为「典型问题 + 沉淀组件」两行细节，避免横向滚动。","mediaStrategy":"无本地素材，删除媒体区；以九段本体作为标志视觉，不伪造产品截图。","forbiddenPatterns":["三卡指标墙（stars/forks/用户数平铺）","浏览器/终端/编辑器外壳","发光网格或粒子代替结构","九段被压成营销 Hero 下的功能卡片"],"compositionFingerprint":"九段蛇形河道，段为厚度不等的横向水帘，段间连续箭头，编号 0–8 贯穿，主体在河道而不在指标。"},{"id":"specimen-catalog","name":"标本档案 Specimen Catalog","projectMetaphor":"这个仓库像一套编号标本档案：每个科研阶段是一张归档卡，卡上记录该阶段的典型问题、可沉淀的组件与最好的项目，由社区不断补录新标本。","signatureVisual":"一摞斜置并扇开的存档卡（ledger/catalog index），每张卡顶部是阶段编号与名称，卡身是可读的清单性字段行，底层露出「社区补录」的开放插槽。","heroGeometry":"左手竖向堆叠成「归档卡扇」——五到七张卡按次序斜排列开，中央卡为正视全展开视图。","sectionSequence":["catalog-index","stage-entry-detail","taxonomy-keys","contribution-slot","provenance","license"],"surfaceModel":"纸张档案面：卡片呈浅暖底纸感、带细刻度线与字段分隔线（ruled），卡片边缘有轻微归档切角，阴影克制仅作堆叠深度，无泛光。","palette":"纸白/砂暖底 + 深墨文字，标签用档案红（编号、分类）+ 墨绿（项目链接），分隔线为文件柜灰。","typography":"等宽字体主导（卡片档案、编号、字段值），标题用衬线做档案名称，正文无衬线；编号系统（Stage-0…Stage-8）与字段标签（Q./Cap./Ref.）贯穿。","desktopComposition":"Hero 为归档卡扇，中央卡为全展开的详细条目字段；下方六段围绕「索引→条目→分类键→补录槽→溯源→许可」成线性台账，表格行而非卡片墙。","mobileTransformation":"扇开卡叠成单列纵向索引：每行=一张卡（编号+阶段+计数），点击翻开为字段明细；补录槽与溯源缩为表单友好的一列，保卡片隐喻、换空间决策。","mediaStrategy":"无本地素材，删除媒体区；以归档索引 + 字段清单作为标志视觉，不伪造项目截图。","forbiddenPatterns":["功能卡片墙（可套任意文案）","发光网格或粒子代替抽屉结构","伪造的图表/开关/输入控件","用统一圆角+阴影让每段看起来同质"],"compositionFingerprint":"编号归档卡扇 + 细粒度字段台账，等宽编号（Stage-0…8）与 Q./Cap./Ref. 字段标签贯通全页，补录以开放插槽呈现。"},{"id":"claim-orbit","name":"证言轨道 Claim Orbit","projectMetaphor":"科研从一个开放问题（核心）向外辐射：围绕它的每一圈轨道是一次扩散阶段，社区把最好的 agent、skill、workflow 放在对应的环带上，像引用网络一圈圈展开影响。","signatureVisual":"同心参考圆环（compass/scale 标尺），中心为「开放研究问题」，向外四到六圈环带分别对应科研阶段，环上以刻度节点标记沉淀的组件。","heroGeometry":"居中核心 + 同心环（concentric rings）：核心是问题点，外环由近及远，当前环高亮并悬停显示其组件刻度。","sectionSequence":["core-orbit","ring-detail","node-index","contribution-ring","provenance-ring"],"surfaceModel":"径向深度模型：同心环从中心向外铺开，环带用描边刻度而非填充色块表达，环间标注距离刻度（如同罗盘），无发光。","palette":"深空底 + 一圈一色调（核心暖白），刻度节点用单一强调色区分真沉淀/待补录，文字冷白；色带仅用于环线。","typography":"极简无衬线主体，环带标签与刻度用等宽字体做坐标读法（R1, θ=…），标题紧贴核心问题文本；编号隐藏，由坐标替代。","desktopComposition":"Hero 为同心环中心问题，向下四段按「环带展开→节点索引→补录环→溯源环」逐层拉开半径，各部分以同心对齐居中排布，形成由内到外的扩张叙事。","mobileTransformation":"同心环无法缩小，重构为「从中心向外的径向列表」：核心问题置顶，每条环带=一行向外扩展的记录，露出刻度；坐标读数保留但转为纵向堆叠，避免裁切环。","mediaStrategy":"无本地素材，删除媒体区；以同心环 + 刻度节点作为标志视觉，不伪造产品截图。","forbiddenPatterns":["三卡指标墙","浏览器/终端/编辑器外壳","发光网格或粒子代替结构","环被压成普通圆形 icon 网格"],"compositionFingerprint":"同心标尺环 + 核心问题居中，环为描边刻度而非填充，坐标式读数（R1、θ）代替编号，叙事由内向外扩张。"}],"selectedDirectionId":"research-current","selectionRationale":"项目核心机制是「9 段科研生命周期被社区持续淘洗、每个阶段沉淀可复用 AI 组件」。research-current 用横向蛇形河道直接可视化这段流动过程：Hero 几何即 9 段流程本体，能把「生命周期」这一最核心证据直接变成签名视觉，而非营销堆叠。specimen-catalog 更偏档案组织方式、claim-orbit 更偏文献引用隐喻，二者对「全流程生命周期」的揭示弱于流向本身；且无本地素材时河道的连续箭头 + 阶段厚度更能表达「阶段之间如何衔接」这一真实关系。故选 research-current。"},"designBrief":{"schemaVersion":"dodot-design-brief-v1","selectedDirectionId":"research-current","projectMetaphor":"科研是有方向的流动：一个问题被提出后，流过方向扫描、文献、方法、实验、可视化、写作、复现、传播这九道被社区淘洗过的水系，每一帘都沉淀出可复用的组件。","signatureVisual":"一条横向蜿蜒的九段流程河道；每一段是一个阶段节点，断面厚度代表该阶段沉淀的组件密度，段与段之间用连续而非断点的箭头衔接，强调「流动而非堆叠」。","compositionFingerprint":"九段蛇形河道，段为厚度不等的横向水帘，段间连续箭头，编号 0–8 贯穿，主体在河道而不在指标。","heroGeometry":"水平蛇形河道（serpentine track）：九段以 S 形从左到右蜿蜒，当前段高亮抬高，显示其上标着的阶段名与沉淀组件计数。","sectionSequence":["flow-map","stage-detail","curation-index","contribution","provenance-community"],"surfaceModel":"多层半透明「水帘」横带堆叠：每层次带承载一阶段，带间有看得见的线性流动轨迹，带面亚光、边缘收敛而非泛光，透气感来自稀疏留白而非光晕。","palette":"深青底 + 三段琥珀（流动主色）、水绿（文献）、石墨青（实验），白/冷灰文字，青色仅在进程轨上用。无渐变光晕。","typography":"衬线标题（科研叙事的沉稳）+ 等宽小字做阶段编号/组件名，正文用无衬线；编号系统（0–8）贯穿全页如水的刻度。","desktopComposition":"Hero 为九段蛇形河道，下方五段正文对齐河道节点成立轴基准；贡献与溯源区以细线列表呈现，让结构像地图上的图例而不像卡片墙。","mobileTransformation":"蛇形河道压平成纵向有序账本/检查清单：每阶段一行，前行有动线箭头，tap 展开为「典型问题 + 沉淀组件」两行细节，避免横向滚动。","mediaStrategy":"无本地素材，删除媒体区；以九段本体作为标志视觉，不伪造产品截图。","evidencePlan":["Hero 河道九段编号 0–8 对应 readme 九大阶段，阶段名取自 readme 标题（方向扫描、文献研究、方法设计、实验执行、科学可视化、论文写作、复现发布、传播影响、全流程端到端）。","stage-detail 段引用 readme 的「典型问题 + 可沉淀组件类型」结构，逐阶段以文字呈现。","curation-index 段以细线台账呈现社区收录模式（agents、skills、workflows、tools、最佳实践）与贡献机制。","provenance-community 段引用「Doing Research with Agents（与 Agent 共事做研究）」沙龙起源这一事实。","无产品截图/样本数据，全部数字仅限仓库事实（stars 397、forks 43、Apache-2.0、Python），不杜撰条目列表内容。"],"localeContract":{"zh-CN":"标题用「研究之流」，slogan 沿用『AI 辅助科研的开放共建知识库』；九段阶段名与典型问题用中文，文字密度允许较高，编号 0–8 保留。","en":"Title uses 'Research Current'; slogan follows 'Open knowledge base for AI-assisted research'; nine-stage names and typical questions in English, lower text density, numbering 0–8 retained."},"forbiddenPatterns":["三卡指标墙（stars/forks/用户数平铺）","浏览器/终端/编辑器外壳","发光网格或粒子代替结构","九段被压成营销 Hero 下的功能卡片","伪造图表、开关、搜索、输入控件、客户评价"]},"visualReview":{"schemaVersion":"dodot-automated-design-review-v1","status":"approved","summary":"页面忠实实现了 design-brief 的 research-current（研究之流）方向：九段编号 0-8 的横向流程河道贯穿全页并配连续箭头与断面密度条，5 个章节与 brief.sectionSequence 完全一致，深青底+琥珀/水绿/石墨青配色、衬线标题+等宽编号均符合 palette/typography 契约；所有数字（397 stars、43 forks、9 阶段、Python、Apache-2.0）与「Doing Research with Agents」沙龙起源均有 material.json 事实依据，未杜撰任何条目列表；手机端 CSS 将蛇形河道压平为纵向账本且无横向滚动，双语章节逐段对应、措辞自然。唯一非阻断提示：桌面 Hero 以直线横排 flex 实现九段河道，未形成 brief 所述的 S 形蜿蜒几何，但编号、连续箭头、断面厚度三要素齐全，签名视觉真实存在。","checks":{"projectFit":{"passed":true,"evidence":"zh-CN.html/en.html 的 data-dodot-section 依次为 flow-map、stage-detail、curation-index、contribution、provenance-community，与 design-brief.json sectionSequence 逐项匹配，共 5 章（契约 min4/max7）；flow-map 为九段河道、stage-detail 为九行台账、curation-index 为细线收录索引、contribution 为贡献、provenance-community 为沙龙起源；章节结构围绕「研究之流」隐喻而非通用展示。"},"evidenceBound":{"passed":true,"evidence":"hero-meta 的 397 Stars、43 Forks、9 流程阶段、Python 主语言、Apache-2.0 许可与 material.json projection.repository 事实一致；provenance-community 的『与 Agent 共事做研究 / Doing Research with Agents』沙龙起源对应 summary.evidence 第 4 条；九段阶段名取自 readme 标题（全流程/方向扫描/文献研究/方法设计/实验执行/科学可视化/论文写作/复现发布/传播影响）；材质条仅标注『断面厚度≈沉淀密度』为示意图，未杜撰具体条目内容或样本数据。"},"signatureVisual":{"passed":true,"evidence":"flow-map 段真实实现了九段横向河道：9 个 .seg 含编号 0-8、衬线阶段名、seg-count 与底部 seg-flow 断面密度条（宽度各异），段间以 → 连续箭头衔接（末段隐藏），并附 river-legend 图例解释厚度与主色含义；编号系统 0-8 贯穿 river 与 stage-detail 两章，满足 compositionFingerprint『段间连续箭头 + 断面厚度 + 编号贯穿，主体在河道而不在指标』。注：桌面为直线 flex 横排，未呈现 brief heroGeometry 所述 S 形蜿蜒，见 issues。"},"mobileTransformation":{"passed":true,"evidence":"@media (max-width:620px) 下 .serpentine 改为 flex-direction:column 纵向有序账本、每段一行，隐藏 .seg-arrow，seg-count 移至段右侧、作为可读的检查清单式计数；qa-report 显示两语言移动端 scrollWidth=clientWidth=390，无横向滚动，匹配 brief『压平成纵向有序账本/检查清单，避免横向滚动』。"},"localeParity":{"passed":true,"evidence":"zh-CN 首页 H1 为『研究之流』、en 为 'Research Current'，分别命中 localeContract 的标题约定；slogan 沿用『AI 辅助科研的开放共建知识库』/'Open knowledge base for AI-assisted research'；九段阶段名与典型问题在两种语言逐段对应、编号 0-8 均保留，en 文本密度更低且用词自然（literature/writing/impact 等），章节标题互为等价翻译，无错位或缺失分节。"},"antiTemplate":{"passed":true,"evidence":"未回退到通用 Hero/三卡指标墙：hero-meta 是细小的等宽元信息行而非平铺指标卡；无浏览器/终端/编辑器外壳；无发光网格或粒子代偿结构（配色明确『无渐变光晕』，body 仅用最简纵向深青渐变底色，河道为主角）；九段未被压成 Hero 下方的营销功能卡，而是河道本体，下方正文以台账/细线列表展开；contribution 与 provenance 为细线 step/panel 列表而非卡片墙，符合 brief 的 desktopComposition。"},"noFakeUi":{"passed":true,"evidence":"seg-flow 密度条是表现断面厚度的示意元素而非可交互伪控件；contribution 三步为真实文字步骤并指向真实仓库链接 https://github.com/modelscope/awesome-vibe-research；无伪造图表、开关、搜索框、输入控件或客户评价；qa-report 两语言×两视口 fakeControlCount=0、brokenImageCount=0（无本地素材/设备壳，符合 mediaStrategy 删除媒体区），assetManifest 为空与页面一致。"}},"issues":[{"code":"gek-serpentine-shape-not-realized","severity":"warning","locale":"all","description":"brief 的 heroGeometry 明确要求九段以 S 形从左到右蜿蜒（'以 S 形从左到右蜿蜒，当前段高亮抬高'），但桌面实现为 .serpentine{display:flex} 的直线等宽横排九卡，未见 S 形蛇形蜿蜒或当前段抬高的几何处理，仅 .seg:hover 有 hover 抬起。虽编号 0-8、连续箭头、断面厚度三要素齐全、签名视觉真实存在，但 S 形蜿蜒这一几何特征是 research-current 区别于普通一行卡带的关键造型，未完全落地。","repairInstruction":"在桌面视口为 .serpentine 或父容器引入错位行距/translateY 使九段呈 S 形蛇形排布（例如奇偶段交替纵移并衔接连续箭头路径），或对当前聚焦段做持续性抬高以还原 brief 要求的『蛇形蜿蜒 + 当前段高亮抬高』；此处为打磨项，不构成阻断。"}]},"repairEvidence":[],"provenance":{"executor":"opencode","model":"deepseek/deepseek-v4-flash","opencodeVersion":"1.1.58","promptVersion":"dodot-showcase-deepseek-v4","inputHash":"86cd76d61d57953955cd4c8bb469521981a5bbb66dfa8de80406224cb1010039","artifactHash":"45b96ca37d68e216e71f6e566808d92b07e65c868bd7bca73fef031fd18917f1","designSkill":"dodot-showcase-design"},"qaEvidence":{"zh-CN":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":4,"sectionCount":5,"textLength":1978,"cjkCount":1186,"latinLetterCount":293,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":4093,"h1Top":142.125,"bottomInset":215.9375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"3e74a7df98b91fbfd16cf5ad42940c6cb460d865d634aa3d16b9c487b1c10038","middle":"2ab9f0989530a86464c9693bf3f1853dd18964385a90653e44b63264ce1f87c6","bottom":"22d266c909f021d9a5a530353941b4645e32c73ce333db384bc1ca0f600b32ad"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":4,"sectionCount":5,"textLength":1946,"cjkCount":1186,"latinLetterCount":293,"scrollWidth":390,"clientWidth":390,"scrollHeight":5710,"h1Top":150.25,"bottomInset":215.734375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"76bde20c7cb3f73799cf76a5746082b86f6f0494732ff4d18af0e6bdaa7f817e","middle":"19ddb74572b422d0bfce9b6dd205f17d9ee722bd7ee04fb23b76908c7da58483","bottom":"28dec12145947d0875c7127b4dac1a5ff9696edc3deadb44cd6adc4ef0687845"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":4,"sectionCount":5,"textLength":4744,"cjkCount":0,"latinLetterCount":3729,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":4339,"h1Top":142.125,"bottomInset":215.578125,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"fd936608b8cced254e203c4aace9018048cc766ef0fac16866b89fb99add0e23","middle":"75ad9afeba839ea620f32270098411556205e6b8850c5df73cd93ecc370bdd7a","bottom":"6c5695b26ae711109d0f40ec4f066d0a7767b56dbba57b3741775c60ef17597a"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":4,"sectionCount":5,"textLength":4712,"cjkCount":0,"latinLetterCount":3729,"scrollWidth":390,"clientWidth":390,"scrollHeight":6658,"h1Top":150.25,"bottomInset":215.59375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"4a16c09c87c37d85d712cca6c8893f2a3391d5d280e090fa563653dfce737031","middle":"0e010f415ca1fa8c9f0a5514040929dbe1b4082831f6ad901f08cede465d42e1","bottom":"2924cdf1ed4144c412706f1aeea61f97fbf3b6ffcdbd8033b84223ac0f5d699d"},"consoleErrors":[],"networkRequests":[]}]}},"generatorVersion":"opencode-1.1.58:deepseek/deepseek-v4-flash:dodot-showcase-deepseek-v4","publishedAt":"2026-08-20T07:06:54.551Z"},"rawExtraction":{"projectRevisionId":"cmt2zh1j6013r62715ps8bz8p","projectVersion":1,"sources":[{"canonicalKey":"github:modelscope/awesome-vibe-research","role":"primary","revisionVersion":1,"url":"https://github.com/modelscope/Awesome-Vibe-Research"}],"facts":[{"key":"repository_owner","value":"modelscope","evidence":[{"path":"repository.owner","excerpt":"modelscope"}]},{"key":"repository_name","value":"awesome-vibe-research","evidence":[{"path":"repository.name","excerpt":"awesome-vibe-research"}]},{"key":"description","value":"An open, collaboratively-built repository for AI-assisted scientific research — collecting and curating agents, skills, workflows, tools, and best practices across the full research lifecycle. 面向 AI 辅助科研的开放共建仓库收集和沉淀科研全流程中的 agents、skills、workflows、tools 与最佳实践","evidence":[{"path":"repository.description","excerpt":"An open, collaboratively-built repository for AI-assisted scientific research — collecting and curating agents, skills, workflows, tools, and best practices across the full research lifecycle. 面向 AI 辅助科研的开放共建仓库收集和沉淀科研全流程中的 agents、skills、workflows、tools 与最佳实践"}]},{"key":"primary_language","value":"Python","evidence":[{"path":"repository.language","excerpt":"Python"}]},{"key":"license","value":"Apache-2.0","evidence":[{"path":"repository.license","excerpt":"Apache-2.0"}]},{"key":"source_count","value":"1","evidence":[{"path":"projectRevision.sources","excerpt":"github:modelscope/awesome-vibe-research"}]}],"taxonomy":{"category":"development_tools","subcategory":"data_ai_tools","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"An open, collaboratively-built repository for AI-assisted scientific research — collecting and curating agents, skills, workflows, tools, and best practices across the full research lifecycle. 面向 AI 辅助科研的开放共建仓库收集和沉淀科研全流程中的 agents、skills、workflows、tools 与最佳实践"}]},"lifecycle":{"stage":"active","confidence":0.75,"evidence":[{"path":"repository.pushedAt","excerpt":"2026-08-17T02:52:43.000Z"}]},"userIntents":[{"slug":"build_software","label":"构建或改进软件","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"An open, collaboratively-built repository for AI-assisted scientific research — collecting and curating agents, skills, workflows, tools, and best practices across the full research lifecycle. 面向 AI 辅助科研的开放共建仓库收集和沉淀科研全流程中的 agents、skills、workflows、tools 与最佳实践"}]},{"slug":"automate_work","label":"自动化工作流程","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"An open, collaboratively-built repository for AI-assisted scientific research — collecting and curating agents, skills, workflows, tools, and best practices across the full research lifecycle. 面向 AI 辅助科研的开放共建仓库收集和沉淀科研全流程中的 agents、skills、workflows、tools 与最佳实践"}]},{"slug":"create_content","label":"创作内容或设计","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"An open, collaboratively-built repository for AI-assisted scientific research — collecting and curating agents, skills, workflows, tools, and best practices across the full research lifecycle. 面向 AI 辅助科研的开放共建仓库收集和沉淀科研全流程中的 agents、skills、workflows、tools 与最佳实践"}]},{"slug":"analyze_data","label":"分析与理解数据","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"An open, collaboratively-built repository for AI-assisted scientific research — collecting and curating agents, skills, workflows, tools, and best practices across the full research lifecycle. 面向 AI 辅助科研的开放共建仓库收集和沉淀科研全流程中的 agents、skills、workflows、tools 与最佳实践"}]},{"slug":"learn_explore","label":"学习与探索","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"An open, collaboratively-built repository for AI-assisted scientific research — collecting and curating agents, skills, workflows, tools, and best practices across the full research lifecycle. 面向 AI 辅助科研的开放共建仓库收集和沉淀科研全流程中的 agents、skills、workflows、tools 与最佳实践"}]}],"readme":{"headings":["🗺️ 科研流程地图","0 🔄 全流程（端到端自动科研）","1 🔭 方向扫描与问题定义","2 📚 文献研究：检索、精读、综述与引用网络","3 🧩 方法设计","4 ⚗️ 实验执行与分析","5 📊 科学可视化：论文插图、科学绘图与可视化表达","6 ✍️ 论文写作、投稿与同行评审","7 📦 复现、发布与归档","8 📡 传播、教学与影响力分析","🏆 最佳实践","📋 收录原则","🤝 如何贡献","🙏 致谢"],"commands":[]},"changeSet":[{"field":"repository","beforeHash":"1a18980a42a9b97629ae767f230f3a2f3b9778b21198d90546ea077c03ccabb0","afterHash":"721346692585ccc703cc644ceab476b466ea48de413b97ce83e9b40b839c1197"}]},"publishedAt":"2026-08-16T12:31:30.202Z","updatedAt":"2026-08-21T13:26:51.928Z","summary":{"id":"cmt2zh1j9013t62714fkg74hh","projectType":"other","oneLine":"收集科研全流程的 AI 工具与最佳实践","overview":"这是一个开放共建的 AI 辅助科研资源仓库，按科研生命周期 9 个阶段（方向扫描、文献研究、方法设计、实验执行、可视化、写作、复现发布等）收录 agents、skills、workflows、tools 与最佳实践。每个阶段列出典型问题和可沉淀的组件类型，并附有项目链接、Stars 和论文。适合希望快速了解 AI 科研工具生态、寻找可复用工作流或贡献自己经验的研究者与开发者。","highlights":["覆盖科研全流程 9 阶段","收录 agents、skills、workflows","社区共建，可自由贡献","附项目链接与论文"],"bestFor":["想探索 AI 科研工具的研究者","需要可复用工作流的开发者","希望贡献经验的社区成员"],"caveats":["仓库为资源列表，不提供直接运行功能","条目质量依赖社区维护，需自行验证","部分项目 Stars 数据可能滞后"],"categorySuggestion":"education","subcategorySuggestion":"community_collections","lifecycleStages":["idea_requirements","product_design","backend_data","deployment_growth"],"audienceIntents":["discover","learn","build","evaluate"],"confidence":0.85,"evidence":[{"claim":"这是一个开放共建的 AI 辅助科研仓库，收集和沉淀科研全流程中的 agents、skills、workflows、tools 与最佳实践。","sourceType":"description","locator":"https://github.com/modelscope/Awesome-Vibe-Research"},{"claim":"仓库将科研生命周期拆分为 9 个阶段，包括全流程、方向扫描、文献研究、方法设计、实验执行、可视化、写作、复现发布和传播影响。","sourceType":"readme","locator":"https://github.com/modelscope/Awesome-Vibe-Research"},{"claim":"每个阶段列出典型问题和可沉淀的 AI 辅助组件类型，并收录已知最好的项目、skill 和 workflow。","sourceType":"readme","locator":"https://github.com/modelscope/Awesome-Vibe-Research"},{"claim":"仓库从「与 Agent 共事做研究」系列沙龙出发，沉淀 speaker 实践经验、观众反馈、开源项目与可复用流程。","sourceType":"readme","locator":"https://github.com/modelscope/Awesome-Vibe-Research"},{"claim":"仓库鼓励社区贡献，用户可直接在对应表格中添加缺失条目。","sourceType":"readme","locator":"https://github.com/modelscope/Awesome-Vibe-Research"}],"publishedAt":"2026-08-17T13:33:49.200Z"},"github":{"url":"https://github.com/modelscope/Awesome-Vibe-Research","ownerAvatarUrl":"https://avatars.githubusercontent.com/u/109945100?v=4","starsCount":397,"forksCount":43,"language":"Python","syncedAt":"2026-08-17T13:30:54.525Z","license":"Apache-2.0","defaultBranch":"main","archived":false,"disabled":false,"lastChangedAt":"2026-08-17T02:52:43.000Z","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"]