research-workflow
Skill2 starsChanged 30 days agoMITView on GitHub
科研知识工作流总入口(Zotero + Obsidian + AI 中间层):多源文献获取与精读、Obsidian 知识库维护(ingest/query/lint/体检)、科研出图(Origin)与三处归档、文档/表格/PDF 转 Markdown。当要处理文献、维护或查询知识库、出图或归档图件、读 PDF/文档/表格、任务收尾做经验总结时使用。Entry point for a Zotero + Obsidian + AI research knowledge workflow.
13 other projects use a version of this.
Adopted by
Most recently taken up by
- frontdesk-alhouse/alto-ostheir own version, first seen Nov 8, 2026
- GEUConn/research-idea-workflow-skilltheir own version, first seen Nov 1, 2026
- crce-ux/research-workflowtheir own version, first seen Oct 20, 2026
- microsoft/ArgusAgenttheir own version, first seen Oct 16, 2026
- sherifican/Agent-FleetOpstheir own version, first seen Oct 7, 2026
- yoloshii/gigaxity-deep-researchtheir own version, first seen Sep 30, 2026
Other versions13 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 13
- microsoft/ArgusAgent★ 140same name
- yoloshii/gigaxity-deep-research★ 61same name
- RipeMangoBox/BITE★ 60same name
- TenureAI/PhD-Zero★ 52same name
- sherifican/Agent-FleetOps★ 34same name
- GEUConn/research-idea-workflow-skill★ 17same name
- arnaudgelas/K-Ops★ 1same name
- frontdesk-alhouse/alto-os★ 1same name
- mageyuki/skills★ 1same name
- crce-ux/research-workflow★ 0same name
3 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository4 other files it gives its agents.
- doc-to-mdSkill
- figure-pipelineSkill
- vault-opsSkill
- zotero-opsSkill
The file
Files
科研知识工作流(Zotero + Obsidian + AI)
采集层 AI 中间层(AI 助手) 沉淀层
┌────────────────────┐ ┌───────────────────────────┐ ┌────────────────────────┐
│ Zotero(入口) │ 读取 │ 精读 / 出图 / 转换 / 归档 │ 写入 │ Obsidian vault(大脑) │
│ Connector·WoS·知网 │ ───→│ 按本仓库各子技能执行 │ ───→│ 卡片·概念·项目·问答 │
│ 主题 collections │ ←───│ │ ←───│ index·log·矛盾清单 │
└────────────────────┘ RPC └───────────────────────────┘ 回链 └────────────────────────┘
ZCode Bridge ──┘ ↑ LabTalk/COM └─ ZotMoov → Raw/文献PDF(只读)
出图(择优)─→ PNG/PDF/TIF ─→ 三处归档
PDF/docx/表格 ── doc-to-md ───→ Markdown ──→ AI 阅读(数值核校仍以原始文件为准)分工:用户提供资料、提问、把关方向;AI 负责一切维护。本仓库就是 AI 的操作手册。
首次使用:先完成本机配置
- 把
LOCAL-CONFIG.example.md复制为LOCAL-CONFIG.md,与用户一起填好各项(该文件不进 git) - 本仓库所有
<vault>、<项目目录>、<git-name>、<cnki-profile>等占位符的取值,一律来自LOCAL-CONFIG.md;AI 使用前先读它,缺失时主动询问用户 - 检查各软件版本,按
README.md「版本差异与兼容」与各子技能的「版本差异」节适配——Zotero 7/9、Obsidian 版本、Origin 2017 与 2021+ 的能力边界都不同
子技能路由(按任务进入对应 SKILL.md)
| 任务 | 进入 |
| Zotero 条目/分类/附件操作、多源文献获取(WoS / 知网 / OA 与文献传递)、文献入库与精读 | skills/zotero-ops/SKILL.md |
| Obsidian 知识库维护(Ingest/Query/Lint/新任务、页面与 git 规范) | skills/vault-ops/SKILL.md |
| 科研出图(Origin)与图件三处归档 | skills/figure-pipeline/SKILL.md |
| PDF/图片/Office/表格 → Markdown(读文档前必做) | skills/doc-to-md/SKILL.md |
路径相对本技能目录;也可把任一子目录单独装进 AI 的 skills 目录独立使用。
全局红线(每次任务都适用)
- 数值红线:转换 / 摘录出的文本只用于阅读;数据校核、定案值引用必须回原始文件(Excel 等)读精确值。
- 绘图工具:不再强制 Origin(2026-10-14 变更)——按“好用 + 出图质量”择优(Origin / Python 绘图库 / 其他均可)。
- 归档红线:每次出图必须三处归档——资料包原位 +
<项目目录>\<项目>\图\<类型>\+<vault>\项目\<项目>\附件\,并在试验页嵌入;同步 index/log + git。缺一不可。 - 经验总结红线:每个任务收尾必须做经验总结(有效做法 / 踩坑与修复 / 迁移提示),按去向分流:工具·网络·自动化坑 → AI 记忆系统;领域知识与方法论 → vault;两条都追加
log.md。 - git 红线:知识库提交用
git -c user.name="<git-name>" -c user.email="<git-email>" commit(取值见 LOCAL-CONFIG),不改动全局 git 配置——每台机器既有的协作身份不受影响。 - Raw 只读:
Raw/层与一切原始资料绝不修改;文献 PDF 的移动/重命名一律走 Zotero(不手动)。 - 项目盘:新项目放空间充裕的盘(如
D:\Projects\),系统盘只放临时脚本与小文件。
版本差异总览
关键差异与完整说明见 README.md「版本差异与兼容」;各子技能内另有「版本差异」节:
- Zotero:本地 API / Connector 需 7+;Bridge 实测于 9.0.x(7+ 理论兼容),7/8 与 9 的插件安装方式不同
- Obsidian:Bases(
.base)需 1.9+,旧版用 Dataview 或纯表格替代 - Origin:能力边界随版本不同——2017 不能脚本化的功能在 2021+ 部分可用,以本机实测为准
- MinerU:本仓库按 4.x 子命令式 CLI 编写(
mineru parse ...)
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

