Scansci Pdf logo

Scansci Pdf

Community
YuanyuanMa03
scansci-pdf

学术论文获取全家桶,基于 scansci-pdf MCP server。多路由下载 PDF(开放获取 → 出版商 API → 机构通道并行竞速)、批量清单下载、引文导出(BibTeX/RIS/EndNote)、推送 Zotero、机构登录(WebVPN/CARSI/EZProxy)、环境诊断排障。当用户要下载论文(单篇或批量)、导出引文、推送 Zotero、配置机构访问,或下载遇到 Cloudflare/验证页/代理问题时使用。用户仅讨论 PDF 工具对比选型而无需获取文献时不要使用。

Overview

PublisherYuanyuanMa03
Repositoryacademic-research-skills
Skill namescansci-pdf
Stars
76
Forks
7
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by YuanyuanMa03 on GitHub. Read the source before you install it.

Installation

Install the Scansci Pdf AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/YuanyuanMa03/academic-research-skills/tree/main/plugins/scansci-pdf/skills/scansci-pdf
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/YuanyuanMa03/academic-research-skills.git /tmp/academic-research-skills
mkdir -p .claude/skills
cp -r /tmp/academic-research-skills/plugins/scansci-pdf/skills/scansci-pdf .claude/skills/scansci-pdf
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Scansci Pdf in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Scansci Pdf on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Scansci Pdf is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

ScanSci PDF — 论文获取 / 引文 / Zotero

本技能桥接开源项目 scansci-pdf(MIT)的 MCP server:20+ 数据源并行竞速,任一成功立即返回;每篇结果都报告实际来源。它取代了本仓库早期的 *-download / *-export / zotero-csl 技能——平台站内检索仍用 cnki-search、pm-search 等技能,拿到条目后交给本技能完成下载、引文与推送。

前置条件 (Prerequisites)

MCP server 需要 Python ≥ 3.11 和 scansci-pdf 命令。安装本插件后若工具列表中没有 scansci_pdf_* 工具,按以下步骤引导(steps 按顺序执行,遇到问题转排障表):

bash
# 1. 检查 Python 版本(需要 >= 3.11)
python3 --version

# 2a. 推荐:独立 venv 安装(不污染用户环境)
python3 -m venv ~/.scansci-pdf/venv && ~/.scansci-pdf/venv/bin/pip install scansci-pdf

# 2b. 或一行式(需要 uv)
uvx --from scansci-pdf scansci-pdf check

# 3. 环境自检,按 readiness 分级处理
scansci-pdf check
  • ready:全部就绪,直接走工作流。
  • partial:核心可用、可选组件缺失(浏览器后端、Tor 等),按提示安装,不阻塞下载。
  • limited:缺关键依赖,先完成安装再继续。

可选浏览器后端强烈建议安装:没有它,下载只覆盖"OA 仓库直链"型文章,出版商官网 OA 页面(Wiley/RSC 等)与机构登录都不可用。无机构通道时执行 pip install "scansci-pdf[camoufox]"(或 scansci-pdf[patchright])可显著扩大可下载范围;机构登录/Cloudflare 场景必须有它。

安装后需重启宿主(或重载插件)让 MCP server 生效。CLI(scansci-pdf get/search/batch/...)与 MCP 工具能力等价,MCP 不可用时用 CLI 兜底。注意:CLI 模式没有引文格式化命令(citation 仅 MCP 提供)——CLI 兜底时由你(agent)用检索结果的 title/authors/year/doi 字段直接构造 BibTeX,或提示用户启用 MCP server。

工作流 (Workflow)

快速工作流(三条技能串起完整课题)

① 站内/平台检索(cnki-search、pm-search 等)拿条目(标题/作者/DOI)
② 本技能:get 下载 PDF(--md 转 Markdown 供精读)→ 引文/Zotero
③ nature-* 系列:精读笔记 → 综述 → 开题 → 文章撰写

场景 → 工具路由

用户想做什么用什么
下载单篇(DOI / arXiv / 文章页 URL)scansci_pdf_download(可带 si=true 拉补充材料、md=true 转 Markdown)
批量下载一份清单(xlsx/csv/txt/bib/APA)scansci_pdf_parse_list → scansci_pdf_batch_download
验证 / 嗅探一批标识符scansci_pdf_prepare_queue(action="verify");找 OA 副本用 action="resolve_oa"
跨源检索文献(含作者、被引过滤)scansci_pdf_search(可选 scansci_pdf_sciverse 语义检索)
导出引文scansci_pdf_citation(bibtex / ris / endnote)
推送到 Zoteroscansci_pdf_zotero_push(走 Zotero Web API,需要 Zotero API key——在 zotero.org 设置页创建;不是本地 Zotero 程序,也不是旧的 Connector 端口)
机构登录 / 渠道状态scansci_pdf_login、scansci_pdf_channel_status、scansci_pdf_schools
环境或下载排障`scansci_pdf_diagnostics(check=health
模糊引用("Wang 2023 CRISPR 那篇")先 scansci_pdf_search 拿候选让用户确认,不要硬猜 DOI

检索源选择提示:scansci_pdf_search 是宽泛发现引擎(13 源),长复合查询的相关性排序偏宽——主题精确检索优先用平台技能(生物医学 → pm-search,其走 PubMed 精确 API;站内 → cnki-search 等),本技能检索用于跨源发现与兜底,结果需做相关性过滤。

策略选择(必读)

用户指定了来源 = 只按用户的来,调度不得扩大授权范围:

用户说的是策略
"从 Sci-Hub 下载"scihub_only
"优先 scihub"scihub_first(OA 仍竞速,更快时最终来源可能是 OA)
"只要免费合法的"legal_only(排除 Sci-Hub / LibGen)
未指定fastest(默认,全源并行竞速)

机构通道两条事实(失败大多源于此)

  1. 登录必须用工具自己打开的浏览器:scansci_pdf_login 打开的可见浏览器窗口才是 cookie 抓取上下文;用户自己浏览器里的登录态、无头模式,都拿不到。
  2. 出口 IP 决定机构权限:出版商按请求源 IP 判权益。系统级代理会把出口 IP 移出校园网段,静默作废 IP 认证。出版商报"无权限"时,先读错误页上它看到的 IP,再下"学校没订阅"的结论。工具只认 network_proxy / SCANSCI_PDF_PROXY,忽略 HTTP_PROXY。

批量(>300 篇)必须分批;连续 Cloudflare 拦截时停下来排障,不要硬冲。

输出 (Output)

  • 每篇下载结果必含:文件路径、实际来源(OA / 出版商 API / 机构通道 / 灰色源)、耗时;失败结果带 error_type、action、source_failures(逐渠道明细)。
  • 引文结果按请求格式输出;批量结果落盘 batch_results.json(含失败清单,可用 batch --retry 补齐)。
  • 诊断结果为分级报告(ready / partial / limited),每项失败附下一步动作。

排障 (Error Handling)

症状首个动作
工具列表没有 scansci_pdf_*命令未安装或插件未启用 → 走前置条件安装引导,或用 CLI 兜底
Sci-Hub 返回 Cloudflare/Turnstilescansci_pdf_diagnostics(check="health") 查浏览器后端;或改走机构路径
日志出现 "browser not available, skipping browser strategy"可选浏览器后端未装 → pip install "scansci-pdf[camoufox]" 后重试;该症状意味着出版商官网型 OA 文章会下载失败
所有源超时scansci_pdf_diagnostics(check="network") 查代理与 DNS
Elsevier 只回 1 页预览API key 无效或无权限 → 重新 setup 并 --validate
结果带 config_needed / ask_user_email缺用户决策(如 Unpaywall 需真实邮箱)→ 问用户,不要替用户猜
机构登录页卡住session 过期 → 重新 scansci_pdf_login
get 返回的路径不在指定 --output 目录缓存命中:返回的是此前下载的文件。换源重下或改输出位置先清缓存(rm -f <out>/.doi_index.json && rm -rf ~/.scansci-pdf/cache/*)
CLI 模式需要 BibTeX/引文citation 仅 MCP 提供 → agent 用检索结果字段(title/authors/year/doi)构造 BibTeX,或启用 MCP server
403 / 超时 ≠ 无全文多为反爬或网络受限,交给浏览器后端或代理,不要直接判死

合规边界 (Operating Boundaries)

  • 灰色源(Sci-Hub / LibGen)与反爬路由仅在用户明确选择时使用,且限于用户有权且适用规则允许的场景;用户选择 legal_only 后调度不得回退。
  • 保留用户指定的来源策略,并在每个结果中如实报告实际来源——不静默换渠道。
  • API key、cookie、代理凭据、机构信息视为机密,不在回复中复述。
  • 打开交互式登录、导入 cookie、更改访问配置前先征得用户同意。

Frequently asked questions

What does the Scansci Pdf AI skill do?

学术论文获取全家桶,基于 scansci-pdf MCP server。多路由下载 PDF(开放获取 → 出版商 API → 机构通道并行竞速)、批量清单下载、引文导出(BibTeX/RIS/EndNote)、推送 Zotero、机构登录(WebVPN/CARSI/EZProxy)、环境诊断排障。当用户要下载论文(单篇或批量)、导出引文、推送 Zotero、配置机构访问,或下载遇到 Cloudflare/验证页/代理问题时使用。用户仅讨论 PDF 工具对比选型而无需获取文献时不要使用。

Why use Scansci Pdf on TypingMind?

Because you install it once and use it with any model. Scansci Pdf is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Scansci Pdf in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/YuanyuanMa03/academic-research-skills/tree/main/plugins/scansci-pdf/skills/scansci-pdf. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Scansci Pdf?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Scansci Pdf?

As many as you like. As long as a model supports skills, you can use Scansci Pdf with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Scansci Pdf AI skill free?

Yes. It is published on GitHub by YuanyuanMa03 under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇