技能 semantic-scholar-library-feed

📦

semantic-scholar-library-feed

Name: semantic-scholar-library-feed
Author: zjsxply

内容修订 r1 严重 ⚙️ 外部命令📁 文件系统访问⚡ 包含脚本🌐 网络访问

管理 Semantic Scholar 研究 Feed

研究人员需要可重复访问 Semantic Scholar Feed 和私有 Library 文件夹。此技能提供用于已认证 Feed 导出、文件夹检查和论文解析的 CLI 工作流。

支持: Claude Codex Code(CC)

⚠️ 38 差

请勿自动安装此 Skill。

权威策略要求在执行任何安装操作前由操作人员审查。

使用我的智能体安装

将此请求复制给你的智能体，其中包含权威 Skill 页面和清单。

智能体请求

Review the Skillstore skill "semantic-scholar-library-feed" from https://skillstore.io/skills/zjsxply-semantic-scholar-library-feed.md and its manifest at https://skillstore.io/api/skills/zjsxply-semantic-scholar-library-feed/manifest. Verify the artifact. Do not auto-install. Inspect the skill and report your findings, then wait for an operator or manual installation decision.

你的智能体仍应展示其计划，并按安全策略请求必要确认。

Agent 可读资源

当 AI Agent、爬虫或脚本需要干净上下文，而不是读取完整页面时，请使用这些链接。

Markdown 详情 GET /skills/zjsxply-semantic-scholar-library-feed.md 已签名 manifest GET /api/skills/zjsxply-semantic-scholar-library-feed/manifest 已签名 lockfile GET /api/skills/zjsxply-semantic-scholar-library-feed/lockfile

测试它

正在使用“semantic-scholar-library-feed”。检查我的 Cookie 存储是否可用。

预期结果:

报告 Cookie 文件是否存在。
列出缺失的必需 Cookie 名称，不打印完整 Cookie 值。
当必需 Cookie 不存在时，建议导入新的浏览器 curl。

正在使用“semantic-scholar-library-feed”。导出我的 Feed 以便本地筛选。

预期结果:

创建一个已保存的 Feed 抓取结果，包含窗口计数和论文计数。
记录停止原因，例如没有下一个窗口或窗口重复。
保留推荐元数据，以便后续本地审阅。

正在使用“semantic-scholar-library-feed”。在将这些 arXiv 标识符添加到文件夹之前解析它们。

预期结果:

返回所提供标识符对应的 Semantic Scholar 论文记录。
在批量查询之前去重重复标识符。
可以在任何文件夹更新前保存已解析记录以供审阅。

安全审计

严重

v6 • 2026/7/8 打开版本化报告

Most command-execution and reconnaissance alerts are false positives from Markdown backticks, help strings, or documented endpoint notes. The real risks are handling reusable Semantic Scholar session cookies, writing private data to local files, outbound authenticated requests, and an unvalidated page URL that can leak cookies. No prompt-injection text was found in the reviewed skill files.

已扫描文件

1,423

分析行数

复核项

已忽略误报

已确认安全问题 (3)

严重

scripts/semantic_scholar_cli.py:279-283 scripts/semantic_scholar_cli.py:333-334 scripts/ss_store.py:213-230

Unvalidated URL Can Receive Session Cookies

The CLI accepts a page URL for SSR and feed bootstrapping, then fetches it with the stored Cookie header. A malicious or mistaken URL can receive reusable Semantic Scholar session cookies.

The code attaches Cookie headers whenever a bundle is present, while ssr-dump and feed-crawl accept page URLs without origin validation. This is a direct credential disclosure path.

高

SKILL.md:39-44 references/auth-and-cookies.md:25-39 scripts/ss_store.py:169-179

Plaintext Semantic Scholar Session Cookie Storage

The workflow imports browser session cookies and writes both structured cookies and a raw Cookie header to local files. These files can be reused to access private Library and Research Feed data.

The documentation explicitly requires sid and s2 cookies, and the implementation writes the raw header to disk. The behavior is intentional but exposes high-value account credentials if file access is weak.

高

SKILL.md:84-91 references/library.md:28-43 scripts/semantic_scholar_cli.py:478-501

Authenticated Library Mutation Without Built-In Confirmation

The skill includes a folder-add command that posts to Semantic Scholar's bulk library endpoint. Without a required confirmation step, an agent can modify the user's private library once cookies are available.

The documented workflow and implementation perform an authenticated POST that changes folder membership. A dry-run flag exists, but confirmation is not enforced by the command.

能力复核项 (22)

这些是真实的本地能力，对此技能可能是预期行为，因此需要复核，但不会按已确认恶意行为计分。

高

references/auth-and-cookies.md:7

Hidden file in home directory

- `~/.auth/semantic-scholar.cookies.json`

The skill directs storage or access of raw Semantic Scholar session cookies under hidden home-directory paths. Those files contain reusable account credentials and no permission or encryption control is shown.

高

references/auth-and-cookies.md:8

Hidden file in home directory

- `~/.auth/semantic-scholar.cookie-header.txt`

高

SKILL.md:41

Hidden file in home directory

- `~/.auth/semantic-scholar.cookies.json`

高

SKILL.md:42

Hidden file in home directory

- `~/.auth/semantic-scholar.cookie-header.txt`

中

references/auth-and-cookies.md:7

Hidden file access

- `~/.auth/semantic-scholar.cookies.json`

中

references/auth-and-cookies.md:8

Hidden file access

- `~/.auth/semantic-scholar.cookie-header.txt`

中

references/auth-and-cookies.md:14

Temp directory access

- `/tmp/custom.cookies.json`

The example path is under /tmp and is used for cookie material or browser-copied curl data. Shared temporary locations can expose reusable Semantic Scholar session credentials.

中

references/auth-and-cookies.md:15

Temp directory access

- `/tmp/custom.cookie-header.txt`

The example path is under /tmp and is used for cookie material or browser-copied curl data. Shared temporary locations can expose reusable Semantic Scholar session credentials.

中

references/auth-and-cookies.md:31

Temp directory access

--curl-file /tmp/semantic-scholar-request.sh

The example path is under /tmp and is used for cookie material or browser-copied curl data. Shared temporary locations can expose reusable Semantic Scholar session credentials.

中

references/library.md:89

Temp directory access

--output /tmp/folder-13895811.json

The example writes private Semantic Scholar feed or folder data under /tmp. This is user-authorized output, but shared temporary storage can disclose private research activity.

中

references/research-feed.md:93

Temp directory access

--output /tmp/research-feed.json

The example writes private Semantic Scholar feed or folder data under /tmp. This is user-authorized output, but shared temporary storage can disclose private research activity.

中

scripts/ss_store.py:236

HTTP client library

def request(

The code performs outbound HTTP requests and attaches authenticated Semantic Scholar cookies for private endpoints. Network access is core functionality, but it creates privacy and credential-handling risk.

中

scripts/ss_store.py:284

HTTP client library

response = request(

中

scripts/ss_store.py:302

HTTP client library

response = request(

中

SKILL.md:41

Hidden file access

- `~/.auth/semantic-scholar.cookies.json`

中

SKILL.md:42

Hidden file access

- `~/.auth/semantic-scholar.cookie-header.txt`

中

SKILL.md:24

Temp directory access

--curl-file /tmp/semantic-scholar-request.sh

The example path is under /tmp and is used for cookie material or browser-copied curl data. Shared temporary locations can expose reusable Semantic Scholar session credentials.

中

SKILL.md:64

Temp directory access

--output /tmp/research-feed.json

The example writes private Semantic Scholar feed or folder data under /tmp. This is user-authorized output, but shared temporary storage can disclose private research activity.

中

SKILL.md:81

Temp directory access

--output /tmp/folder.json

The example writes private Semantic Scholar feed or folder data under /tmp. This is user-authorized output, but shared temporary storage can disclose private research activity.

低

scripts/ss_store.py:13

Python HTTP libraries

from urllib.request import Request, urlopen

低

scripts/ss_store.py:15

Hardcoded URL

SEMANTIC_SCHOLAR_ORIGIN = "https://www.semanticscholar.org"

低

scripts/ss_store.py:16

Hardcoded URL

GRAPH_API_ORIGIN = "https://api.semanticscholar.org"

风险因素

⚙️ 外部命令 (32)

📁 文件系统访问 (16)

references/auth-and-cookies.md:7 references/auth-and-cookies.md:8 references/auth-and-cookies.md:7 references/auth-and-cookies.md:8 references/auth-and-cookies.md:14 references/auth-and-cookies.md:15 references/auth-and-cookies.md:31 references/library.md:89 references/research-feed.md:93 SKILL.md:41 SKILL.md:42 SKILL.md:41 SKILL.md:42 SKILL.md:24 SKILL.md:64 SKILL.md:81

⚡ 包含脚本 (1)

scripts/semantic_scholar_cli.py:11-30

🌐 网络访问 (7)

scripts/ss_store.py:236 scripts/ss_store.py:284 scripts/ss_store.py:302 scripts/ss_store.py:13 scripts/ss_store.py:15 scripts/ss_store.py:16 scripts/ss_store.py:22

审计者: codex 查看审计历史 →

分享与引用此报告

分享版本化评估报告、中性徽章、嵌入卡片和引用。Skillstore 只报告证据，不判断该 Skill 是否安全。

打开版本化报告

复制报告链接

https://skillstore.io/skills/zjsxply-semantic-scholar-library-feed/audits/6?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 徽章

[![Skillstore security assessment](https://skillstore.io/badges/skills/zjsxply-semantic-scholar-library-feed/security.svg)](https://skillstore.io/skills/zjsxply-semantic-scholar-library-feed?utm_source=security_passport_badge)

HTML 徽章

<a href="https://skillstore.io/skills/zjsxply-semantic-scholar-library-feed?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/zjsxply-semantic-scholar-library-feed/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

嵌入卡片

<iframe src="https://skillstore.io/embed/skills/zjsxply-semantic-scholar-library-feed.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

学术引用 (APA · BibTeX · CFF)

APA 引用

zjsxply. (2026). semantic-scholar-library-feed security audit report (audit version 6) [Author version unspecified]. Skillstore. https://skillstore.io/skills/zjsxply-semantic-scholar-library-feed/audits/6

BibTeX 引用

@techreport{zjsxply-zjsxply-semantic-scholar-library-feed-2026,
  author = {zjsxply},
  title = {semantic-scholar-library-feed security audit report (audit version 6)},
  institution = {Skillstore},
  year = {2026},
  number = {6},
  url = {https://skillstore.io/skills/zjsxply-semantic-scholar-library-feed/audits/6},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "semantic-scholar-library-feed security audit report (audit version 6)"
version: "unspecified"
type: report
authors:
  - name: "zjsxply"
date-released: "2026-07-08"
url: "https://skillstore.io/skills/zjsxply-semantic-scholar-library-feed/audits/6"
identifiers:
  - type: other
    value: "skillstore:zjsxply-semantic-scholar-library-feed:audit:6"
    description: "Skillstore immutable audit report identifier"

Skillstore 评分

评分依据证据置信度: 中

架构

可维护性

内容

社区

规范符合性

你能构建什么

导出 Research Feed

收集推荐窗口，以便后续审阅、筛选或归档。

审查 Library 文件夹

列出文件夹条目，将其与参考文献列表比较，并识别缺失的论文。

解析论文标识符

在更新文件夹之前，将 arXiv 或 DOI 标识符转换为 Semantic Scholar 论文记录。

试试这些提示

检查 Cookie 就绪状态

检查我的 Semantic Scholar Cookie 存储是否已准备好。告诉我缺少哪些 Cookie，以及我应该刷新什么。

导出 Feed 历史

抓取我的 Semantic Scholar Research Feed，并在每个窗口有进展时保存。分页完成后停止。

将文件夹与参考文献列表比较

导出我的目标 Semantic Scholar 文件夹，解析我的参考文献列表中的标识符，并报告缺失哪些论文。

准备安全的文件夹更新

解析这些论文标识符，先显示拟添加到文件夹的内容，并且只有在我批准最终列表后才添加论文。

最佳实践

为 Cookie 和导出的私有数据使用由用户控制的私有目录。
在运行任何会更改 Library 的命令前，审查拟添加到文件夹的内容。
当私有端点返回 401 时，从成功的 Semantic Scholar 浏览器请求中刷新 Cookie。

避免

不要将已认证的 curl 命令粘贴到共享终端或日志中。
不要将 Semantic Scholar 会话 Cookie 存储在共享临时目录中。
在确认目标文件夹和论文列表之前，不要运行 folder-add。

常见问题

此技能需要我的 Semantic Scholar 密码吗？

不需要。它使用从已认证浏览器会话复制的 Cookie。

它能读取私有 Library 文件夹吗？

可以，只要有有效的会话 Cookie，并且该账户有权访问这些文件夹。

它会更改我的 Semantic Scholar 账户吗？

可以。folder-add 工作流可以将论文添加到 Library 文件夹。

它是否只使用官方 Semantic Scholar API？

不是。它使用公共 Graph API 以及从 Semantic Scholar 页面观察到的私有 Web 端点。

它会在每个输出中隐藏我的 Cookie 值吗？

Cookie 摘要会掩码 Cookie header，但导入的 curl 文件和 Cookie 存储仍然是敏感信息。

Claude、Codex 和 Claude Code 能使用此技能吗？

可以。报告列出了对 Claude、Codex 和 Claude Code 的支持。

开发者详情

作者

zjsxply

许可证

MIT

Skillstore 修订

版本提示

作者未声明版本。

仓库

https://github.com/zjsxply/skills/tree/main/semantic-scholar-library-feed/

引用

88a205c7f635a966e31156313b590d59007c5caa

维护新鲜度

2026/7/18

使用量

34 次下载 · 184 次浏览

文件结构

📁 references/

📄 auth-and-cookies.md

📄 library.md

📄 research-feed.md

📁 scripts/

📄 semantic_scholar_cli.py

📄 ss_store.py

📄 SKILL.md