← Back to Skills Marketplace
paudyyin

文献综述自动器

by paudyyin · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
68
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install literature-review-automator
Description
自动检索学术文献(Semantic Scholar/arXiv/CrossRef),进行相关性筛选、主题聚类分析,并生成综述草稿(支持本地模板或大模型润色)。适用于快速了解某个研究方向的前沿动态。
README (SKILL.md)

文献智能检索与综述生成器

功能描述

根据用户提供的研究主题(关键词或问题),自动在多个公共学术数据库(Semantic Scholar、arXiv、CrossRef)中检索相关文献,经相关性筛选、去重、主题聚类分析后,生成结构化综述文档(含摘要、分主题讨论、趋势分析、参考文献列表)。

使用示例

  • "帮我检索近3年关于'联邦学习在工业视觉中的应用'的文献,写一篇综述。"
  • "追踪'大模型微调技术'的最新进展,给我一份热点概览。"
  • "我想了解'柔性机器人'领域的研究趋势,输出综述报告。"

执行流程

该技能按以下阶段顺序执行,每个阶段失败时会自动重试或给出明确错误提示。

1. 请求解析

  • 从用户输入中提取:核心关键词、时间范围、最大文献数、输出格式等。
  • 若用户未提供,使用配置文件中的默认值。
  • 自动生成英文同义词和字段变体。

2. 文献检索

  • 并发调用以下API(优先使用Semantic Scholar,返回结果快且丰富):
    • Semantic Scholar APIhttps://api.semanticscholar.org/graph/v1/paper/search
    • arXiv APIhttp://export.arxiv.org/api/query
    • CrossRef APIhttps://api.crossref.org/works
  • 检索参数:关键词、出版年份范围、按相关性排序、每页最多100条。
  • 合并去重(基于DOI或标题相似度)。

3. 相关性筛选(本地)

  • 对摘要和标题进行关键词匹配或TF-IDF向量化,计算与用户关键词的余弦相似度。
  • 按相似度降序排序,结合引用数加权,保留前 max_papers 篇。
  • human_review_papers = true,则生成候选列表并询问用户是否采纳或手动剔除。

4. 主题聚类分析(可选,本地)

  • 如果安装了 sentence-transformersbertopic,自动对摘要进行BERTopic聚类。
  • 否则使用基于关键词的简单分组。
  • 输出:每个聚类的主题词、代表性论文、论文数量。
  • 同时统计每年发文量,生成趋势文本描述。

5. 综述草稿生成

支持两种模式:

模式A(纯本地,零API费用)

  • 根据聚类结果按大纲自动生成结构化综述,包含:
    • 摘要(基于检索到的论文数量和主要聚类方向)
    • 主要研究方向(按主题分类,附带代表性论文关键信息)
    • 研究趋势与挑战(年度发文趋势、热门期刊、潜在研究机会)
    • 参考文献列表(含DOI和链接)

模式B(大模型润色,可选)

  • 将本地生成的草稿和论文摘要输入大模型,生成更流畅的综述文本。

文件结构

lit_review/
├── skill.md           # 技能元数据定义
├── lit_review.py      # 核心实现(主入口)
├── config.json        # 配置文件
├── requirements.txt   # Python依赖
├── README.md          # 使用说明
└── test.py            # 测试脚本
Usage Guidance
This skill appears safe for normal public literature-review tasks. Before installing, use a virtual environment, be careful with unpinned Python dependencies, keep LLM writing disabled for sensitive research topics unless you trust the provider, protect any LLM API key, and verify the generated citations and summaries.
Capability Analysis
Type: OpenClaw Skill Name: literature-review-automator Version: 1.0.0 The skill is a legitimate tool for academic literature review, fetching data from Semantic Scholar, arXiv, and CrossRef. However, it is classified as suspicious due to several security vulnerabilities. Specifically, the LLM API endpoint (llm_api_base) is user-configurable in 'main.py' and 'config.json', which could allow for the exfiltration of the 'llm_api_key' if redirected to a malicious server. Additionally, the tool is vulnerable to indirect prompt injection because it fetches untrusted paper abstracts from external APIs and incorporates them directly into LLM prompts within 'main.py' without sanitization. No evidence of intentional malice or data exfiltration to unauthorized domains was found.
Capability Tags
cryptorequires-sensitive-credentials
Capability Assessment
Purpose & Capability
The described behavior—searching Semantic Scholar/arXiv/CrossRef, ranking papers, clustering topics, and drafting a review—is coherent with the skill purpose. The main user-visible sensitivity is that research queries and paper metadata leave the local machine.
Instruction Scope
The instructions are bounded to literature retrieval, analysis, and review generation. No artifact-backed prompt override, hidden role change, destructive workflow, or unsafe autonomous action was found.
Install Mechanism
The registry says there is no install spec, while the skill and requirements file document Python package installation. This is common for Python skills but users should install dependencies cautiously, preferably in a virtual environment.
Credentials
Network access to public academic APIs and optional LLM endpoints is proportionate to the stated purpose. No broad local file indexing, protected-path writes, or unrelated system access is shown.
Persistence & Privilege
There is no evidence of background persistence or privilege escalation. The optional LLM API key is stored/read through configuration, so users should protect that config if they enable LLM writing.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install literature-review-automator
  3. After installation, invoke the skill by name or use /literature-review-automator
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
学术文献自动检索(Semantic Scholar/arXiv/CrossRef),相关性筛选、主题聚类分析、综述草稿生成
Metadata
Slug literature-review-automator
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is 文献综述自动器?

自动检索学术文献(Semantic Scholar/arXiv/CrossRef),进行相关性筛选、主题聚类分析,并生成综述草稿(支持本地模板或大模型润色)。适用于快速了解某个研究方向的前沿动态。 It is an AI Agent Skill for Claude Code / OpenClaw, with 68 downloads so far.

How do I install 文献综述自动器?

Run "/install literature-review-automator" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is 文献综述自动器 free?

Yes, 文献综述自动器 is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does 文献综述自动器 support?

文献综述自动器 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created 文献综述自动器?

It is built and maintained by paudyyin (@paudyyin); the current version is v1.0.0.

💬 Comments