← 返回 Skills 市场
410
总下载
0
收藏
6
当前安装
1
版本数
在 OpenClaw 中安装
/install literature-search-workflow
功能描述
Standardized literature search workflow integrating tavily-search, pubmed-database, bgpt-paper-search, openalex-database and other academic search skills. Pr...
使用说明 (SKILL.md)
Literature Search Workflow
Overview
标准化文献搜索工作流，整合多个学术搜索技能，提供 6 阶段完整文献搜索流程。
Workflow Stages
- 查询分析 (1-2 分钟): 意图识别、关键词提取
- 初步搜索 (3-5 分钟): tavily-search
- 深度搜索 (5-10 分钟): PubMed/BGPT/OpenAlex
- 结果整理 (3-5 分钟): 去重、排序、评估
- 文献获取 (5-10 分钟): web_fetch 获取全文
- 输出报告 (2-3 分钟): Markdown/BibTeX
Usage
# 学术论文搜索
python literature_search.py "主观幸福感量表 validation" --type academic_paper
# 量表工具搜索
python literature_search.py "SHS scale Chinese validation" --type scale_search
# 综述文献搜索
python literature_search.py "AI psychology review" --type review --max-results 50
# 研究方法搜索
python literature_search.py "experimental design psychology" --type methodology
Search Engines
- Primary: tavily-search
- Fallback: pubmed-database, bgpt-paper-search, openalex-database
- Specialized: research-lookup (方法学)
Scenarios
| 场景 | 引擎 | 输出 |
|---|---|---|
| academic_paper | tavily-search, pubmed-database, bgpt-paper-search | 论文列表 + 引用 |
| scale_search | tavily-search, pubmed-database | 量表验证报告 |
| review | tavily-search, bgpt-paper-search | 综述文献列表 (50 篇) |
| methodology | research-lookup, tavily-search | 方法学资源列表 |
Output Format
Markdown report with:
- 搜索摘要
- 关键文献（含 DOI）
- 来源引用
- 全文链接
- BibTeX 引用
Dependencies
Required Skills:
- tavily-search
- web_fetch
- citation-management
Optional Skills:
- pubmed-database
- bgpt-paper-search
- openalex-database
- research-lookup
API Requirements
- Tavily API: Required (1000 searches/month free)
- PubMed API: Free
- OpenAlex API: Free
- BGPT MCP: Free 50 searches
Example Output
# 文献搜索结果报告
**查询**: 主观幸福感量表 validation psychometric 2025
**搜索时间**: 2026-03-14 12:38
**数据库**: Tavily, PubMed, BGPT
**结果数量**: 10 篇
## 🔍 关键文献
### 1. Psychometric Evaluation of the Chinese Version of the SHS
**作者**: Cheung F, Lucas RE
**期刊**: Quality of Life Research
**年份**: 2014
**DOI**: 10.1007/s11136-014-0721-1
**样本量**: N = 2,635
**信度**: α = 0.82
**链接**: https://pmc.ncbi.nlm.nih.gov/articles/PMC4107280/
## 📊 文献统计
| 年份 | 数量 | 百分比 |
|------|------|--------|
| 2025 | 3 | 30% |
| 2024 | 2 | 20% |
| 2020-2023 | 3 | 30% |
## 🔗 相关链接
- [PMC 全文](https://pmc.ncbi.nlm.nih.gov/articles/)
- [ResearchGate](https://www.researchgate.net/)
Quality Assessment
| 指标 | 优秀 | 良好 | 需改进 |
|---|---|---|---|
| 相关性 | >0.8 | 0.5-0.8 | \x3C0.5 |
| 时效性 | \x3C1 年 | 1-3 年 | >3 年 |
| 权威性 | SCI/SSCI | 核心 | 普通 |
| 完整性 | 全文可获取 | 摘要 | 仅标题 |
Best Practices
- 明确查询: 使用具体关键词，包含时间范围
- 多轮搜索: 从广泛到精确，多数据库验证
- 记录来源: 保存 DOI，记录访问时间
- 质量评估: 检查研究设计、样本量、期刊质量
References
- Workflow documentation:
workspace/guides/文献搜索工作流_v1.0.md - Test report:
reports/文献搜索技能测试报告_20260314.md
安全使用建议
This skill mostly does what it says (search Tavily and assemble a report), but there are a couple of red flags you should address before installing or publishing: 1) The Python script contains a hard-coded default TAVILY_API_KEY. That can expose someone else's API access or let the script run under another account. Replace the default with no fallback, require users to set TAVILY_API_KEY, and do not commit keys to source control. 2) The skill metadata does not declare required environment variables even though the script uses one — update the manifest so users know what secrets are required. 3) Review and confirm whether additional components (web_fetch, pubmed) will be invoked by your agent; those can fetch full texts and may implicate copyright or require additional credentials. 4) Run the script in an isolated environment and rotate any exposed keys if they’ve already been used. If the embedded key was intentional for demo purposes, remove it before publishing. These inconsistencies look like sloppy engineering rather than overtly malicious behavior, but they do create a real risk of credential leakage or unintended network usage.
功能分析
Type: OpenClaw Skill
Name: literature-search-workflow
Version: 1.0.0
The skill bundle contains a hardcoded API key for the Tavily search service in `scripts/literature_search.py` (`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`), which constitutes a security vulnerability. Additionally, `UPLOAD_GUIDE.md` includes hardcoded local file paths (e.g., `C:\Users\13600\...`) from the author's environment, indicating poor hygiene. While the code logic is aligned with the stated purpose of performing academic literature searches and lacks clear malicious intent, the inclusion of hardcoded credentials warrants a suspicious classification.
能力评估
Purpose & Capability
Name, description, SKILL.md and code all describe a literature-search workflow using Tavily, PubMed, BGPT, OpenAlex, and web_fetch. The Python script implements the core search/report behavior and calls Tavily’s API, which is coherent with the stated purpose. Minor mismatch: SKILL.md lists several other back-end skills (pubmed-database, web_fetch, etc.) that the script references only as optional/placeholder; the script currently only invokes Tavily.
Instruction Scope
SKILL.md describes a 6-stage workflow and using other skills (web_fetch, citation-management). The runtime script stays mostly scoped to searching, deduping, and report generation. It does perform network calls (requests.post to Tavily) and writes a Markdown file locally; these actions are consistent with a search/report skill. Instructions do reference local paths (workspace/guides, reports/...) but these are documentation references rather than code reading arbitrary system files.
Install Mechanism
No install spec (instruction-only + small script). No downloads or package installs are performed by the skill itself; risk from installation mechanism is low.
Credentials
openclaw.skill.json declares no required environment variables, but the script reads TAVILY_API_KEY from the environment and includes a hard-coded default value ('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'). That default API key embedded in source is unexpected and disproportionate — it may leak credentials or cause requests to be billed/attributed to someone else. The skill also documents that a Tavily API is required but does not explicitly declare the environment variable name in metadata.
Persistence & Privilege
The skill is not always:true and does not request elevated/persistent system privileges. It writes output files to the current working directory (a normal, limited action) and does not modify other skills' configs or system settings.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install literature-search-workflow - 安装完成后，直接呼叫该 Skill 的名称或使用
/literature-search-workflow触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
Initial release - Standardized literature search workflow with 6 stages
元数据
常见问题
Literature Search Workflow 是什么？
Standardized literature search workflow integrating tavily-search, pubmed-database, bgpt-paper-search, openalex-database and other academic search skills. Pr... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 410 次。
如何安装 Literature Search Workflow？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install literature-search-workflow」即可一键安装，无需额外配置。
Literature Search Workflow 是免费的吗？
是的，Literature Search Workflow 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Literature Search Workflow 支持哪些平台？
Literature Search Workflow 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Literature Search Workflow？
由 Roma（@earthwalking）开发并维护，当前版本 v1.0.0。
推荐 Skills