← 返回 Skills 市场
jirboy

Research Review Assistant

作者 JIRBOY · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
61
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install research-review-assistant
功能描述
自动检索并结构化总结科研文献,支持相关性评估、多轮优化及SCI格式综述草稿生成。
安全使用建议
What to consider before installing: - Capability mismatch: the SKILL.md/README advertise multi-source retrieval (arXiv, PubMed, Google Scholar) but the included Python code only implements arXiv queries. Ask the author or inspect the rest of the code to confirm support for PubMed/Google Scholar before relying on those features. - Packaging oddness: package.json (an npm manifest) lists Python libraries and a different version than the registry metadata. This is likely a packaging mistake but could cause confusion about how to install dependencies; verify dependency installation instructions and run in an isolated environment. - Network behavior: the script performs outbound HTTP requests to arXiv (public API). This will leak search queries and requested paper metadata to that remote service — expected for this skill, but worth noting for sensitive queries. The script uses an http arXiv endpoint; consider network encryption policies in your environment. - Source provenance: the skill source/homepage and owner details are minimal/unknown. If you need stronger assurance, request the upstream repository, tests, or an author identity and a signed release. - Next steps to raise confidence: confirm full implementation (PubMed/Google Scholar), correct packaging metadata or provide a proper install spec, and review the remainder of the Python file (truncated in the manifest) for any unexpected external endpoints or file I/O. If you cannot obtain those, run the skill in a sandboxed environment and avoid providing private credentials or sensitive project text as inputs.
功能分析
Type: OpenClaw Skill Name: research-review-assistant Version: 1.0.0 The skill is a legitimate academic research assistant designed to search, analyze, and summarize papers from arXiv. The core logic in `review_generator.py` uses the official arXiv API and performs basic keyword-based analysis to generate Markdown reports. No evidence of data exfiltration, unauthorized execution, or malicious prompt injection was found in the code or instructions.
能力评估
Purpose & Capability
SKILL.md and README claim multi-source retrieval (arXiv/PubMed/Google Scholar) and features like SCI-format output and fund-support modules. The included Python code implements a working arXiv search and analysis pipeline but I only see arXiv-specific network calls; no PubMed/Google Scholar implementations are present in the visible code. Also package/metadata versions differ (registry metadata 1.0.0 vs package.json 2.0.0). These mismatches are unexplained and reduce confidence in the stated capabilities.
Instruction Scope
The runtime instructions in SKILL.md are narrowly scoped to research review workflows and do not instruct the agent to read arbitrary files, system credentials, or other unrelated resources. The skill is described as a compatibility shim forwarding to a unified 'research' entrypoint, which is a reasonable integration note.
Install Mechanism
There is no install spec (instruction-only) which is low-risk. However, package.json is present despite this being a Python script; it lists Python packages (requests, beautifulsoup4) under an npm manifest, which is inconsistent and could indicate sloppy packaging or mis-publishing. That mismatch is suspicious because it may lead to confusion about how to install dependencies or hide additional install steps elsewhere.
Credentials
The skill requests no environment variables, no credentials, and references no config paths. The code uses only outward HTTP(S) requests to public APIs (arXiv) and does not access secrets, so requested privileges are proportionate to its stated purpose.
Persistence & Privilege
The skill is not marked always:true and does not request persistent system-wide settings. There is no code in the visible files that writes to other skills' configs or requests elevated privileges.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install research-review-assistant
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /research-review-assistant 触发
  4. 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
- This skill has been integrated into the unified research entry; please use the research skill for literature review tasks. - The original functionality remains for backward compatibility and will automatically forward requests to the new entry point. - Updated usage guide provided for both new (research review) and old (review) commands. - Maintains all core features: automated retrieval, structured summary, iterative refinement, and multiple output formats. - Added configuration and migration instructions for a smooth transition.
元数据
Slug research-review-assistant
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Research Review Assistant 是什么?

自动检索并结构化总结科研文献,支持相关性评估、多轮优化及SCI格式综述草稿生成。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 61 次。

如何安装 Research Review Assistant?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install research-review-assistant」即可一键安装,无需额外配置。

Research Review Assistant 是免费的吗?

是的,Research Review Assistant 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Research Review Assistant 支持哪些平台?

Research Review Assistant 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 Research Review Assistant?

由 JIRBOY(@jirboy)开发并维护,当前版本 v1.0.0。

💬 留言讨论