← 返回 Skills 市场
298
总下载
1
收藏
1
当前安装
1
版本数
在 OpenClaw 中安装
/install research-paper-compare
功能描述
Compare research papers by retrieving full PDFs from titles, URLs, or files and synthesizing differences, strengths, weaknesses, and evidence-backed trade-offs.
安全使用建议
This skill appears coherent with its purpose, but check a few operational points before installing: 1) Confirm the agent environment has a safe, local PDF extraction capability (or understand which external service would be used) because the skill assumes full-text extraction but declares no tooling. 2) Verify network policies and sandboxing—the skill will download PDFs from arbitrary URLs and may attempt to fetch from publisher sites; ensure downloads are allowed and handled safely. 3) If you expect paywalled PDFs, decide whether you'll provide credentials; the skill does not request them and will stop if full text is unavailable. 4) Confirm the agent will not upload PDF contents to third-party services unless you approve that data flow. If these operational details are acceptable, the skill is internally consistent with its description.
功能分析
Type: OpenClaw Skill
Name: research-paper-compare
Version: 1.0.0
The skill is a legitimate tool designed to compare research papers by retrieving full PDF texts from academic sources like arXiv and OpenReview. The instructions across all files (SKILL.md, input-handling.md, evidence-policy.md) focus on data extraction, evidence quality, and structured synthesis without any indicators of malicious intent, data exfiltration, or unauthorized execution.
能力评估
Purpose & Capability
The name and description (compare research papers by obtaining full PDFs and synthesizing differences) match the instructions: the SKILL.md explicitly describes resolving titles/URLs/local PDFs, downloading PDF full text, normalizing records, and performing structured comparison. No unrelated credentials, binaries, or installs are requested, so required capabilities are proportionate to the stated purpose.
Instruction Scope
The SKILL.md requires web search and downloading arbitrary PDF URLs to obtain full text (expected for this purpose). It correctly enforces a policy to stop if full PDF text is unobtainable. However, it does not specify operational guardrails: it doesn't state whether PDF parsing is local vs sent to third-party services, whether downloads are sandboxed, or how to handle paywalled content or copyright-restricted PDFs. It also assumes the runtime can extract readable text from PDFs but does not declare the needed extraction tooling.
Install Mechanism
Instruction-only skill with no install spec and no code files—this is low-risk from an installation perspective (nothing will be written to disk or downloaded as part of a packaged install).
Credentials
The skill declares no environment variables or credentials (proportional). That said, in practice retrieving PDFs from publisher sites may require credentials or access tokens; the skill does not request these and states it will stop if full text isn't accessible. Verify whether your platform's agent will attempt to use any agent-level credentials for retrieval or will upload PDFs to external extraction services—those behaviors are not described by the skill.
Persistence & Privilege
always is false and the skill does not request persistent presence or elevated privileges. It does not attempt to modify other skills or system-wide settings based on the provided files.
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 research-paper-compare - 安装完成后，直接呼叫该 Skill 的名称或使用
/research-paper-compare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 of paper-compare skill with core framework for side-by-side research paper analysis.
- Added structured input handling guide for DOIs, URLs, search, and PDF files (`input-handling.md`).
- Defined standard comparison schema and output patterns for clarity and consistency.
- Introduced evidence policies and memory dimensions referenced in outputs.
- Implemented rule: always use PDF full text before analysis; block and report any papers missing full PDF access.
- Removed legacy metadata file (`_meta.json`).
元数据
常见问题
Research Paper Compare 是什么？
Compare research papers by retrieving full PDFs from titles, URLs, or files and synthesizing differences, strengths, weaknesses, and evidence-backed trade-offs. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 298 次。
如何安装 Research Paper Compare？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install research-paper-compare」即可一键安装，无需额外配置。
Research Paper Compare 是免费的吗？
是的，Research Paper Compare 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Research Paper Compare 支持哪些平台？
Research Paper Compare 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 Paper Compare？
由 quqxui（@quqxui）开发并维护，当前版本 v1.0.0。
推荐 Skills