← 返回 Skills 市场
Paper Reproduction by Python
作者
Celynn Moonlight
· GitHub ↗
· v1.0.1
· MIT-0
315
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install paper-repro-python
功能描述
This skill should be used when the user asks to "reproduce a paper", "implement paper methods in Python", "extract paper content to Markdown", or works on pa...
安全使用建议
This skill appears to do what it says: extract paper sources and implement reproducible code. Before installing or invoking it, (1) ensure the execution environment has the expected tooling (LaTeX toolchain, pandoc/pdftotext, tesseract or other OCR, and the Python packages you expect) because the SKILL.md assumes these but does not install them; (2) review the workspace for any sensitive files (API keys, private datasets, credentials) because the skill instructs the agent to read files in the working folder and will save logs/checkpoints there; run it in a sandboxed or dedicated reproduction repository if you have confidentiality concerns; (3) if you need network isolation or specific tools, provide them explicitly or modify the workflow to declare required binaries; and (4) verify outputs before sharing externally (logs may contain debug output or data you did not intend to expose).
功能分析
Type: OpenClaw Skill
Name: paper-repro-python
Version: 1.0.1
The skill bundle 'paper-repro-python' provides a structured workflow for an AI agent to extract content from scientific papers (TeX or PDF) and implement them in Python. The instructions in SKILL.md focus on modular code design, logging, result verification, and bilingual documentation without any indicators of data exfiltration, malicious execution, or prompt injection.
能力评估
Purpose & Capability
The name/description (paper reproduction in Python) matches the SKILL.md: it specifies extracting TeX/PDF content, building a reproduction plan, and producing Python code and logs. However, the workflow implicitly expects external tooling (LaTeX tooling, PDF extractors, OCR, pandoc, Python libraries) but declares no required binaries or install steps; the absence of declared dependencies is a documentation gap the user should address.
Instruction Scope
The instructions explicitly direct the agent to inspect and parse project files in the working folder (.tex, .bib, .md, .json, images, PDFs) and to save logs/outputs. That is consistent with reproduction work, but means the agent will access any files present in the workspace (which may include secrets or unrelated data). The SKILL.md does not instruct any network exfiltration or use of unexpected external endpoints.
Install Mechanism
This is an instruction-only skill with no install spec or code files, which is the lowest install risk. Note: the instructions mention actions (OCR, TeX parsing, PDF extraction) that typically require third-party binaries or Python packages, but none are declared or installed by the skill.
Credentials
The skill requests no environment variables or credentials. It does direct the agent to read files and write logs/checkpoints in the workspace; these outputs could include sensitive information depending on the repository contents, but no unrelated credentials are requested by the skill itself.
Persistence & Privilege
The skill does not request always:true and does not modify other skills or system-wide settings. It will create project-local outputs (logs, results, checkpoints) as part of normal operation.
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 paper-repro-python - 安装完成后，直接呼叫该 Skill 的名称或使用
/paper-repro-python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
v1.0.1
**Added support for user-preprocessed documents and implemented experiment logging, output persistence, and results verification.**
- Expanded input priority: Now supports user-preprocessed files (Markdown, JSON, paper images) as secondary sources, before falling back to PDF.
- New logging and persistence rules: All experiments must save logs, outputs, and configuration snapshots for reproducibility.
- Added mandatory results verification: Requires direct comparison between reproduction outputs and paper-reported results, with detailed discrepancy analysis.
- Refined workflow steps and clarified source extraction order.
- README and documentation requirements remain, but now assume richer provenance for extracted content.
v1.0.0
Initial release of paper-repro-python skill:
- Guides end-to-end reproduction of scientific papers focusing on TeX-first extraction, modular Python implementation, and bilingual documentation.
- Prioritizes TeX source extraction with structured PDF fallback; ensures faithful, non-inventive content conversion to Markdown.
- Enforces validation of extraction completeness and explicit reporting of any missing or unresolved content.
- Details planning and modular Python implementation best practices, emphasizing reproducibility and minimal dependencies.
- Mandates bilingual (English and Chinese) README files with complete paper metadata, aligned content, and embedded images.
- Specifies exact output contract separating verbatim source content from implementation/engineering notes, with reproduction status reporting.
元数据
常见问题
Paper Reproduction by Python 是什么？
This skill should be used when the user asks to "reproduce a paper", "implement paper methods in Python", "extract paper content to Markdown", or works on pa... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 315 次。
如何安装 Paper Reproduction by Python？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install paper-repro-python」即可一键安装，无需额外配置。
Paper Reproduction by Python 是免费的吗？
是的，Paper Reproduction by Python 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Paper Reproduction by Python 支持哪些平台？
Paper Reproduction by Python 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 Paper Reproduction by Python？
由 Celynn Moonlight（@celynnmoonlight）开发并维护，当前版本 v1.0.1。
推荐 Skills