No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →
by MrGeDiao · Codex Skill · ★ 100
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
🔒 Is paper-reading-zh safe to install? View the security audit →
paper-reading-zh 给 AI 加一套论文阅读的证据规则：未核验的不补，读不到的不编，比较前先对口径。 An evidence-rule pack for AI-assisted paper reading. Docs and outputs are in Chinese by design. 是一个中文论文精读规则包，面向 Codex、Claude Code、Claude Project 和 ChatGPT Project。它不是论文翻译器，也不是文献管理器；它的目标是让 AI 读论文时少一点顺滑猜测，多一点可复查的证据标注。 证据边界：venue、年份、CCF、代码链接未核验就写“未核验”；实验数字必须锚定原文 Table / Figure 或具体段落。 防顺滑编造：公式抽取乱码不硬补，图表读不到不描述，只能读到摘要时明确写“仅基于摘要”。 跨论文口径审计：比较多篇论文前先检查数据集、指标定义、模型规模、训练预算和测试 setting。 双入口复用：同一套规则同时提供 Agent Skill（CLI / Agent 环境）和 Web Prompt Kit（网页端项目）。 和直接把论文丢给 AI 的区别 经常用 AI 读论文的人大多见过这些行为：查不到 venue 就补一个像样的，只读到摘要却写出全文精读，两篇论文口径不同也直接判胜负。这套规则把它们逐条挡住： 左列是常见失败模式的示意，不是对某个具体产品的实测记录；右列是规则的硬性要求。 输出长什么样 默认输出是约 2000 到 3500 中文字的中等深读，骨架固定：关键词（不超过 5 个）、一段话总结（不超过 150 字）、论文基本信息（标题 / venue/年份 / 链接 / 任务领域 4 项）、核心问题与贡献、方法深度解析、实验与结果、批判性讨论。 证据标注落在输出里是这样的（依
| Stars | 100 |
| Forks | 1 |
| Language | Python |
| Category | Codex Skill |
| License | MIT |
| Quality Score | 68.4626720571692/100 |
| Last Updated | 2026-09-14 |
| Created | 2026-05-27 |
| Platforms | claude-code, codex, python |
| Est. Tokens | ~14k |
These tools work well together with paper-reading-zh for enhanced workflows:
Looking for a paper-reading-zh alternative? If you're comparing paper-reading-zh with other codex skill tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.
Bring Agent Skills to Any AI Agent and Coding Agent — via CLI or MCP. Manage once, serve anywhere.
60 production-ready AI marketing skills (29 VN + 30 Global) for Claude Code, ChatGPT, Gemini & Copilot. 4 regi
Claude Code skill that transforms vague prompts into structured, expert-level prompts using 7 research-backed
Better 'AskUserQuestion' - A lightweight MCP server/OpenCode plugin/Agent Skills + CLI tool that allows your L
中华术数 MCP 排盘服务：八字・紫微斗数・六爻・梅花易数・大六壬・奇门遁甲｜Deterministic Chinese-metaphysics chart calculator for AI agents (MCP).
114 source-backed /goal contracts for coding agents, plus 200 separated seed patterns.
Explore other popular codex skill tools:
paper-reading-zh is Evidence-aware Chinese paper reading rules for Codex, Claude Code, Claude Project, and ChatGPT Project. It is categorized as a Codex Skill with 100 GitHub stars.
paper-reading-zh is primarily written in Python. It covers topics such as academic-papers, agent-skills, chatgpt.
You can find installation instructions and usage details in the paper-reading-zh GitHub repository at github.com/MrGeDiao/paper-reading-zh. The project has 100 stars and 1 forks, indicating an active community.
paper-reading-zh is released under the MIT license, making it free to use and modify according to the license terms.
The top alternatives to paper-reading-zh on Agent Skills Hub include skillport, fullstack-mkt-skills, prompt-architect. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: