Borrowing it
Nothing to install: this file belongs to ZimoLiao/scholaraio. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/ZimoLiao/scholaraio/main/.claude/skills/literature-review/SKILL.mdgit clone --depth 1 https://github.com/ZimoLiao/scholaraioWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/zimoliao/scholaraio/literature-review)<a href="https://agentmods.dev/skills/zimoliao/scholaraio/literature-review"><img src="https://agentmods.dev/badge/skills/zimoliao/scholaraio/literature-review/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/zimoliao/scholaraio/literature-review"><img src="https://agentmods.dev/badge/skills/zimoliao/scholaraio/literature-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00036 | $0.04357 |
| Opus 5.5 | $0.00014 | $0.01743 |
| Sonnet 5.5 | $0.00007 | $0.00871 |
| Haiku 4.5 | $0.00004 | $0.00436 |
Grade A, and why
literature-review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 253 lines — stays where its author put it; the contents beside it link to each section on GitHub.
文献综述写作
基于工作区中的论文,撰写结构化的文献综述。
前提
用户必须指定一个 workspace(--ws NAME)。如果用户未指定:
- 运行
scholaraio ws list列出已有工作区 - 让用户选择或创建一个
综述输出写入 workspace/<name>/ 目录。
模式选择
本 skill 支持两种工作模式:
- 手动模式(默认):逐步确认大纲和章节,适合精密打磨的综述写作
- 快速模式:用户明确要求快速生成时,跳过人工确认,自动完成文献扫描、分组、分析和初稿生成
输出格式选择(必须在开始时明确确认)
在了解写作需求时,必须主动询问用户期望的输出格式:
- Markdown 速览稿:轻量、便于在线阅读、快速迭代
- LaTeX 正式学术综述:遵循
docs/writing-guide/academic-survey-writing-guide.md的规范,适用于长文深度分析,最终输出 PDF
如果用户未明确选择,默认追问一次。 若用户选择 LaTeX 正式学术综述,执行以下准备步骤:
- 必读文档:
docs/writing-guide/academic-survey-writing-guide.md(项目内置精简版流程)- 若宿主环境提供项目记忆中的学术写作指南文件(如
academic_writing_guidelines.md、academic_writing_rich_elements.md、chinese_latex_typesetting.md),可一并参考;若不可用,则以仓库内文档为准
- 硬性要求:
- 使用
\documentclass[12pt,a4paper]{article}+ctex编译;中文综述用xelatex,英文综述用lualatex - 作者字段必须包含
Claude (AI Assistant) - 所有非原创论述必须标注数字上标引用
[n],使用 BibTeX 管理;最终编译前必须执行引用闭合检查:确保所有\cite{}都有对应 bibitem,所有 bibitem 都被引用或经\nocite{}声明,无孤儿文献 - 禁止生成 AI 风格的密集 bullet point 罗列;优先使用连贯段落、批判性分析、因果推演
- 每章至少包含 2–3 种丰富元素(源论文插图 + 数据表格/伪代码算法/TikZ 概念图)
- 图注描述必须与图片实际内容严格对应;插入任何图片前必须先打开查看
- 学术综述必须包含独立的“批判性讨论”章节:对资料不对称、方法论局限、厂商宣称与独立验证的鸿沟、封闭生态权衡等进行分析
- 若涉及硬件/微架构演进,必须区分同代内的不同变体(如训练版 vs 推理版),并在结论中逐一总结每一代/变体的核心贡献;如资料充足,建议补充专利布局分析
- 正文完成后执行“补充插入源论文插图”步骤,最后编译 PDF 并做视觉检查
- 使用
快速模式(Fast Mode)
触发条件(满足任一即可):
- 用户明确说"快速生成"、"直接写"、"跳过确认"、"先给我一版"
- 用户在请求中体现出明显的时间紧迫或对完整度的要求高于精雕细琢
快速模式执行流程:
重要原则:快速模式 ≠ 低质量模式。 快速模式跳过的是与用户的人工确认环节,而不是深度分析环节。以下流程强制执行三道质量控制:Critic 骨架重组、Subagent 批判性精读模板、每章最低丰富元素。
- 自动摸底文献:运行
ws show <name>和ws search <name> "<主题>",快速扫描工作区所有论文的 L1-L2 - 自动构建骨架:
- 如已存在主题模型,参考
scholaraio topics的分组结构(注意:topics反映的是全局主库分布,若工作区与主库差异较大,应优先基于工作区论文摘要直接推断分组) - 否则直接基于摘要由 LLM 推断最优分组(主题式/方法论式/时间线式/争议式),不中断用户确认
- 如已存在主题模型,参考
- Critic 骨架重组(强制):
在写初稿前,必须由 critic subagent 审查自动生成的骨架。审查标准:
- 分组是否按"方法论演进"或"核心争议"组织,而非机械的年份/机构罗列
- 各章节标题是否暗示了明确的论点,而非模糊的"XX 年研究"
- 章节间是否有清晰的逻辑递进(问题 → 方法 → 对比 → 局限) 若骨架被判定为"清单式"而非"论证式",必须重写骨架后再进入下一步。
- 自动深度阅读(强制提示词模板):
对每章节的核心论文,启动 subagent 并行提取 L3/L4 关键发现。Subagent 必须使用
_templates/critic-reading.md中的批判性精读模板,不得简化。- 将值得保留的发现自动追加到各论文的
notes.md。优先使用 CLI:scholaraio show "<paper-id>" --append-notes "## YYYY-MM-DD | <workspace> | literature-review-fast\n- Finding 1\n- Finding 2";若需要批量自动化,再由主 agent 直接读写文件实现 - 利用
images/和公式进行多模态分析(如适用)
- 将值得保留的发现自动追加到各论文的
- 自动生成初稿(丰富元素强制):
- 若用户选择 Markdown:按骨架一次性撰写完整综述(含开头、各章节、结尾),保存为
literature-review.md。每章必须包含至少 1 个数据对比表或自动生成的概念图(scholaraio diagram --from-text ...),禁止纯文字大段罗列 - 若用户选择 LaTeX:按骨架一次性撰写
literature-review.tex,每章至少包含 2 种丰富元素(源论文插图 / 数据表格 / 伪代码算法 / TikZ 概念图),编译输出literature-review.pdf
- 若用户选择 Markdown:按骨架一次性撰写完整综述(含开头、各章节、结尾),保存为
- 保存产物:
- Markdown 模式:
- 综述正文:
workspace/<name>/literature-review.md - 章节大纲:
workspace/<name>/literature-review-outline.md - 参考文献:
workspace/<name>/references.bib
- 综述正文:
- LaTeX 模式:
- LaTeX 源文件:
workspace/<name>/literature-review.tex - 编译后的 PDF:
workspace/<name>/literature-review.pdf - 章节大纲:
workspace/<name>/literature-review-outline.md - 参考文献:
workspace/<name>/references.bib
- LaTeX 源文件:
- Markdown 模式:
- 向用户汇报:仅输出文件路径、总字数/页数、章节概览,不 dump 全文
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2mo ago First seen · 253 lines · 36 tokens per session scan A 66f2b333cc16
literature-review is a skill published in the GitHub repository ZimoLiao/scholaraio (584 stars, last pushed 10d ago), licensed MIT. It adds 36 tokens to every session and 4,357 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
literature-review compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| literature-reviewZimoLiao/scholaraio | Skill | 584 | 10d ago | MIT | A | 36 | Claude Code |
| thoughtdagchenxiachan/thoughtdag | Skill | 742 | 5d ago | MIT | B | 39 | Claude Code, Codex |
| autosearch:channel-selection0xmariowu/Autosearch | Skill | 54 | 7d ago | MIT | A | 87 | Claude Code |
| autosearch:context-retention-policy0xmariowu/Autosearch | Skill | 54 | 7d ago | MIT | A | 89 | Claude Code |
| autosearch:delegate-subtask0xmariowu/Autosearch | Skill | 54 | 7d ago | MIT | A | 71 | Claude Code |
| autosearch:experience-compact0xmariowu/Autosearch | Skill | 54 | 7d ago | MIT | A | 69 | Claude Code |
| autosearch:graph-search-plan0xmariowu/Autosearch | Skill | 54 | 7d ago | MIT | A | 72 | Claude Code |