Karpathy Wiki
基于 Karpathy LLM Wiki 模式,为科研工作建立和维护持久化知识库。 当用户提到建立知识库、LLM Wiki、Karpathy 方法、Obsidian 知识管理、论文管理、 研究笔记、摄入论文、维护 Wiki 时使用。 Skill: Karpathy Wiki Owner: zhangyanbo2007 Summary: 基于 Karpathy LLM Wiki 模式,为科研工作建立和维护持久化知识库。 当用户提到建立知识库、LLM Wiki、Karpathy 方法、Obsidian 知识管理、论文管理、 研究笔记、摄入论文、维护 Wiki 时使用。 Tags: karpathy:1.1.3, knowledge-base:1.1.3, latest:1.1.3, llm:1.1.3, research:1.1.3, wiki:1.1.3 Version history: v1.1.3 | 2026-04-22T01:48:14.157Z | user 修复 log.md 链接路径:从 wiki/ 子目录使用相对路径,VSCode Preview 可点击跳转 v1.1.2 | 2026-04-22T01:32:40.036Z | user lo
Rank
62
Safety
84
Downloads
1.1k
Updated
Oct 11, 2026
Version
1.1.3
Source
CLAWHUB
About
What it does, and when to use it.
Capability contract not published. No trust telemetry is available yet. 1.1K downloads reported by the source. Last updated 10/11/2026.
Avoid when
- Contract metadata is missing or unavailable for deterministic execution.
Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing
Public facts
Every fact links back to the source it came from.
- Vendor
- Clawhubvendor · observed Oct 11, 2026
- Protocol compatibility
- OpenClawcompatibility · observed Oct 11, 2026
- Adoption signal
- 1.1K downloadsadoption · observed Oct 11, 2026
- Latest release
- 1.1.3release · observed Apr 22, 2026
- Handshake status
- UNKNOWNsecurity
Install and run
Setup complexity: low.
clawhub skill install s1743fjt8p75r07r4a35af525183p97g:karpathy-wiki-cn- Setup complexity is LOW. This package is likely designed for quick installation with minimal external side-effects.
- Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data.
Contract: missing
curl -s "https://www.xpersona.co/api/v1/agents/clawhub-zhangyanbo2007-karpathy-wiki-cn/snapshot"
Documentation
CLAWHUB
25,821 characters of source documentation, loaded on request.
Extracted files
3 files captured from the source.
SKILL.md
--- name: karpathy-wiki description: 基于 Karpathy LLM Wiki 模式,为科研工作建立和维护持久化知识库。 当用户提到建立知识库、LLM Wiki、Karpathy 方法、Obsidian 知识管理、论文管理、 研究笔记、摄入论文、维护 Wiki 时使用。 --- # Karpathy LLM Wiki — 科研知识库 ## 触发条件 - 建立 LLM 知识库 / Karpathy Wiki - 管理论文、研究笔记 - 摄入(Ingest)新的论文或资料 - 对知识库提问(Query) - 维护知识库健康(Lint) ## 核心思想 和传统 RAG 不同——RAG 每次提问都从零开始检索、重新拼凑答案——Karpathy 模式让 LLM **持续构建和维护一个持久化的 wiki**。新资料进来后,LLM 不是简单索引它,而是提取关键信息、整合到现有 wiki 中、更新实体页、修订主题摘要、标记新旧资料之间的矛盾。知识被编译一次,然后持续保持最新,而不是每次查询都重新推导。 **Wiki 是一个持久化、不断复利的产物。** 交叉引用已经在那里了,矛盾已经被标记,综合摘要已经反映了你读过的所有内容。 你从不自己写 wiki 页面——LLM 写和维护所有内容。你负责提供资料、提问、引导方向。LLM 负责摘要、交叉引用、归档和簿记。 ## 系统架构 ### 三层结构 - **Raw sources(原始资料)**:你收集的论文、文章、数据文件。不可变——LLM 只读不改,这是真相来源。 - **The wiki(知识库)**:LLM 生成的 markdown 文件集合。LLM 全权负责这个层,你只读不写。 - **The schema(配置)**:你正在读的 CLAUDE.md/SKILL.md。它告诉 LLM wiki 的结构、约定和工作流。这个文件不是一成不变的——你和 LLM 随着使用共同演化它。 ### 目录结构 ``` research-wiki/ ├── CLAUDE.md # Schema:告诉 LLM 如何维护 Wiki ├── raw/ # 原始资料(不可变,只进不改) │ ├── papers/ # PDF 论文 │ ├── articles/ # Markdown/Web 文章 │ ├── images/ # 图片资源 │ └── data/ # 数据文件 ├── wiki/ # LLM 维护的 Wiki(LLM 全权负责) │ ├── index.md # 内容目录(每个页面 + 一行摘要) │ ├── log.md # 操作日志(只追加) │ ├── overview.md # 研究主题概述 │ ├── entities/ # 研究实体页 │ ├── concepts/ # 核心概念页 │ ├── papers/ # 论文摘要页 │ └── synthesis/ # 综合分析(Query 产物存回) └── assets/ # 图片附件 ``` > 页面格式和目录结构只是起点,不是铁律。根据你的研究领域和偏好,和 LLM 一起调整。 ## 核心操作 ### 1. Ingest(摄入新资料) 当用户提供论文、文章或其他资料时: 1. 将资料保存到 `raw/` 对应子目录 2. 读取资料内容,与用户讨论关键要点 3. 在 `wiki/papers/` 创建论文摘要页 4. 在 `wiki/concepts/` 更新或创建相关概念页 5. 在 `wiki/entities/` 更新或创建相关实体页(学者、机构、方法名等) 6. 更新 `wiki/index.md` 添加新页面索引 7. 追加 `wiki/log.md` 记录本次摄入(格式:`## [YYYY-MM-DD] ingest | 论文标题`)。文件引用使用 Markdown 标准链接 `[文件名](路径)`,确保 VSCode Markdown Preview 中可点击 8. 向用户汇报:触及了多少个 Wiki 页面 一篇资料可能触及 10-15 个 wiki 页面。建议逐篇摄入、保持参与——阅读摘要、检查更新、引导 LLM 强调重点。当然也可以批量摄入,取决于你的工作风格。 ### 2. Query(对知识库提问) 当用户提出问题: 1. 先读 `wiki/index.md` 找到相关页面 2. 深入阅读相关 Wiki 页面 3. 综合回答,引用具体来源 4. **如果回答有持续价值(对比分析、主题梳理、新发现),存回 `wiki/synthesis/` 作为新页面**——你的探索也在为知识库积累知识,不应消失在聊天记录里 5. 追加 `wiki/log.md` 记录本次查询 Query 产物可以是多种格式:markdown 页面、对比表、演示文稿(Marp)、图表(matplotlib)。好的分析应该像 ingested sources 一样复利增长。 ### 3. Lint(健康检查) 当用户要求维护或定期检查: 1. 扫描所有 Wiki 页面,查找: - 页面间的矛盾或不一致 - 过时声明(被新论文/来源推翻的旧观点) - 孤立页面(没有入站链接) - 文中提到但未建立页面的重要概念 - 缺失的交叉引用 - 可以填补的数据缺口 2. 自动修复能发现的问题 3. 报告需要人工判断的问题 4. 建议值得追问的新问题和值得补充的新来源 5. 追加 `wiki/log.md` 记录维护结果 ## 页面格式规范 以下格式是建议起点,可根据领域调整。 ### 论文摘要页(wiki/papers/) ```markdown --- title: "论文完整标题" authors: ["作者1", "作者2"] year: 2024 venue: "会议/期刊名" tags: [标签1, 标签2] ingested: 2026-04-21 --- ## 核心问题 本文要解决什么问题? ## 方法 核心方法是什么? ## 关键发现 - 主要结果1 - 主要结果2 ## 与Wiki的关系 [[相关概念页]] [[其他相关论文]] ## 笔记 任何额外的观察和笔记 ``` ### 概念页(wiki/concepts/) ```markdown --- topic: 概念名称 tags: [类别标签] created: 2026-04-21 --- ## 定义 这个概念是什么? ## 关键要点 - 要点1 - 要点2 ## 相关论文 [[论文1]] [[论文2]] ## 相关概念 [[相关概念]] ## 笔记 补充信息 ``` ## index.md 和 log.md **index.md** 是内
_meta.json
{
"ownerId": "kn75tf09qgpgt30ht1z56qxft982k6gc",
"slug": "karpathy-wiki-cn",
"version": "1.1.3",
"publishedAt": 1776822494157
}skill-card.md
## Description: 基于 Karpathy LLM Wiki 模式,为科研工作建立和维护持久化知识库。 This skill is ready for commercial/non-commercial use. ## Publisher: [zhangyanbo2007](https://clawhub.ai/user/zhangyanbo2007) ### License/Terms of Use: MIT-0 ## Use Case: Researchers and developers use this skill to create, query, and maintain a persistent Markdown research wiki from papers, articles, notes, and related sources. ### Deployment Geography for Use: Global ## Known Risks and Mitigations: Risk: The skill can create or update persistent local research-wiki files, so incorrect summaries, cross-references, or logs could become durable knowledge-base content. Mitigation: Review generated diffs, keep raw sources immutable, and use version control before accepting wiki changes. Risk: The bundled arXiv downloader can retrieve PDFs from the network and save them locally. Mitigation: Run the downloader only when recent arXiv PDFs are expected, choose an explicit output directory, and review downloaded files before ingesting them. ## Reference(s): - [Karpathy LLM Wiki gist](https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f) - [qmd local Markdown search engine](https://github.com/tobi/qmd) - [Marp Markdown presentation tool](https://marp.app/) - [Karpathy Wiki ClawHub page](https://clawhub.ai/zhangyanbo2007/skills/karpathy-wiki-cn) ## Skill Output: **Output Type(s):** [text, markdown, code, shell commands, configuration, guidance] **Output Format:** [Markdown, shell commands, and local file updates] **Output Parameters:** [1D] **Other Properties Related to Output:** [May create or update local raw/ and wiki/ files when the user asks the agent to ingest sources, answer durable research questions, or run wiki maintenance.] ## Skill Version(s): 1.1.3 (source: server release metadata) ## Ethical Considerations: Users should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
AionUi
Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!
activepieces
AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents
cherry-studio
AI productivity studio with smart chat, autonomous agents, and 300+ assistants.
CopilotKit
The Frontend for Agents & Generative UI. React + Angular
Machine-readable data
The same record, as JSON, for agents and crawlers.
{
"facts": [
{
"factKey": "vendor",
"category": "vendor",
"label": "Vendor",
"value": "Clawhub",
"href": "https://clawhub.ai/zhangyanbo2007/skills/karpathy-wiki-cn",
"sourceUrl": "https://clawhub.ai/zhangyanbo2007/skills/karpathy-wiki-cn",
"sourceType": "profile",
"confidence": "medium",
"observedAt": "2026-10-11T11:35:50.294Z",
"isPublic": true
},
{
"factKey": "protocols",
"category": "compatibility",
"label": "Protocol compatibility",
"value": "OpenClaw",
"href": "https://www.xpersona.co/api/v1/agents/clawhub-zhangyanbo2007-karpathy-wiki-cn/contract",
"sourceUrl": "https://www.xpersona.co/api/v1/agents/clawhub-zhangyanbo2007-karpathy-wiki-cn/contract",
"sourceType": "contract",
"confidence": "medium",
"observedAt": "2026-10-11T11:35:50.294Z",
"isPublic": true
},
{
"factKey": "traction",
"category": "adoption",
"label": "Adoption signal",
"value": "1.1K downloads",
"href": "https://clawhub.ai/zhangyanbo2007/karpathy-wiki-cn",
"sourceUrl": "https://clawhub.ai/zhangyanbo2007/karpathy-wiki-cn",
"sourceType": "profile",
"confidence": "medium",
"observedAt": "2026-10-11T11:35:50.294Z",
"isPublic": true
},
{
"factKey": "latest_release",
"category": "release",
"label": "Latest release",
"value": "1.1.3",
"href": "https://clawhub.ai/zhangyanbo2007/karpathy-wiki-cn",
"sourceUrl": "https://clawhub.ai/zhangyanbo2007/karpathy-wiki-cn",
"sourceType": "release",
"confidence": "medium",
"observedAt": "2026-04-22T01:48:14.157Z",
"isPublic": true
},
{
"factKey": "handshake_status",
"category": "security",
"label": "Handshake status",
"value": "UNKNOWN",
"href": "https://www.xpersona.co/api/v1/agents/clawhub-zhangyanbo2007-karpathy-wiki-cn/trust",
"sourceUrl": "https://www.xpersona.co/api/v1/agents/clawhub-zhangyanbo2007-karpathy-wiki-cn/trust",
"sourceType": "trust",
"confidence": "medium",
"observedAt": null,
"isPublic": true
}
],
"events": [
{
"eventType": "release",
"title": "Release 1.1.3",
"description": "修复 log.md 链接路径:从 wiki/ 子目录使用相对路径,VSCode Preview 可点击跳转",
"href": "https://clawhub.ai/zhangyanbo2007/karpathy-wiki-cn",
"sourceUrl": "https://clawhub.ai/zhangyanbo2007/karpathy-wiki-cn",
"sourceType": "release",
"confidence": "medium",
"observedAt": "2026-04-22T01:48:14.157Z",
"isPublic": true
}
]
}Record generated Oct 11, 2026.
