@toolboxmd

创作者 · toolboxmd

最近更新 · 2026年8月20日

karpathy-wiki-read

审查 · 73已收录

Read protocol for the main agent. Load on demand when answering ANY user question (per Iron Rule 4 in `using-karpathy-wiki/SKILL.md`). Defines the deterministic 6-step orientation ladder for finding wiki coverage of a question, when to inline-read vs spawn an Explore subagent vs

OpenAgentSkill 信任评分
73/100

仅限沙盒

质量67/100
审计82/100
Stars101
Verified installs0

安装目标

Codex 安装提示词

Install the "karpathy-wiki-read" agent skill from https://github.com/toolboxmd/karpathy-wiki/tree/main/skills/karpathy-wiki-read. Read its SKILL.md or equivalent instructions first, install only the files needed for this workspace, and summarize any required setup before using it. Skill purpose: Read protocol for the main agent. Load on demand when answering ANY user question (per Iron Rule 4 in `using-karpathy-wiki/SKILL.md`). Defines the deterministic 6-step orientation ladder for finding wiki coverage of a question, when to inline-read vs spawn an Explore subagent vs fall through to web search, and the cite contract every wiki-grounded answer must satisfy. After the install attempt, report the result to https://www.openagentskill.com/api/agent/outcome with POST JSON {"event_id":"install_<unique-id>","skill_slug":"toolboxmd-karpathy-wiki-read","task":"Install karpathy-wiki-read","agent":"codex","outcome":"success","install_used":true}. Replace event_id with a unique value and outcome with success or failed. Report success only after the skill is installed and a minimal verification passes.

供给资产档案

研究与知识工作

Deep research, source comparison, literature review, RAG, knowledge search, and reports.

浏览赛道

场景

研究 Agent

I need my agent to research a topic, compare sources, and produce a concise report.

适配 Agent

Claude Code + CLI + Codex

适用于 Codex、Claude Code、Cursor、CLI 或自定义 Agent。

安装

就绪

npx skills add toolboxmd/karpathy-wiki --skill karpathy-wiki-read

维护状态

新鲜

距上次推送 3 天

风险

可安全尝试

Quality score needs review

GitHub 质量

101

67/100 质量 · 81/100 信任

覆盖标签

研究研究 Agentagent-skill

审查说明

Quality score needs review · Stars/forks activity: 101 stars, 24 forks; issue activity unavailable in current metadata

Agent 采用评分卡

一眼查看信任、审计与安装准备度

这些分数综合公开仓库元数据、OpenAgentSkill 审查信号、维护新鲜度与安装准备度。它用于候选筛选,不替代人工审查。

质量

有潜力
67

有用的候选项,但采用前应与替代方案比较。

信任

仅限沙盒
73

有用但信任信号不足或混杂的候选项。在结果闭环证明任务匹配前,请保持在隔离工作区内使用。

审计

可安全尝试
82

对安装准备度、安全元数据、维护情况与采用风险的机器可读审查。

OpenAgentSkill 信任评分 v5

安装前需人工审查

仅在沙盒中运行,并在用于真实工作前比较接近的替代方案。

CodexClaude CodeCursorOpenAgentSkill CLI

Stars

101 个 GitHub Stars

仓库活跃度

101 个 Star,24 个 Fork

维护状态

距上次推送 3 天

许可证

MIT

安装

npx skills add toolboxmd/karpathy-wiki --skill karpathy-wiki-read

安装安全性

标准软件包或运行时安装路径

权限范围

数据库访问

Agent 结果

暂未有 Agent 结果数据

文档

README/SKILL.md 上下文充分

风险摘要

生产前审查

  • Quality score needs review
  • Stars/forks activity: 101 stars, 24 forks; issue activity unavailable in current metadata

安装准备度

安装路径可用

  • 安装路径可用
  • 仓库证据可用
  • 已声明许可证
  • 暂无 Agent 验证结果证据

Agent 可读元数据

这个 Skill 的机器可读决策数据。

使用此区块或内嵌 JSON 判断 Agent 是否应安装该 Skill、选择替代方案,或先请求人工审查。

View technical data+

适用任务

  • 研究 Agent 工作流
  • Claude Code 团队
  • builders willing to evaluate younger projects
  • 检索来源

适用 Agent

CodexClaude CodeCursorOpenAgentSkill CLICLI

安装决策

命令
npx skills add toolboxmd/karpathy-wiki --skill karpathy-wiki-read
策略
审查
人工审查

信任与风险

信任
73/100
审计
82/100
风险级别
可安全尝试

结果闭环

端点
/api/agent/outcome
事件 ID
resolve
结果
5

安装命令

npx skills add toolboxmd/karpathy-wiki --skill karpathy-wiki-read

不适用场景

  • 需要厂商支持 SLA 的团队
  • 没有内部安全审查的高合规环境
  • 当前元数据中未发现重大风险信号
  • Quality score needs review
  • Stars/forks activity: 101 stars, 24 forks; issue activity unavailable in current metadata

Agent 安全 v2

62/100 · 安装前审查

已审查并附权限说明审查

可用候选,但 Agent 在安装前应展示权限与审计说明。

在真实工作区安装前需要人工批准。

通过 API 解析

网络访问

Skill 可能访问远程页面、API、仓库或外部服务。

文件系统访问

Skill 可能读取或写入项目文件、文档、生成产物或本地工作区状态。

数据库访问

Skill 可能检查 Schema、查询数据库或处理持久化存储。

  • Quality score needs review

Agent 解析计划

让 Agent 在安装前验证匹配度。

Resolve API 返回首选 Skill、替代方案、安全策略、审计说明、安装目标和可直接执行的提示词,无需抓取此页面。

打开文本计划

Agent 应检查

  • 从 Resolve API 检查任务匹配与替代方案。
  • 检查审计评分、信任评分和安全策略警告。
  • 检查 Codex、Claude Code、Cursor 或 CLI 的安装目标兼容性。

复制提示词

Task: Use karpathy-wiki-read in this workspace.
Resolve first: https://www.openagentskill.com/api/agent/resolve?task=Use%20karpathy-wiki-read%20for%20an%20agent%20workflow&agent=codex&max_risk=medium
Review install handoff: https://www.openagentskill.com/api/skills/toolboxmd-karpathy-wiki-read/install
Install command: npx skills add toolboxmd/karpathy-wiki --skill karpathy-wiki-read
Before running it, summarize audit warnings, required permissions, and the fallback skill if install is risky.

Agent 交接

把安装路径交给 Agent,而不是再给一个目录页。

通过公开安装端点获取命令、安全清单、目标提示词和该 Skill 的规范链接。

打开安装 API

Agent 提示词

Use karpathy-wiki-read for this task. Review https://www.openagentskill.com/api/skills/toolboxmd-karpathy-wiki-read/install, then install with: npx skills add toolboxmd/karpathy-wiki --skill karpathy-wiki-read

Registry 元数据

用于自动选择 Skill 的 Agent 可读档案。

本页通过 Registry API 提供相同的决策、信任、审计、场景和安装信号,让 Agent 无需抓取界面即可排序。

打开 Manifest

适配 Agent

69/100

研究 Agent

平台

Claude Code

审计报告

可安全尝试 · 82/100

对安装准备度、安全元数据、维护情况与采用风险的机器可读审查。

查看审计报告查看评估报告

Agent 决策面板

Fallback candidate for Research agents

先用此 Skill 做原型验证,并保留备选方案。

69
就绪度
原型验证
阶段

栈中角色

备选候选

主要匹配

研究 Agent

信任标签

先做原型验证

安装路径

命令已就绪

适用场景

  • 研究 Agent 工作流
  • Claude Code 团队
  • builders willing to evaluate younger projects

证据

  • 仓库近期活跃
  • 已提供安装命令或 GitHub 仓库
  • 67/100 质量档案
  • 9 个 OpenAgentSkill 交互事件

先审查

  • 当前元数据中未发现重大风险信号

实施路径

  1. 1在沙盒 Agent 中安装它,并端到端完成一次研究 Agent任务。
  2. 2Compare output quality, latency, and failure behavior against at least one alternative.
  3. 3Promote it into production only after reviewing repository permissions, license, and maintenance signals.

信任档案

仅限沙盒

有用但信任信号不足或混杂的候选项。在结果闭环证明任务匹配前,请保持在隔离工作区内使用。

73
OpenAgentSkill 信任评分

GitHub 采用度

信息

101 个 GitHub Stars

Star/Fork 活跃度

检查

101 个 Star,24 个 Fork; 当前元数据中没有议题活跃度信息

近期维护

通过

距上次推送 3 天

许可证清晰度

通过

MIT

积极信号

  • AI 审查已通过
  • 安装路径可用
  • 仓库证据可用
  • 近期维护的仓库
  • 安装命令未发现明显高风险模式
  • 结果闭环已就绪,但需要首次真实 Agent 运行

安装前审查

  • Quality score needs review
  • Stars/forks activity: 101 stars, 24 forks; issue activity unavailable in current metadata
  • 暂未有真实 Agent 结果报告
  • 无人值守安装前需要人工审查

建议操作

仅在沙盒中运行,并在用于真实工作前比较接近的替代方案。

质量档案

有潜力 适用于 Agent 工作流的候选

有用的候选项,但采用前应与替代方案比较。

67
GitHub Stars
101
新鲜度
3 天前
安装就绪
许可证
MIT

工作流匹配

在这些场景使用此 Skill

工作流匹配

加入完整工作流

替代方案短名单

安装前对比

可能适合该任务的相近 Skill。

对比全部

概览

--- name: karpathy-wiki-read description: | Read protocol for the main agent. Load on demand when answering ANY user question (per Iron Rule 4 in `using-karpathy-wiki/SKILL.md`). Defines the deterministic 6-step orientation ladder for finding wiki coverage of a question, when to inline-read vs spawn an Explore subagent vs fall through to web search, and the cite contract every wiki-grounded answer must satisfy. ---

# karpathy-wiki read

You loaded this skill because the user asked a question (per Iron Rule 4 in the loader: NO ANSWERING ANY USER QUESTION WITHOUT ORIENTING FIRST). Your job is to determine whether the wiki covers the answer, retrieve the coverage, and either cite-and-answer from the wiki or fall through to web search — without ever pre-classifying the question as "wiki-irrelevant."

## The 6-step ladder

Every step is deterministic. There is no agent judgement at branch points; each branch is gated on a counted or boolean condition.

### Step A — Orient

Read these files in order:

1. `<wiki>/schema.md` — taxonomy, conventions, thresholds. 2. The relevant `<wiki>/<category>/_index.md` for the question's apparent topic. If the question crosses categories or you cannot tell which category applies, read `<wiki>/index.md` (the root MOC).

If you have already oriented earlier in this session, skip to Step B — the schema and index content are already in your working memory. Orientation is once per session.

### Step B — Count signal-matching candidates

Extract the question's signal terms: meaningful nouns, proper-noun phrases, technical terms, version numbers, tool names. Skip stopwords ("the", "what", "how", "do", "is").

Walk the relevant `_index.md` (already in memory from Step A). A page is a candidate if ANY signal term:

- Substring-matches its title (case-insensitive), OR - Matches a tag (exact, case-insensitive), OR - Appears in its one-line summary.

Count the candidates. Branch on count:

- **0 candidates** → Step F (cold-result path) - **1-5 candidates** → Step C (inline read) - **6+ candidates** → Step E (Explore subagent)

The threshold is 5/6, not your judgement. Do not "feel" your way to a different branch.

### Step C — Inline read

Read all candidate pages in full. After reading, ask exactly:

> Does the union of these pages contain every claim my answer would make?

Branch on the answer:

- **YES** → cite + answer (see "Cite contract" below). Done. - **NO** → Step D (gap-fill via web search).

### Step D — Gap-fill via web search

The wiki covered part of the answer; the rest needs fresh information. For each claim your answer would make that the wiki did NOT cover:

1. Web search for the specific claim (use `WebFetch` or `WebSearch`). 2. Cite the web source(s) for the gap-filled claim.

Compose the answer using BOTH the wiki citations (for what it covered) and the web citations (for what it didn't).

After answering, **always** write a capture noting the gap. The wiki should grow toward questions it failed to answer fully. The capture's title is the question; the body documents what the wiki had and what was missing. Use `karpathy-wiki-capture/SKILL.md` for the capture authoring procedure.

### Step E — Explore subagent

Spawn an Explore subagent with this prompt shape:

> Question: `<the user's question, verbatim>` > > Wiki path: `<wiki absolute path>` > > Run the orient procedure (read schema.md and the relevant _index.md), identify candidate pages by signal-term match, read all candidates in full (no cap on page count — your context is isolated), and return a synthesis. > > The synthesis must: > - Cite every page it draws from (path + one-line relevance note per page). > - Cover the question completely (no truncation for brevity if the answer is genuinely long). > - Be as terse as possible while preserving every wiki-specific claim, decision, and contradiction. Terseness is a property of the writing, not a target word count. > - If the wiki does not cover the question, say so explicitly and return an empty page-list. Do NOT attempt to fall through to web search yourself; the main agent handles that.

When the subagent returns:

- **If it returned a synthesis with cited pages** → use the synthesis as the answer's basis. Add conversational framing as needed. Cite the pages the subagent cited. - **If it reported "wiki does not cover"** → Step D (web search + capture the gap).

The subagent dispatch maps to the patterns in your `~/.claude/CLAUDE.md` Task Delegation rule: this is a context-isolation case where the main agent needs the synthesis, not the page contents.

### Step F — Cold result (no candidates)

Zero signal matches in `_index.md`. The wiki has no coverage for this question.

1. Web search for the answer (`WebFetch` or `WebSearch`). 2. Cite the web source(s) in the answer. 3. **Always** write a capture so the next session is not cold for this topic. The capture's title is the question; the body documents what was found and where (the URLs you searched). Use `karpathy-wiki-capture/SKILL.md`.

A cold result is not a failure — it is the wiki telling you it has a gap, and you closing the gap.

## Cite contract

Every wiki-grounded answer must include citations the user can verify. Format:

- For each wiki page you drew from: cite as `<wiki>/<category>/<page>.md` followed by a brief relevance note OR a short quote (≤2 lines). - For each web source you drew from: cite as `[<title>](<url>)`. - Citations appear inline near the claim they support, not in a footer block, so the user can check each claim against its source.

If your answer makes a claim with no citation, that claim came from training data — flag it explicitly: *"(from training data; not in wiki)."* This is the **only** case where uncited claims are allowed, and they must be flagged.

## What goes in this skill — and what doesn't

This skill covers the read-from-wiki protocol only. It does NOT cover:

- Capture authoring (when a question reveals a wiki gap, the capture flow lives in `karpathy-wiki-capture/SKILL.md`). - Page-format conventions, manifest protocol, validator contract — those live in `karpathy-wiki-ingest/SKILL.md` and apply only to the spawned ingester. - Iron laws and the orient-first rule itself — those live in the loader (`using-karpathy-wiki/SKILL.md`).

If you find yourself needing capture or ingest mechanics while running this protocol, load the appropriate sibling skill — do not improvise.

技术详情

版本
1.0.0
许可证
MIT
最近更新
2026年8月20日
发布时间
2026年8月20日

决策摘要

备选候选

69
就绪
原型验证
阶段

仓库近期活跃

审计

安装审查

安装与采用审查

82
可安全尝试
安全性
87/100
维护状态
100/100
安装
92/100
打开完整审计查看评估报告

Agent 验证证据

Agent 验证证据

来自解析、审查、安装和一次小范围运行后的结果报告。

0
已验证
Needs first agent run自动安装: 先审查最近: 未知
成功率
近期失败
结果
0
输出质量
失败
0
不相关
0
安装次数
0
风险拦截
0
需要配置
0
生产环境
0

暂时没有 Agent 结果数据。首次 Agent 执行可以通过 /api/agent/outcome 报告成功、需要设置、风险拦截、失败或不相关。

安装

加入 Agent 工作流

免费且开源. 在生产 Agent 中安装前请先审查报告。

增长闭环

分享工具包

X

为 karpathy-wiki-read 准备的场景化草稿,可手动发布到 X。

策展说明
karpathy-wiki-read: Read protocol for the main agent. Load on demand when answering ANY user question (per Iron R...

101 stars

https://www.openagentskill.com/skills/toolboxmd-karpathy-wiki-read?ref=x
打开 X 草稿
可选:带安装命令的回复
Listing + install path for karpathy-wiki-read:
https://www.openagentskill.com/skills/toolboxmd-karpathy-wiki-read?ref=x

Install: npx skills add toolboxmd/karpathy-wiki --skill karpathy-wiki-read
打开回复草稿

收录来源

Registry 收录

可认领

此列表来自公开来源,维护者认领获批前不会标记为官方。

创作者
toolboxmd
收录方
OpenAgentSkill 社区索引

归属链接指向公开仓库或创作者主页。创作者可认领列表以更新所有权信号。

认领此 Skill

所有者认领

认领此 Skill 页面

这条 Registry 收录 列表归属于 toolboxmd,但尚未标记为官方。认领后可增加已验证所有者信号,使后续发布、安装和审计更新更值得信赖。

创作者外链工具包

将证据徽章加入你的 README

在开发者评估仓库的位置展示规范页面、当前信任与审计信号,以及真实的 Agent 验证证据。

[![Listed on OpenAgentSkill](https://www.openagentskill.com/api/badge/toolboxmd-karpathy-wiki-read?metric=listed&label=Listed)](https://www.openagentskill.com/skills/toolboxmd-karpathy-wiki-read)
[![OpenAgentSkill Trust](https://www.openagentskill.com/api/badge/toolboxmd-karpathy-wiki-read?metric=trust&label=Trust)](https://www.openagentskill.com/skills/toolboxmd-karpathy-wiki-read)
[![OpenAgentSkill Audit](https://www.openagentskill.com/api/badge/toolboxmd-karpathy-wiki-read?metric=audit&label=Audit)](https://www.openagentskill.com/skills/toolboxmd-karpathy-wiki-read/audit)
[![Agent Proven](https://www.openagentskill.com/api/badge/toolboxmd-karpathy-wiki-read?metric=proven&label=Agent%20Proven)](https://www.openagentskill.com/skills/toolboxmd-karpathy-wiki-read)

作者

T

toolboxmd

@toolboxmd

平台适配

健康信号

GitHub Stars
101
质量评分
37/100
最近 GitHub 推送
2026年8月20日
框架提示
未知
OpenAgentSkill 浏览量
9
复制安装命令
0
跳转点击
0

社区信号

告诉我们这个 Skill 是否对你的 Agent 工作流有帮助。汇总反馈会持续改善排序。

信任与安全

仅限沙盒

73
  • GitHub 采用度101 个 GitHub Stars信息
  • Star/Fork 活跃度101 个 Star,24 个 Fork; 当前元数据中没有议题活跃度信息检查
  • 近期维护距上次推送 3 天通过
  • 许可证清晰度MIT通过
  • README/SKILL.md 完整度元数据包含足够的用法与工作流上下文通过
  • 依赖与运行时风险公开元数据中未发现主要依赖风险提示通过