Karpathy LLM Wiki
Astro-Han/karpathy-llm-wiki
A skill your agent uses when building or maintaining a personal LLM-powered knowledge base.
Co-creates a personal investment-research LLM Wiki (Karpathy's pattern) by interviewing the user for THEIR OWN analysis framework, not a template — a living CLAUDE.md of pure markdown + wikilinks…
Install LLM Wiki Setup
$ npx skills add daymade/claude-code-skills --skill llm-wiki-setup -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install daymade/claude-code-skills llm-wiki-setup --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/daymade/claude-code-skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/llm-wiki-setup .claude/skills/llm-wiki-setup && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "llm-wiki-setup" agent skill from https://github.com/daymade/claude-code-skills/tree/main/llm-wiki-setup into .claude/skills/llm-wiki-setup/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki-setup", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/daymade/claude-code-skills/tree/main/llm-wiki-setupType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add daymade/claude-code-skills --skill llm-wiki-setup -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install daymade/claude-code-skills llm-wiki-setup --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/daymade/claude-code-skills.git skills-src && mkdir -p .agents/skills && cp -r skills-src/llm-wiki-setup .agents/skills/llm-wiki-setup && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "llm-wiki-setup" agent skill from https://github.com/daymade/claude-code-skills/tree/main/llm-wiki-setup into .agents/skills/llm-wiki-setup/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki-setup", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add daymade/claude-code-skills --skill llm-wiki-setup -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install daymade/claude-code-skills llm-wiki-setup --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/daymade/claude-code-skills.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/llm-wiki-setup .cursor/skills/llm-wiki-setup && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "llm-wiki-setup" agent skill from https://github.com/daymade/claude-code-skills/tree/main/llm-wiki-setup into .cursor/skills/llm-wiki-setup/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki-setup", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/daymade/claude-code-skills.git --path llm-wiki-setup--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add daymade/claude-code-skills --skill llm-wiki-setup -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install daymade/claude-code-skills llm-wiki-setup --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/daymade/claude-code-skills.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/llm-wiki-setup .gemini/skills/llm-wiki-setup && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "llm-wiki-setup" agent skill from https://github.com/daymade/claude-code-skills/tree/main/llm-wiki-setup into .gemini/skills/llm-wiki-setup/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki-setup", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install daymade/claude-code-skills llm-wiki-setupInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add daymade/claude-code-skills --skill llm-wiki-setup -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/daymade/claude-code-skills.git skills-src && mkdir -p .github/skills && cp -r skills-src/llm-wiki-setup .github/skills/llm-wiki-setup && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "llm-wiki-setup" agent skill from https://github.com/daymade/claude-code-skills/tree/main/llm-wiki-setup into .github/skills/llm-wiki-setup/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki-setup", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add daymade/claude-code-skills --skill llm-wiki-setup -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install daymade/claude-code-skills llm-wiki-setup --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/daymade/claude-code-skills.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/llm-wiki-setup .opencode/skills/llm-wiki-setup && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "llm-wiki-setup" agent skill from https://github.com/daymade/claude-code-skills/tree/main/llm-wiki-setup into .opencode/skills/llm-wiki-setup/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki-setup", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
llm-wiki-setup- GitHub stars
- 1.4k
- Token cost
- ~885 tokens
- SKILL.md length
- 239 words
- Files
- 25 (incl. scripts, references)
- Skills in repo
- 103
- Repo updated
- First seen
- Licence
- MIT
At a glance
Co-creates a personal investment-research LLM Wiki (Karpathy's pattern) by interviewing the user for THEIR OWN analysis framework, not a template — a living CLAUDE.md of pure markdown + wikilinks…
- Works in 5 steps: 判断意图 → scaffold 机制层 → 访谈共创 CLAUDE.md ★核心步骤 → …
- Build a compounding 投研第二大脑 / 投研知识库 / 个人投研 wiki
- SKILL.md covers ★ 先读这一条（这个 skill 的灵魂）, 不碰的红线（Karpathy 原意，别…, 机制层 vs 规则层（贯穿全程的区分） and 工作流, plus 3 more sections
- Runs Python scripts from its folder; calls python, git and uv
What it does
LLM Wiki Setup is an agent skill from daymade/claude-code-skills. Co-creates a personal investment-research LLM Wiki (Karpathy's pattern) by interviewing the user for THEIR OWN analysis framework, not a template — a living CLAUDE.md of pure markdown + wikilinks, no RAG in place of compilation. Use to build a compounding 投研第二大脑 / 投研知识库 / 个人投研 wiki, ingest research reports / earnings calls into an existing wiki, or run post-earnings prediction→fulfillment reviews.
Its SKILL.md is about 890 tokens, which your agent loads only when the skill is triggered. The skill folder holds 32 other files, including scripts and reference files (for example `examples/investment-research-CLAUDE.md`, `references/counter_review.md` and `references/fulfillment_sop.md`).
It sits in Knowledge Management, covering LLM wikis. The repository describes itself as: Professional Claude Code skills marketplace featuring production-ready skills for enhanced development workflows. The licence is MIT.
When your agent uses it
- Build a compounding 投研第二大脑 / 投研知识库 / 个人投研 wiki
- Ingest research reports / earnings calls into an existing wiki
- Run post-earnings prediction→fulfillment reviews
Example prompts
- “/llm-wiki-setup”
Requirements
- Python 3
Workflow steps
5 steps, taken from the step headings in SKILL.md.
What it can do on your machine
Read from SKILL.md and the folder at commit 872127b. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Ships 2 files in scripts/ (Python, from the files we listed), which the agent can run.
Shell commands in SKILL.md call:
pythongituv
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md. Its commands use git and uv, which can reach the network depending on how they are called.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
LLM Wiki Setup loads about 885 tokens when it runs, and up to ~4.3k if it reads all its reference files. Until then it costs about 104 tokens; SKILL.md has 239 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.
SKILL.md
The full file from daymade/claude-code-skills at commit 872127b, republished under its MIT licence (© daymade). 239 words, ~885 tokens.
Download SKILL.mdSave it as
.claude/skills/llm-wiki-setup/SKILL.md (or your agent's skills folder). This skill also uses 24 other files; get the full folder from GitHub.- name
- llm-wiki-setup
- description
- Co-creates a personal investment-research LLM Wiki (Karpathy's pattern) by interviewing the user for THEIR OWN analysis framework, not a template — a living CLAUDE.md of pure markdown + wikilinks, no RAG in place of compilation. Use to build a compounding 投研第二大脑 / 投研知识库 / 个人投研 wiki, ingest research reports / earnings calls into an existing wiki, or run post-earnings prediction→fulfillment reviews.
LLM Wiki Setup（投研第二大脑共创）
帮用户搭一个金融投研专用 LLM Wiki（Karpathy 模式）：纯 markdown 文件 + [[wikilink]] 互联 + LLM 维护，知识随用复利。
但核心不是给一份投研模板——是引导用户把他自己的投资判断方式，提炼成他专属的 CLAUDE.md。
★ 先读这一条（这个 skill 的灵魂）
每个人用自己的语言、自己的投资偏好，建自己的 CLAUDE.md。
两个投资者看同一家公司，关注点可能完全不同——一个看「下季度订单能否超市场预期」，另一个看「管理层电话会上的语气和信心」。给他们同一份模板，就抹掉了让 wiki 有用的那个东西。
- ✅ 你的工作 = 访谈用户 → 提炼他的关注维度 → 用他的话写进 CLAUDE.md
- ❌ 你的失败 = 套一份「标准投研 schema」让他填空，或让他照抄
examples/
examples/investment-research-CLAUDE.md 是一个人长成的样子，给用户看可能性，禁止照抄。它像模板一样被搬走，这个 skill 就失败了。
不碰的红线（Karpathy 原意，别 over-engineer）
纯 markdown + wikilink。不用检索替代编译：知识靠预编译进结构化页「复利」，不是每次 query 重新检索原始文档——这是本模式相对 RAG 的根本区别，也是 Karpathy 的核心 idea。搜索层不在禁区：他的 gist 写明百来个来源靠 index.md 就够，wiki 长大后要正经搜索，并点名 qmd（本地 BM25 + 向量 + 重排）——搜的是编译好的 wiki 页，不是拿原始文档的片段替代编译。别加 knowledge graph / 自动 health-check 之类机制（社区有些版本加了，那是 over-engineer）。
机制层 vs 规则层（贯穿全程的区分）
| 内容 | 处置 | |
|---|---|---|
| 机制层 | 三层目录 + wikilink + lint + git hook | ✅ 通用工程结构，scripts/init_vault.py 直接装 |
| 规则层 | 看哪些维度 / 怎么记观点 / 要不要分析师归属 / 怎么复盘 / 要长报告还是三行 | ❌ 用户的投资大脑，访谈长出来，绝不给模板 |
机制层照抄没问题（它是 Karpathy 模式的工程卫生，跟「你怎么投资」无关）。规则层照抄 = 背叛方法论。
工作流
Phase 0 — 判断意图
- 新建 vault → Phase 1
- 已有 vault，ingest 一份源 → 直接读
references/ingest_sop.md - 已有 vault，财报后复盘某标的 →
references/fulfillment_sop.md - query → 读 vault 的
index.md+ 相关页，带 citation 综合答；好答案回填 synthesis
Phase 1 — scaffold 机制层
python scripts/init_vault.py <目标目录>建空骨架（三层目录 + lint + hook 占位 + 空 index/log + CLAUDE 骨架）。这一步只装机制层，不写任何 schema。
Phase 2 — 访谈共创 CLAUDE.md ★核心步骤
读 references/interview.md，按它的 8 个维度一条条访谈用户，把回答用他自己的话写进 <vault>/CLAUDE.md 规则层的占位。
- 一次问一个维度，别一口气灌
- 用户不在乎的维度直接砍（极简 > 全面）
- 卡住才翻
examples/给灵感，明说「别抄，挑你戳中的」 - 自检：写好的 CLAUDE.md 像不像「这个人」？像通用模板就重来
Phase 3 — 启用防腐
cd <vault> && git init
git config core.hooksPath .githooks # local 配置，换机/重 clone 要重设
PYTHONUTF8=1 uv run --no-project --with pyyaml python3 scripts/lint-vault.py wiki # 确认绿灯已有 vault — 刷新机制层工具
更新本 skill 后，显式刷新已复制进 vault 的 linter 与 hook：
python scripts/init_vault.py --refresh-tools <vault>只更新 scripts/lint-vault.py 与 .githooks/pre-commit，不碰 wiki/、raw/ 或用户的 CLAUDE.md。文件有变化时先保留 .before-refresh 备份；若备份已存在则 fail-fast，先审阅并移走旧备份再重跑。
Phase 4 — 首次 ingest 演示
拿用户一份真实的源（研报 / 电话会 / 纪要），按 references/ingest_sop.md 走一遍 HITL 5 卡点，让他亲眼看到 wiki 怎么从源长出来。用用户自己的素材，不要用 examples。
后续运营（按需读 references）
| 场景 | 读 |
|---|---|
| ingest 新源 | references/ingest_sop.md（doc_type 用用户自己定的分类） |
| 财报后复盘 | references/fulfillment_sop.md（分析师回测调 analyst-track-record skill，别重造） |
| vault 卫生（派生值漂移） | references/prune_discipline.md |
| 复盘页对抗审查 | references/counter_review.md |
| 怎么访谈提炼用户的投资大脑 | references/interview.md（Phase 2 的完整方法） |
为什么这个 skill 是 inline（不设 context: fork）
它要调 analyst-track-record skill（复盘回测）、跑 Bash（scaffold / lint）、可能并行 Task 取财报数据——subagent 不能调 skill 或 spawn subagent，所以必须 inline。
Next Step
vault 搭好、用户开始 ingest 卖方研报后，如果他想回测某分析师过去准不准 → 建议接 analyst-track-record skill（双维度命中率，有 validated 脚本）。
© daymade, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
SKILL.md and 24 other files (scripts, references) in llm-wiki-setup of daymade/claude-code-skills.
- SKILL.md
- examples/investment-research-CLAUDE.md
- references/counter_review.md
- references/fulfillment_sop.md
- references/ingest_sop.md
- references/interview.md
- references/prune_discipline.md
- scripts/init_vault.py
- scripts/lint-vault.py
- templates/CLAUDE-skeleton.md
- templates/pre-commit.snippet
- templates/vault/raw/.gitkeep
- templates/vault/wiki/analysts/.gitkeep
- … and 12 more
Open the folder on GitHubat commit 872127b
Compare with similar skills
LLM Wiki Setup next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| LLM Wiki Setup this skilldaymade/claude-code-skills | 1.4k | — | ~885 | Automated safety check: Pass | MIT | |
| Karpathy LLM WikiAstro-Han/karpathy-llm-wiki | 2.4k | — | ~3.6k | Automated safety check: Pass | MIT | |
| LLM Wikilewislulu/llm-wiki-skill | 655 | — | ~3.7k | Automated safety check: Pass | None | |
| Wiki Builderrohitg00/pro-workflow | 2.9k | — | ~1k | Automated safety check: Pass | None | |
| Codex History IngestAr9av/obsidian-wiki | 3.5k | — | ~2.2k | Automated safety check: Notes | MIT | |
| Arkon Editnduckmink/arkon | 1.5k | — | ~1.6k | Automated safety check: Pass | Custom licence |
Similar skills
LLM Wiki
lewislulu/llm-wiki-skill
Build and maintain a Karpathy-style LLM knowledge base — a self-compiling Obsidian markdown wiki where an Agent ingests raw sources, compiles cross-linked concept/entity/summary pages, answers…
Wiki Builder
rohitg00/pro-workflow
Start, structure, and grow a persistent research wiki indexed in pro-workflow's SQLite knowledge base.
Codex History Ingest
Ar9av/obsidian-wiki
Ingest Codex CLI conversation/session history into Obsidian as distilled knowledge.
Arkon Edit
nduckmink/arkon
Propose or directly apply edits to Arkon wiki pages, including proposing brand new pages.
Research Wiki Builder
dair-ai/dair-academy-plugins
Creates and maintains configurable research wikis: scaffold a folder, add sources, compile pages and indexes, and file query answers back.
More from daymade/claude-code-skills
All 103 skills in this repoVideo Comparer
daymade/claude-code-skills
This skill should be used when comparing two videos to analyze compression results or quality differences.
CLI Demo Generator
daymade/claude-code-skills
Generates professional animated CLI demos as GIFs using VHS terminal recordings.
Doc To Markdown
daymade/claude-code-skills
Converts DOCX/PDF/PPTX and saved HTML/HTM to high-quality Markdown with automatic post-processing.
Interaction Design Board
daymade/claude-code-skills
Generates several distinct, clickable HTML interaction prototypes for one product surface into a Design Board and collects selection/remix feedback before implementation.
Auto Repo Setup
daymade/claude-code-skills
Diagnoses and repairs repository setup and guarded Git workflows for Claude Code or Codex — environment repair, startup sync, hook auditing, collaborator handoff.
Bigdata Skill
daymade/claude-code-skills
Pulls Bigdata.com (RavenPack) financial and news data via the official bigdata-client SDK and /v1/ REST endpoints — structured financials, prices, analyst estimates, entity-sentiment series…
Related
Categories
Questions about LLM Wiki Setup
What does LLM Wiki Setup do?
Co-creates a personal investment-research LLM Wiki (Karpathy's pattern) by interviewing the user for THEIR OWN analysis framework, not a template — a living CLAUDE.md of pure markdown + wikilinks…. LLM Wiki Setup is an agent skill from daymade/claude-code-skills.md of pure markdown + wikilinks, no RAG in place of compilation.
When should I use LLM Wiki Setup?
LLM Wiki Setup fits situations like: build a compounding 投研第二大脑 / 投研知识库 / 个人投研 wiki; ingest research reports / earnings calls into an existing wiki; run post-earnings prediction→fulfillment reviews.
How do I install LLM Wiki Setup in Claude Code?
Run `npx skills add daymade/claude-code-skills --skill llm-wiki-setup -a claude-code`. Or copy the skill folder (llm-wiki-setup in daymade/claude-code-skills) into .claude/skills/llm-wiki-setup in your project. Claude Code loads it when a task matches its description.
How do I install LLM Wiki Setup in Codex?
Run `npx skills add daymade/claude-code-skills --skill llm-wiki-setup -a codex`. Or copy the skill folder (llm-wiki-setup in daymade/claude-code-skills) into .agents/skills/llm-wiki-setup in your project. Codex loads it when a task matches its description.
Can I use LLM Wiki Setup in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add daymade/claude-code-skills --skill llm-wiki-setup -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/llm-wiki-setup, .gemini/skills/llm-wiki-setup, .github/skills/llm-wiki-setup and .opencode/skills/llm-wiki-setup in your project.
What does LLM Wiki Setup need to run?
Going by SKILL.md and its folder, LLM Wiki Setup needs Python for the scripts in its folder and the command-line tools its instructions call (python, git and uv). Our summary lists: Python 3.
Does LLM Wiki Setup access the network?
SKILL.md contains no URLs. Its commands use git and uv, which can reach the network depending on how they are called. This is read from the text; nothing was executed.
Is LLM Wiki Setup safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.
What licence does LLM Wiki Setup use?
LLM Wiki Setup is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does LLM Wiki Setup use?
About 885 tokens (SKILL.md is roughly 3.5k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 3.5k tokens, read only when the agent opens those files.
What are the alternatives to LLM Wiki Setup?
Skills that share tags, products or a category with LLM Wiki Setup: Karpathy LLM Wiki (Astro-Han/karpathy-llm-wiki, 2.4k stars), LLM Wiki (lewislulu/llm-wiki-skill, 655 stars), Wiki Builder (rohitg00/pro-workflow, 2.9k stars) and Codex History Ingest (Ar9av/obsidian-wiki, 3.5k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains LLM Wiki Setup?
daymade (a GitHub user) maintains it in daymade/claude-code-skills, which has 1,447 GitHub stars. The repository holds 103 skills in this directory. The repository was last updated on October 9, 2026.
Source: daymade/claude-code-skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.