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…
Xây dựng và duy trì knowledge base cá nhân theo pattern LLM Wiki (Karpathy).
$ npx skills add mduongvandinh/llm-wiki --skill llm-wiki -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install mduongvandinh/llm-wiki llm-wiki --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/mduongvandinh/llm-wiki.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/llm-wiki .claude/skills/llm-wiki && 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" agent skill from https://github.com/mduongvandinh/llm-wiki/tree/main/skills/llm-wiki into .claude/skills/llm-wiki/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki", 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/mduongvandinh/llm-wiki/tree/main/skills/llm-wikiType 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 mduongvandinh/llm-wiki --skill llm-wiki -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install mduongvandinh/llm-wiki llm-wiki --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mduongvandinh/llm-wiki.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/llm-wiki .agents/skills/llm-wiki && 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" agent skill from https://github.com/mduongvandinh/llm-wiki/tree/main/skills/llm-wiki into .agents/skills/llm-wiki/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki", 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 mduongvandinh/llm-wiki --skill llm-wiki -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install mduongvandinh/llm-wiki llm-wiki --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mduongvandinh/llm-wiki.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/llm-wiki .cursor/skills/llm-wiki && 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" agent skill from https://github.com/mduongvandinh/llm-wiki/tree/main/skills/llm-wiki into .cursor/skills/llm-wiki/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki", 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/mduongvandinh/llm-wiki.git --path skills/llm-wiki--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 mduongvandinh/llm-wiki --skill llm-wiki -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install mduongvandinh/llm-wiki llm-wiki --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mduongvandinh/llm-wiki.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/llm-wiki .gemini/skills/llm-wiki && 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" agent skill from https://github.com/mduongvandinh/llm-wiki/tree/main/skills/llm-wiki into .gemini/skills/llm-wiki/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki", 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 mduongvandinh/llm-wiki llm-wikiInstalls 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 mduongvandinh/llm-wiki --skill llm-wiki -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/mduongvandinh/llm-wiki.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/llm-wiki .github/skills/llm-wiki && 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" agent skill from https://github.com/mduongvandinh/llm-wiki/tree/main/skills/llm-wiki into .github/skills/llm-wiki/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki", 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 mduongvandinh/llm-wiki --skill llm-wiki -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install mduongvandinh/llm-wiki llm-wiki --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mduongvandinh/llm-wiki.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/llm-wiki .opencode/skills/llm-wiki && 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" agent skill from https://github.com/mduongvandinh/llm-wiki/tree/main/skills/llm-wiki into .opencode/skills/llm-wiki/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-wiki", 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.
llm-wikiXây dựng và duy trì knowledge base cá nhân theo pattern LLM Wiki (Karpathy).
LLM Wiki is an agent skill from mduongvandinh/llm-wiki. Xây dựng và duy trì knowledge base cá nhân theo pattern LLM Wiki (Karpathy). Hỗ trợ init, ingest, query, lint, discover, run, digest, pain-rank, setup, book-summary, competitive-brief, interview-prep.
Its SKILL.md is about 4.8k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.
It sits in Knowledge Management, covering LLM wikis and Interview preparation. It works with Obsidian. The repository describes itself as: Hệ thống knowledge base cá nhân hoàn toàn tự động, vận hành bởi LLM. Dựa trên pattern LLM Wiki của Andrej Karpathy. The licence is MIT.
4 steps, taken from the first numbered list in SKILL.md.
Read from SKILL.md and the folder at commit 4d7227e. It shows what the files ask for, not the result of running them.
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.
No scripts in the folder and no shell commands in SKILL.md (its code samples are markdown).
From the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
LLM Wiki loads about 4.8k tokens when it runs. Until then it costs about 52 tokens; SKILL.md has 1,935 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.
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); files beside SKILL.md are not scanned.
The full file from mduongvandinh/llm-wiki at commit 4d7227e, republished under its MIT licence (© mduongvandinh). 1,935 words, ~4,807 tokens.
.claude/skills/llm-wiki/SKILL.md (or your agent's skills folder).Hệ thống knowledge base cá nhân. LLM xây dựng và duy trì wiki từ nguồn thô. Dựa trên pattern của Andrej Karpathy, mở rộng với auto-discovery.
Skill này hỗ trợ các sub-commands sau. Parse argument đầu tiên để xác định command:
| Command | Mô tả | Ví dụ |
|---|---|---|
init | Tạo wiki mới cho một chủ đề | /llm-wiki init "AI Agents" |
ingest | Xử lý nguồn mới trong raw/ | /llm-wiki ingest |
query | Hỏi đáp dựa trên wiki | /llm-wiki query "So sánh RAG vs Wiki" |
lint | Kiểm tra sức khỏe wiki | /llm-wiki lint |
discover | Tự tìm nguồn mới | /llm-wiki discover |
run | Chạy full cycle: discover → ingest → lint | /llm-wiki run |
status | Xem trạng thái wiki | /llm-wiki status |
digest | Daily brief — tóm tắt thay đổi wiki 24h | /llm-wiki digest |
pain-rank | Xếp hạng pain points theo cơ hội kinh doanh | /llm-wiki pain-rank |
setup | Khởi tạo wiki theo variant template có sẵn | /llm-wiki setup book-companion |
book-summary | Tóm tắt có cấu trúc toàn bộ wiki sách | /llm-wiki book-summary |
competitive-brief | Battlecard 1-pager cho một competitor | /llm-wiki competitive-brief "Linear" |
interview-prep | 1-pager chuẩn bị interview cho một công ty | /llm-wiki interview-prep "Stripe" |
Nếu không có sub-command → hiển thị trạng thái và hỏi user muốn làm gì.
WIKI_ROOT = <current working directory>Xác định WIKI_ROOT bằng cách tìm folder chứa CLAUDE.md + config.yaml + wiki/ + raw/. Thường là thư mục project hiện tại.
Luôn đọc WIKI_ROOT/CLAUDE.md trước khi thực hiện bất kỳ command nào — đó là schema quy định mọi quy tắc.
Mục đích: Khởi tạo wiki mới hoặc thêm topic mới vào wiki hiện tại.
Quy trình:
CLAUDE.md và config.yamlconfig.yaml → topicsVí dụ:
/llm-wiki init "Rust Programming"
→ Thêm topic "Rust Programming" vào config.yaml
→ Keywords tự sinh: ["Rust language", "Rust programming", "cargo", "rustc"]Mục đích: Xử lý mọi file mới/chưa xử lý trong raw/.
Quy trình:
CLAUDE.md (schema & quy tắc).discoveries/history.json → lấy danh sách đã xử lýraw/ → tìm file chưa có trong historywiki/sources/
c. Trích xuất entities → tạo/cập nhật wiki/entities/
d. Trích xuất concepts → tạo/cập nhật wiki/concepts/
e. Thêm cross-references [[links]] vào các trang liên quan
f. Phát hiện contradictions → ghi chú vào trang liên quanwiki/INDEX.mdwiki/LOG.md.discoveries/history.jsonQuy tắc QUAN TRỌNG:
raw/[Nguồn: filename](../raw/path)schedule.ingest.batch_sizeMục đích: Hỏi đáp dựa trên nội dung wiki.
Quy trình:
wiki/INDEX.md → tìm trang liên quan đến câu hỏi[[trang-wiki]]wiki/syntheses/ hoặc outputs/Quy tắc:
Ví dụ:
/llm-wiki query "So sánh RAG truyền thống vs LLM Wiki pattern"
→ Đọc INDEX.md → tìm trang về RAG, LLM Wiki
→ Đọc các trang liên quan
→ Tạo bảng so sánh
→ Lưu vào wiki/syntheses/rag-vs-llm-wiki.mdMục đích: Kiểm tra sức khỏe wiki, phát hiện vấn đề.
Quy trình:
wiki/INDEX.mdwiki/[[link]] trỏ đến trang chưa tồn tạioutputs/lint-YYYY-MM-DD.md.discoveries/gaps.json (cho discover dùng)config.yaml → schedule.lint.auto_fix = true → tự sửa lỗi đơn giảnOutput format:
# Lint Report — YYYY-MM-DD
## Tóm tắt
- Tổng trang: N
- Contradictions: N
- Orphans: N
- Missing pages: N
- Gaps: N
## Chi tiết
### Contradictions
...
### Đề xuất
- Tạo trang mới: [danh sách]
- Tìm nguồn cho: [danh sách gaps]Mục đích: Tự động tìm nguồn mới từ internet.
Quy trình:
config.yaml → topics, feeds, discovery settings.discoveries/gaps.json → knowledge gaps cần lấp.discoveries/history.json → tránh trùng lặpsite:reddit.com theo subreddits + keywords trong config.yaml → tìm pain points, use cases, ý tưởng. Lưu vào raw/reddit/YYYY-MM-DD-slug.md. Trích xuất: vấn đề gốc, giải pháp được đề xuất, upvotes, sentiment.
b. github_trending: WebSearch GitHub trending repos theo languages/topics filter
c. github_watch: Kiểm tra repos/orgs/people trong config → new releases, new repos
d. web_search: WebSearch theo keywords của mỗi topic
e. feed_poll: Kiểm tra RSS feeds, Hacker News
f. gap_fill: WebSearch theo gaps từ lint
g. snowball: Đọc references trong wiki → follow links chưa córaw/articles/YYYY-MM-DD-slug.md với frontmatter:---
title: "Tiêu đề"
url: "https://..."
discovered: YYYY-MM-DD
topic: "tên topic"
---.discoveries/history.jsonQuy tắc:
config.yaml → schedule.discover.max_sourcesraw/reddit/ (tách riêng khỏi raw/articles/)Mục đích: Chạy full cycle tự động.
Quy trình:
discover → ingest → lint → (nếu có gaps mới → discover lại)discover → tìm nguồn mới (Reddit, GitHub trending, web search)ingest → xử lý mọi file mới trong raw/lint → kiểm tra sức khỏeoutputs/run-YYYY-MM-DD.mdGiới hạn: Tối đa 2 vòng discover-ingest để tránh vòng lặp vô hạn.
Mục đích: Hiển thị trạng thái wiki hiện tại.
Output:
LLM Wiki Status
═══════════════
Wiki: My LLM Wiki
Topics: 3 (LLM Agents, Claude Code, AI Engineering)
Raw sources: N files
Wiki pages: N pages (E entities, C concepts, S sources, Y syntheses)
Last ingest: YYYY-MM-DD
Last lint: YYYY-MM-DD
Last discover: YYYY-MM-DD
Knowledge gaps: N
Orphan pages: N
Health: Good | Warning | Needs AttentionMục đích: Tạo daily brief — tóm tắt mọi thay đổi wiki trong 24h, highlights insights mới.
Quy trình:
wiki/LOG.md → lọc entries trong 24h quaOutput format:
# Daily Digest — YYYY-MM-DD
## Nguồn mới (N)
- [tên nguồn] — 1 dòng tóm tắt
## Wiki pages mới (N)
- [tên trang] — 1 dòng mô tả
## Top 3 Insights
1. [Insight quan trọng nhất — trích từ syntheses hoặc cross-references mới]
2. [Insight thứ hai]
3. [Insight thứ ba]
## Pain Points mới phát hiện (từ Reddit)
| Pain Point | Domain | Upvotes | Cơ hội |
|------------|--------|---------|--------|
| ... | ... | ... | ... |
## Knowledge Gaps cần lấp
- [gap 1]
- [gap 2]
## Thống kê
- Wiki: N pages (+X hôm nay)
- Sources: N (+Y hôm nay)
- Health: Good/WarningLưu vào: outputs/digest-YYYY-MM-DD.md
Mục đích: Xếp hạng pain points từ Reddit và các nguồn khác theo tiềm năng kinh doanh.
Quy trình:
raw/reddit/wiki/concepts/ai-pain-points.md và wiki/concepts/micro-saas-pattern.mdScoring Framework (mỗi tiêu chí 1-10, tổng max 50):
| Tiêu chí | Mô tả | Trọng số |
|---|---|---|
| Urgency | Người dùng cần giải pháp ngay? Hay "nice to have"? | x2 |
| Market Size | Bao nhiêu người/doanh nghiệp có vấn đề này? | x2 |
| Willingness to Pay | Sẵn sàng trả tiền? Đang trả cho alternatives? | x3 |
| AI Solvability | AI/LLM có thể giải quyết tốt không? | x2 |
| Competition | Ít cạnh tranh = điểm cao | x1 |
Output format:
# Pain Point Ranking — YYYY-MM-DD
## Top 10 Cơ hội
| Rank | Pain Point | Domain | Score | Urgency | Market | WTP | AI-Fit | Comp |
|------|-----------|--------|-------|---------|--------|-----|--------|------|
| 1 | ... | B2B | 42/50 | 9 | 8 | 9 | 8 | 8 |
| 2 | ... | Consumer | 38/50 | 7 | 9 | 7 | 9 | 6 |
## Chi tiết Top 3
### #1: [Tên Pain Point] — Score: 42/50
- **Vấn đề:** [Mô tả cụ thể]
- **Target user:** [Ai có vấn đề này]
- **Giải pháp đề xuất:** [MVP concept]
- **Revenue model:** [Cách kiếm tiền]
- **Nguồn Reddit:** [Links/upvotes]
- **Next step:** [Hành động cụ thể tiếp theo]
### #2: ...
### #3: ...
## Idea-to-Spec Pipeline (cho #1)
- Problem Statement: ...
- Target User Persona: ...
- MVP Features (3-5): ...
- Tech Stack Suggestion: ...
- Estimated effort: ... (human) / ... (CC)Lưu vào: outputs/pain-rank-YYYY-MM-DD.md
Quy tắc:
[[kebab-case-name]][Nguồn: filename](../raw/path)Mục đích: Khởi tạo wiki theo một variant template có sẵn trong variants/.
Syntax: /llm-wiki setup <variant-name>
Variants có sẵn: book-companion, competitive-intel, job-search
Quy trình:
WIKI_ROOT/variants/<variant-name>/ tồn tạiconfig.yaml chưa có → copy variants/<variant-name>/config.yaml vào WIKI_ROOT/variants/<variant-name>/CLAUDE.md → merge schema extensions vào WIKI_ROOT/CLAUDE.mdraw/, wiki/entities/, wiki/concepts/, wiki/sources/, wiki/syntheses/, outputs/, .discoveries/variants/<variant-name>/sample-data/* → WIKI_ROOT/raw/articles/ingest tự động trên sample data vừa copyOutput ví dụ (book-companion):
Setting up book-companion variant...
✓ Config copied → config.yaml
✓ Schema extensions merged → CLAUDE.md
✓ Folder structure created
✓ Sample data: 3 files → raw/articles/
✓ Ingest complete: 12 wiki pages created
- 5 characters (Paul, Jessica, Duncan, Gurney, Stilgar)
- 3 locations (Caladan, Arrakis, Giedi Prime)
- 2 factions (Atreides, Harkonnen)
- 2 concepts (The Spice, Bene Gesserit)
Wiki ready! Try:
/llm-wiki query "Who is Paul Atreides?"
/llm-wiki query "What is the relationship between House Atreides and Arrakis?"
/llm-wiki book-summaryQuy tắc:
config.yaml đã tồn tại → hỏi user có muốn overwrite không, mặc định là KHÔNGMục đích: Tạo tóm tắt có cấu trúc toàn bộ wiki sách — characters, timeline, themes, mysteries.
Yêu cầu: Wiki phải ở book_mode: true trong config.yaml
Quy trình:
config.yaml → lấy tên sách, current_chapterwiki/INDEX.md → xác định tất cả pageswiki/entities/ (characters, locations, factions)wiki/concepts/ (themes, events, quotes)Output format:
# [Tên sách] — Book Summary (đến Chapter N)
## Cast of Characters
| Nhân vật | Vai trò | Faction | Trạng thái |
|----------|---------|---------|-----------|
| Paul Atreides | Protagonist | House Atreides | Alive |
| ... | | | |
## Timeline of Key Events
1. [Event 1] — Chapter N
2. [Event 2] — Chapter N
...
## Major Factions
- **House Atreides:** [mô tả ngắn, members, goals]
- **House Harkonnen:** [mô tả ngắn, members, goals]
## Key Locations
- **Arrakis:** [mô tả, significance]
- **Caladan:** [mô tả, significance]
## Major Themes
1. [Theme 1] — [giải thích ngắn]
2. [Theme 2] — [giải thích ngắn]
## Unresolved Mysteries / Open Questions
- [Question 1 chưa có câu trả lời đến chapter hiện tại]
- [Question 2]
## Notable Quotes
> "[Quote 1]" — [Nhân vật], Chapter N
> "[Quote 2]" — [Nhân vật], Chapter NLưu vào: outputs/book-summary-YYYY-MM-DD.md
Quy tắc:
current_chapter trong config (spoiler protection)Mục đích: Tạo battlecard 1-pager cho một competitor cụ thể.
Syntax: /llm-wiki competitive-brief "<Tên Competitor>"
Quy trình:
wiki/entities/wiki/changes/ nếu có (recent changes detected)Output format:
# Competitive Brief: [Competitor Name]
*Generated: YYYY-MM-DD | Sources: N wiki pages*
## One-Line Positioning
[Tagline hoặc how they describe themselves]
## Pricing
| Tier | Price | Key Limits |
|------|-------|-----------|
| Free | $0 | [limits] |
| Pro | $X/mo | [limits] |
| Enterprise | Custom | [limits] |
## Top Features
- [Feature 1]: [mô tả ngắn]
- [Feature 2]: [mô tả ngắn]
- [Feature 3]: [mô tả ngắn]
## Recent Moves (30 ngày qua)
- [Date]: [Change/announcement]
## Known Weaknesses (từ reviews & Reddit)
- [Weakness 1]
- [Weakness 2]
## Job Postings Signal
[Họ đang hire vị trí gì → signal về roadmap]
## How We Differ
[Điểm khác biệt của chúng ta so với họ]Lưu vào: outputs/battlecard-<competitor>-YYYY-MM-DD.md
Quy tắc:
Drop competitor info into raw/ and run /llm-wiki ingest⚡ CHANGEDMục đích: Tổng hợp 1-pager chuẩn bị interview cho một công ty cụ thể.
Syntax: /llm-wiki interview-prep "<Tên Công ty>"
Quy trình:
wiki/entities/wiki/sources/ liên quan đến công ty đóOutput format:
# Interview Prep: [Company Name]
*Generated: YYYY-MM-DD | Sources: N wiki pages*
## Company Overview
[3-4 câu: business model, stage, size]
## Tech Stack
[Những gì đã biết về tech stack của họ]
## Engineering Culture
- [Culture signal 1 — từ reviews/blogs]
- [Culture signal 2]
- [Culture signal 3]
## Recent News (30 ngày qua)
- [Date]: [News item]
## The Role
[Tóm tắt JD đã collect được nếu có]
## Interview Process (từ Glassdoor/Reddit)
1. [Round 1]: [format, duration, focus]
2. [Round 2]: [format]
...
## Known Interview Questions
- [Question 1 — thường gặp]
- [Question 2]
## Green Flags
- [Positive signal 1]
- [Positive signal 2]
## Red Flags / Watch Out For
- [Warning 1 — từ reviews]
## Questions to Ask Them
- [Question 1 — meaningful, shows research]
- [Question 2]
- [Question 3]Lưu vào: outputs/interview-prep-<company>-YYYY-MM-DD.md
Quy tắc:
Khi book_mode: true trong config.yaml, áp dụng thêm logic sau trong command ingest:
Quy trình bổ sung sau bước đọc file:
chapter: Nconfig.yaml → current_chapterchapter > current_chapter → SKIP file này[SKIP] raw/articles/[filename] — chapter N > current_chapter M (spoiler protection)Nếu không có field chapter trong frontmatter → ingest bình thường (coi là safe)
Cập nhật current_chapter: User tự sửa config.yaml → current_chapter khi đọc đến chương mới.
Gợi ý thêm vào setup: "Update current_chapter in config.yaml as you read each chapter."
Khi change_detection: true trong config.yaml, áp dụng thêm logic sau trong command discover:
Quy trình bổ sung khi fetch source đã có trong history:
wiki/changes/YYYY-MM-DD-<entity-name>.md với diff
b. Cập nhật entity page, ghi ⚡ CHANGED: [mô tả thay đổi] ở đầu
c. Ghi LOG.md với tag [CHANGE DETECTED]
d. Include trong digest reportFormat file change:
---
type: change
entity: competitor-name
detected: YYYY-MM-DD
---
# Change Detected: [Competitor Name] — YYYY-MM-DD
## What Changed
[Mô tả thay đổi cụ thể]
## Old Value
[Giá trị cũ]
## New Value
[Giá trị mới]
## Significance
[Tại sao thay đổi này quan trọng]© mduongvandinh, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Just SKILL.md in skills/llm-wiki of mduongvandinh/llm-wiki.
Open the folder on GitHubat commit 4d7227e
LLM Wiki 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 this skillmduongvandinh/llm-wiki | 214 | — | ~4.8k | Automated safety check: Pass | MIT | |
| LLM Wikilewislulu/llm-wiki-skill | 655 | — | ~3.7k | Automated safety check: Pass | None | |
| Codex History IngestAr9av/obsidian-wiki | 3.5k | — | ~2.2k | Automated safety check: Notes | MIT | |
| LLM Wikizosmaai/pi-llm-wiki | 608 | — | ~4.4k | Automated safety check: Pass | MIT | |
| LLM Wikipraneybehl/llm-wiki-plugin | 118 | — | ~5.7k | Automated safety check: Pass | MIT | |
| Hermes History IngestAr9av/obsidian-wiki | 3.5k | — | ~2.2k | Automated safety check: Notes | MIT |
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…
Ar9av/obsidian-wiki
Ingest Codex CLI conversation/session history into Obsidian as distilled knowledge.
zosmaai/pi-llm-wiki
Build and maintain a persistent, interlinked Obsidian-compatible markdown wiki using Karpathy's LLM Wiki pattern.
praneybehl/llm-wiki-plugin
Build and maintain an LLM-curated knowledge base from papers, articles, transcripts, notes and project findings.
Ar9av/obsidian-wiki
Ingest Hermes agent history into Obsidian as distilled knowledge.
SherwinQ/karpathy-wiki
A skill your agent uses when building or maintaining a personal knowledge base with LLM assistance.
Works with
Categories
Xây dựng và duy trì knowledge base cá nhân theo pattern LLM Wiki (Karpathy). LLM Wiki is an agent skill from mduongvandinh/llm-wiki. Xây dựng và duy trì knowledge base cá nhân theo pattern LLM Wiki (Karpathy).
LLM Wiki fits situations like: tasks that involve LLM wikis; tasks that involve Interview preparation.
Run `npx skills add mduongvandinh/llm-wiki --skill llm-wiki -a claude-code`. Or copy the skill folder (skills/llm-wiki in mduongvandinh/llm-wiki) into .claude/skills/llm-wiki in your project. Claude Code loads it when a task matches its description.
Run `npx skills add mduongvandinh/llm-wiki --skill llm-wiki -a codex`. Or copy the skill folder (skills/llm-wiki in mduongvandinh/llm-wiki) into .agents/skills/llm-wiki in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add mduongvandinh/llm-wiki --skill llm-wiki -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, .gemini/skills/llm-wiki, .github/skills/llm-wiki and .opencode/skills/llm-wiki in your project.
SKILL.md names no scripts, command-line tools or credentials: LLM Wiki is instructions for the agent only.
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
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. Review the folder before installing.
LLM Wiki is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 4.8k tokens (SKILL.md is roughly 19k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
Skills that share tags, products or a category with LLM Wiki: LLM Wiki (lewislulu/llm-wiki-skill, 655 stars), Codex History Ingest (Ar9av/obsidian-wiki, 3.5k stars), LLM Wiki (zosmaai/pi-llm-wiki, 608 stars) and LLM Wiki (praneybehl/llm-wiki-plugin, 118 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
mduongvandinh (a GitHub user) maintains it in mduongvandinh/llm-wiki, which has 214 GitHub stars. The repository was last updated on April 16, 2026.
Source: mduongvandinh/llm-wiki on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.