wiki-ingest
Skill34 starsChanged 3 months agoMITView on GitHub
How to ingest a raw source into the LLM wiki (v2.1 lazy) — write a L2-episodic evidence page and NEW L3 concept drafts only; defer merge, confidence recompute, relations, and L2→L3 promotion to the batch consolidation at lint. Cheap & fast. Use when ingesting/processing/파일링 a source dropped into raw/.
47 other projects use a version of this.
Adopted by
Most recently taken up by
- microsoft/llmwikitheir own version, first seen Nov 15, 2026
- tboome33/obsidian-mcp-routertheir own version, first seen Nov 15, 2026
- methylo/claude-skill-booktheir own version, first seen Nov 8, 2026
- Ar9av/obsidian-wikitheir own version, first seen Nov 8, 2026
- barakgut/llm-wikitheir own version, first seen Nov 1, 2026
- chobizzy/llm-wikitheir own version, first seen Nov 1, 2026
Used with
Projects with it also use wiki-capture, wiki-status and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
jehrhardtSkillwiki-capture
Quickly capture raw thoughts, links, images, or notes to today's Obsidian daily note from anywhere in Pi. Use when the user wants to jot something down for later wiki processing.
Also in 6 projects that use this
nexuschainsSkillwiki-status
Reports the current state of the wiki vault — page counts, category breakdown, hub pages, orphans, stale entries, and a delta showing what has changed since the last ingest. Use this skill when the operator asks "what's the status", "wiki dashboard", "how many pages", "what's stale", "what's been ingested", "show me the delta", or "wiki health overview". Also surfaces structural insights (top hubs, fragmented clusters, bridge pages) when the operator asks "wiki insights", "what's central", or "show me the hubs". Read-only: this skill reports and recommends; it does not modify any files.
Also in 5 projects that use this
MicrosoftSkillwiki-research
Research a topic across the user's LLM Wiki — gather every relevant page, synthesize what the wiki knows, flag contradictions, and list knowledge gaps with suggested new pages. Use when the user wants an overview, literature-style synthesis or gap analysis of a topic in their wiki.
Also in 6 projects that use this
nexuschainsSkillcross-linker
Scans the wiki vault and adds missing [[wikilinks]] between pages that are related but not yet connected. Use this skill when the operator says "link my pages", "find missing links", "cross-reference the wiki", "connect my wiki", "add wikilinks", or after any large ingest to weave new pages into the existing knowledge graph. Also trigger when orphaned pages need connecting or when wiki-lint reports fragmented tag clusters. This is a write-heavy skill — it modifies page files to insert links, unlike wiki-lint which only reports.
Also in 4 projects that use this
tboome33Skillwiki-export
Export a vault's wiki either as a single portable file (llms.txt or llms-full.txt per the llmstxt.org standard) or as an OKF knowledge bundle (Google's Open Knowledge Format v0.1 — a shareable directory of markdown files any AI agent can consume). Use when the user says "export my wiki", "make an llms.txt", "share my wiki", "export as OKF", "make an OKF bundle", "publish my wiki as a bundle", "backup the wiki to a single file", "/wiki-export", or asks for a portable dump of their knowledge base.
Also in 4 projects that use this
Ivy-Chen1999Skillwiki-setup
Initialize a new Obsidian wiki vault with the correct structure, special files, and configuration. Use this skill when the user wants to set up a new wiki from scratch, initialize the vault structure, create the .env file, or says things like "set up my wiki", "initialize obsidian", "create a new vault", "get started with the wiki". Also use when the user needs to reconfigure their existing vault or fix a broken setup.
Also in 5 projects that use this
- +1 more from nexuschains
Other versions47 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 47
- AgriciDaniel/claude-obsidian★ 16ksame name
- sanyuan0704/code-review-expert★ 3.9ksame name
- sanyuan0704/sanyuan-skills★ 3.9ksame name
- Ar9av/obsidian-wiki★ 3.7ksame name
- kfchou/wiki-skills★ 189same name
- IssacW228/student-llm-wiki★ 173same name
- microsoft/llmwiki★ 116same name
- bozhouDev/bozhou-skills★ 78same name
- vanillaflava/llm-wiki-skills★ 65same name
- Ivy-Chen1999/llm-wiki-kit★ 21same name
37 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository8 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- meeting-minutesSkill
- weekly-reviewSkill
- wiki-consolidateSkill
- wiki-lintSkill
- wiki-mocSkill
- wiki-opsSkill
- wiki-querySkill
The file
Files
wiki-ingest (v2.1 — lazy·경량)
한 소스를 위키에 넣는다. 핵심: 인제스트는 싸고 빠르게, 정제는 나중에 배치로.
- 인제스트가 하는 것: L2 증거 페이지 + 신규 개념 초안만.
- 인제스트가 안 하는 것: 기존 페이지 병합·신뢰도 재계산·관계 구조화·L2→L3 승격 → 전부 lint/
wiki-consolidate배치로 지연(opus).
왜: v2에서 인제스트가 소스마다 기존 페이지를 전수 Read하고 병합·관계까지 하니 위키가 클수록 O(n)로 토큰이 폭증했다. 병합은 본질적으로 누적 후 일괄이 효율적이다(같은 주장 3회+ 승격도 원래 lint 시점). 인제스트는 증거만 남기고 빠진다.
규약 정본: CLAUDE.md.
주의: 페이지 파일에
</content>같은 닫는 태그·래퍼를 절대 붙이지 마라. 파일은 순수 마크다운(frontmatter + 본문)으로 끝난다. (과거 인제스트가 EOF에 stray</content>를 남겨 85개 파일을 sweep한 적 있음.)
절차
- 소스 읽기 + 중복 체크.
raw/대상. 본문 immutable. frontmatteringest_status: done이면 이미 인제스트 — 사용자에게 알리고 중단(재인제스트는 갱신 모드 별도). 소스 유형 판별(official/code/normal/verbal). - 기존 개념 파악 — 본문 Read 금지. 다음 한 줄 실행으로 기존 L3/L4 카탈로그만 본다(45개 페이지 본문을 읽지 마라):
출력(slug|title|aliases|confidence)으로 소스의 개념이 이미 있는지(기존) vs 없는지(신규) 판별한다.python3 .claude/skills/wiki-ingest/scripts/concept-index.py . - L2 증거 페이지 →
wiki/L2-episodic/source-<slug>.md. 이것이 인제스트의 주 산물.
본문: 2–3문장 요약 +--- type: source title: <한국어 제목> tags: [] created: <오늘> updated: <오늘> last_confirmed: <오늘> decay_class: episodic source_url: <URL/raw 경로> source_kind: official | code | normal | verbal ---## 주장 (claims)에 핵심 사실을- claim:: <내용>으로 8~12개. 기존/신규 개념 모두 claim::으로 기록 — 이 재등장 신호가 lint의 병합·신뢰도 상승·승격 입력이 된다(그래서 인제스트가 기존 페이지를 안 건드려도 된다). - 신규 개념만 L3 초안 생성. 2단계에서 없다고 판별된 개념만
wiki/L3-semantic/concept-<slug>.md(또는 entity-) 생성:
본문은 한국어 골격(요약 + 핵심 몇 줄) + 관련type: concept confidence: 0.6 # 신규 단일소스 고정 — 재계산은 lint sources: [<이 소스 슬러그>] last_confirmed: <오늘> decay_class: concept[[ ]]링크 몇 개. 관계## 관계구조화는 생략(lint 배치). 기존 개념과 겹치면 새로 만들지도, 기존을 열지도 마라 — claim::만 남기면 lint가 병합한다. - MoC 편입(가볍게, 남발 금지). 신규 L3는 기존 MoC 우선 편입(한 줄 추가). 새 MoC는 정말 새 주제 영역이고 편입 페이지 ≥2일 때만 생성 — 1페이지짜리 얇은 MoC는 만들지 말고 인접 기존 MoC나
home-moc의 "미분류" 섹션에 임시 배치. 깊은 큐레이션·MoC 재구성은wiki-cartographer/lint 몫(MoC가 12개 넘으면 자동 권고). - index.md + log.md + raw 스탬프.
- index: 신규 페이지만 추가.
- log:
## [오늘] ingest | <제목>+ 신규 N개·claim M개. - raw 스탬프: 파일 맨 위 메타만(본문 무변경) —
ingested·wiki_source: [[source-<slug>]]·ingest_status: done.
- 보고 + 지연 안내. 신규 페이지·claim 수 보고 + "병합·신뢰도·관계·승격은 다음 lint에서 배치 처리" 명시.
모델 라우팅 (비용)
이 경량 인제스트는 추출·파일쓰기 위주라 sonnet으로 충분하다(wiki-ingestor 기본 sonnet). 병합 판단·관계 추론 같은 고추론은 lint의 wiki-consolidate(opus)가 배치로 담당한다.
지연되는 일 (→ lint/consolidate)
- 기존 개념과의 병합(사실 통합·소스 추가)
- 신뢰도 재계산(2소스 0.85·3+ 0.95) — L2 claim:: 재등장 카운트로 배치 산정
- 관계
## 관계구조화 - L2→L3 승격(같은 주장 3회+)·L3→L4
상세:
.claude/skills/wiki-consolidate/SKILL.md.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

