skills.rest
lfuuu
lfuuuCOMMUNITY
Explore all skills by this author →
GitHub

kb

Builds and maintains a markdown wiki knowledge base from raw source documents.

Updated Jun 2, 2026
Search Tags:#summarization#markdown#knowledge base#obsidian#wiki#ingest#personal knowledge management
One-click install
npx skills add https://github.com/lfuuu/claude-rules --skill kb-lfuuu
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: kb
Source: https://github.com/lfuuu/claude-rules/tree/main/global-skills/knowledge-base
Command: npx skills add https://github.com/lfuuu/claude-rules --skill kb-lfuuu

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Raw articles, papers, and notes pile up without structure, making it hard to retrieve, connect, and reuse the knowledge they contain. This Skill compiles scattered source documents into an organized, cross-linked markdown wiki that you can query, lint, and turn into polished outputs. ## Core Features & Use Cases - Ingest and Compile: Process raw sources (.md, .txt, .pdf, .html, images) into summaries, then synthesize concept articles with Obsidian-compatible wikilinks, categories, and a master index. - Query and Lint: Answer questions grounded in wiki content with citations, and run health checks detecting broken links, orphan articles, stale summaries, and thin content. - Generate Outputs: Produce articles, Marp slide decks, reports, Mermaid diagrams, tables, and timelines from wiki data, saved to an outputs directory. - Use Case: Drop a folder of research papers into raw/, run ingest and compile, then ask questions across all papers or generate a slide deck summarizing a topic. ## Quick Start Ask the agent to initialize a knowledge base in the current directory and ingest the documents in the raw folder.

Frequently Asked Questions about kb

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create a personal knowledge base from raw documents?▼

Run the init command to create the directory structure with raw/, wiki/, and outputs/ folders plus a kb.json metadata file. Then place source documents in raw/ and run ingest followed by compile to generate cross-linked concept articles.

How to query a markdown wiki with an AI assistant?▼

Use the query command with your question. The skill reads the wiki index and categories, retrieves relevant articles via grep, and synthesizes a cited answer grounded in the wiki content, optionally saving substantial answers to the outputs folder.

What file formats can be ingested into the knowledge base?▼

The ingest command supports .md, .txt, .pdf, .html, and image files placed in the raw/ directory. Each source gets a summary with frontmatter recording the source path, ingestion date, and auto-detected tags.

Does the wiki work with Obsidian?▼

Yes, all internal cross-references use Obsidian-compatible wikilinks in the [[concept-name]] format. The generated wiki folder can be opened directly as an Obsidian vault with working graph and backlink features.

How do I find broken links and gaps in my wiki?▼

Run the lint command to generate a health report. It detects broken wikilinks, orphan articles, stale summaries, thin articles under 200 words, duplicate content, and suggests new article candidates based on undefined terms.

Related Skills

Explore More Education & Research Skills→
SebasArenSebasAren
2

obsidian-wiki-ingest

Classify raw materials and generate wiki pages from PDFs, notes, and URLs.

Community
Advanced
erlebacherlebach

marp-deck-visual-designer

Translate MARP deck plans into Mermaid diagrams or Excalidraw visuals.

Community
Advanced
bordenetbordenet
8

wiki-debunker

Validate wiki page claims against git history, issue trackers, and meeting transcripts.

Community
Advanced
JacobJ215JacobJ215

AgentDB Vector Search

Perform semantic vector search over large document collections with AgentDB.

Community
Advanced