llm-wiki

Build and maintain a persistent markdown knowledge base from ingested source documents.

Updated Feb 23, 2026
One-click install
npx skills add https://github.com/deepparser/skills --skill llm-wiki-deepparser
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: llm-wiki
Source: https://github.com/deepparser/skills/tree/main/.agents/skills/llm-wiki
Command: npx skills add https://github.com/deepparser/skills --skill llm-wiki-deepparser

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Re-deriving answers from raw sources on every query wastes effort and loses cross-references; this Skill builds a persistent, interlinked markdown wiki that compounds knowledge as sources are ingested. ## Core Features & Use Cases - Wiki Initialization: Scaffold a wiki directory with raw/, wiki/, index.md, log.md, and a domain-customized SCHEMA.md using the included init script and schema template. - Incremental Ingestion: Process new source documents into summary, entity, concept, and synthesis pages with Obsidian-compatible [[wikilinks]] and YAML frontmatter. - Query and Lint Operations: Answer questions with citations from wiki pages and health-check the wiki for contradictions, orphan pages, stale claims, and missing cross-references. - Use Case: A researcher collecting papers on a topic adds each PDF to raw/, ingests it to update entity and concept pages, then queries the wiki for synthesized comparisons with citations. ## Quick Start Ask the assistant to initialize a new LLM wiki for your chosen topic and then ingest your first source document from the raw folder.

Frequently Asked Questions about llm-wiki

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

FAQPage Schema
How do I build a personal knowledge base with an LLM?

Use the LLM Wiki pattern: initialize a wiki directory with raw sources and LLM-maintained markdown pages, then incrementally ingest each source. The LLM creates summary, entity, and concept pages with cross-references that persist across queries.

How to ingest documents into a markdown wiki?

Place the source file in the raw/ directory and run the ingest operation. The LLM reads SCHEMA.md and index.md, creates a source summary page, updates related entity and concept pages, and appends an entry to log.md.

Does the LLM wiki work with Obsidian?

Yes, the wiki uses Obsidian-compatible [[wikilinks]] for all internal cross-references and plain markdown files with YAML frontmatter. You can open the wiki directory as an Obsidian vault and use graph view to visualize connections.

What does the wiki lint operation check for?

Lint checks for contradictions between pages, stale claims superseded by newer sources, orphan pages with no inbound links, missing cross-references, concepts lacking their own pages, and outdated frontmatter. Findings are presented as a prioritized list and fixed with user approval.

What are the limitations of a markdown-based wiki at scale?

The index.md file is sufficient for navigation only at small scale, roughly under 100 sources. For larger wikis, the documentation recommends adding dedicated search tooling such as qmd to find relevant pages efficiently.