llm-wiki

Build and maintain interlinked markdown knowledge bases with schema-aligned metadata.

Updated Apr 9, 2026
One-click install
npx skills add https://github.com/MarbleSodas/Mavis --skill llm-wiki-marblesodas
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: llm-wiki
Source: https://github.com/MarbleSodas/Mavis/tree/main/skills/research/llm-wiki
Command: npx skills add https://github.com/MarbleSodas/Mavis --skill llm-wiki-marblesodas

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you turn scattered sources, notes, and research into a persistent, cross-referenced markdown knowledge base that keeps improving over time instead of restarting from scratch per query.

Core Features & Use Cases

  • Compounding wiki knowledge base: builds an interlinked network of markdown pages (entities, concepts, comparisons, and filed query results) that grows as you ingest more sources.
  • Source-governed ingestion and maintenance: organizes immutable raw sources separately from agent-owned wiki pages, with provenance and drift detection designed into the workflow.
  • Session orientation, querying, and auditing: reads schema/index/log for continuity, supports answering from the wiki, and provides linting for orphans, broken wikilinks, index completeness, contradictions, stale content, and tag taxonomy issues.

Quick Start

Ask the agent to initialize a new wiki at your preferred location and start ingesting your first set of research sources so it creates SCHEMA.md, index.md, log.md, and the initial set of interlinked pages.

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 persistent markdown knowledge base from research sources?

You build a persistent markdown knowledge base by ingesting URLs, PDFs, and pasted text into interlinked markdown files with wikilink cross-referencing and provenance markers. The wiki compounds over time, maintaining continuity through schema, index, and log files instead of restarting per query.

What is the best way to organize raw research sources separately from wiki notes?

Organizing raw research sources separately from wiki notes requires storing immutable original sources apart from agent-owned wiki pages. This workflow uses provenance markers and drift detection to maintain source governance while updating the interlinked markdown wiki.

How does wiki linting detect broken wikilinks and orphan pages in markdown files?

Wiki linting detects broken wikilinks and orphan pages by auditing markdown file health against schema-aligned metadata. It checks for index completeness, contradictions, stale content, and tag taxonomy issues to maintain wiki integrity.

Can I use this markdown wiki approach as a RAG alternative for answering questions?

Yes, you can use this interlinked markdown wiki as a RAG alternative for answering questions. The agent reads the existing wiki schema, index, and recent logs to retrieve answers directly from the cross-referenced knowledge base.

How do I ingest PDFs and URLs into an interlinked markdown wiki?

You ingest PDFs and URLs by asking the agent to process the sources and create or update wiki pages with frontmatter validation, tag taxonomy, and wikilink cross-referencing. The system separates immutable raw sources from the generated wiki content.

What are the limitations of maintaining a self-compounding markdown wiki?

Maintaining a self-compounding markdown wiki requires session orientation by reading schema, index, and log files for continuity, and depends on controlled updates with frontmatter validation to prevent orphans, broken wikilinks, and tag taxonomy conflicts.