LLM-Wiki

This mode turns scattered information sources into a structured knowledge base. Not just a pile of notes — entity pages with schemas, templates, and cross-links.

It works on any Markdown vault — Obsidian, Logseq, or a plain git wiki.

Core Capabilities

Schema-aware entity pages. Each entity type (person, organization, concept, event, tool, project) has a fixed field structure. The AI follows the schema on every write, so your notes stay consistent.

8-template library. Eight built-in note templates cover common patterns: concept pages, comparison pages, timelines, resource lists, argument maps, reading notes, meeting notes, and decision records. Feed in a source, the AI picks the right template.

Wikilink integrity. New notes automatically link to existing pages in your Vault. You can also run integrity checks later — find orphan notes, broken links, and pages with no backlinks.

Default Tools

Tool	Purpose
Vault Search	Find existing entities, avoid duplicates
File Ops	Create and update structured Markdown files
Document Tools	Read Word, PDF, and other sources
PDF Tools	Extract PDF text as knowledge base input
Image Tools	OCR text from images
Audio Transcription	Convert audio to archivable text
Excel Tools	Process tabular data sources
HTML Tools	Archive content from web pages

Behavior

• Persistent memory: AI remembers your Vault structure preferences and schema definitions
• Workspace aware: AI sees the entire Vault for global linking and deduplication
• Vector search: Judges relevance between new and existing content by meaning

When to Use LLM-Wiki

• Turning a book, course, or article series into a knowledge base
• Maintaining structured notes for a long-term research topic
• Auditing your Vault for orphan notes and broken links
• You want consistent note formatting instead of every page looking different

Related Tool Reference

• Local Knowledge Base — all data stored locally
• Semantic Note Search — search your knowledge base by meaning