llm-wiki

To build a persistent markdown knowledge base, you can ingest new sources into raw layers and synthesize them into durable, interlinked wiki pages. This process preserves provenance and cross-references, allowing your research to compound across sessions instead of starting over.

Maintaining an interlinked markdown wiki involves performing orientation reads before ingesting or querying data to prevent duplication. You can also run linting workflows to check for orphan pages, broken wikilinks, frontmatter validation, and contradiction flags to ensure consistency.

Yes, entity tracking with a markdown wiki requires a specific folder structure containing SCHEMA.md, index.md, and log.md. The system uses a configured $WIKI_PATH, defaulting to ~/wiki, to store and organize your interlinked markdown files.

An interlinked markdown wiki serves as a RAG alternative by synthesizing durable pages with explicit wikilinks and schema validation rather than relying solely on vector retrieval. This approach preserves cross-references and provenance, compounding learning across sessions with built-in linting.

Yes, you can check a markdown wiki directory for broken wikilinks and schema validation issues by running linting and health-checking workflows. This maintenance process identifies orphan pages, validates frontmatter against SCHEMA.md, and flags contradictions to ensure knowledge base consistency.

A markdown knowledge base loses context without persistent interlinked pages and orientation reads. By applying schema validation and maintaining an index with logs during research ingestion, the system preserves cross-references and compounds learning across multiple sessions.