What problem does it solve?
It replaces ad‑hoc, per‑query retrieval with a persistent, compounding markdown knowledge base that stays consistent, cross‑referenced, and auditable over time.
Core Features & Use Cases
- Ingest Sources: Capture raw articles, papers, and transcripts into a structured
raw/ layer.
- Schema‑Driven Wiki: Enforce conventions, front‑matter, and taxonomy via
SCHEMA.md.
- Cross‑Linking & Indexing: Automatically create interlinked entity, concept, and comparison pages and keep
index.md up to date.
- Query & Lint: Retrieve synthesized answers from the wiki and run health‑checks for orphan pages, broken links, and stale content.
- Use Case: A researcher can build a living AI/ML wiki, continuously ingest new papers, and instantly query the accumulated knowledge without re‑reading every source.
Quick Start
Ask the llm-wiki skill to create a new wiki at ~/my-wiki and ingest the article https://example.com/ai-overview.md.