What problem does it solve?
It eliminates the inefficiency of repeatedly regenerating answers from scratch by turning scattered sources into a persistent, interlinked markdown knowledge base that compounds over time.
Core Features & Use Cases
- Persistent interlinked wiki: Maintains entity, concept, comparison, and query pages that cross-reference each other with wikilinks.
- Ingest-to-synthesis workflow: Ingests sources into an immutable raw layer, then synthesizes and updates curated wiki pages with provenance markers.
- Orientation + quality control: Orients on SCHEMA.md, index.md, and recent log entries to avoid duplicates, and supports linting/auditing (orphans, broken wikilinks, frontmatter validation, drift, contradictions).
- Use cases: Build research notes for a domain (papers, articles, transcripts), answer questions using the wiki’s compiled structure, and file new deep-dive comparisons or query results back into the vault.
Quick Start
Ask the agent to initialize a new LLM wiki at your configured path and then ingest your first set of sources so it can create SCHEMA.md, index.md, and log.md.