What problem does it solve? Karpathy-pattern LLM wikis accumulate raw sources, markdown articles, and wikilinks over time, but their implicit structure and cross-references remain hard to explore. This Skill parses the wiki deterministically, enriches it with LLM-extracted entities and claims, and produces an interactive knowledge graph dashboard. ## Core Features & Use Cases - Deterministic Wiki Parsing: Detects the Karpathy three-layer pattern (raw sources, wiki markdown, schema file), extracts wikilinks, headings, frontmatter, and categories from index.md into a scan manifest. - LLM-Enriched Analysis: Dispatches batched article-analyzer subagents to extract implicit entities, claims, and relationships that wikilinks alone do not capture. - Graph Assembly and Dashboard: Merges scan and analysis results with entity deduplication, edge normalization, layer building, and tour generation, then saves a validated knowledge-graph.json and triggers the dashboard. - Use Case: Point the Skill at a research wiki containing hundreds of markdown notes with wikilinks; it produces a force-directed knowledge graph showing articles, topics, entities, and claims organized by category. ## Quick Start Run the understand-knowledge skill on my wiki directory to build an interactive knowledge graph of all articles and their relationships.
