What problem does it solve? Standard RAG workflows re-derive answers from raw documents on every query with no knowledge accumulation. This Skill sets up a persistent, interlinked markdown wiki that an LLM agent incrementally builds and maintains, so knowledge compounds across sources instead of being re-computed each time. ## Core Features & Use Cases - Guided wiki setup: Walks through choosing a NanoClaw group, designing the wiki domain, and creating the wiki/ and sources/ directory structure with index.md and log.md files. - Schema and agent configuration: Generates a container skill (container/skills/wiki/SKILL.md) and a marker-wrapped CLAUDE.md section that turn the agent into a disciplined wiki maintainer with ingest, query, and lint operations. - Source handling and maintenance: Checks whether the agent can handle planned source types (PDFs, images, voice notes, URLs) and optionally schedules periodic lint health checks. - Use Case: A researcher wants a companion wiki while reading papers over several months. Run this Skill to create the wiki structure, then drop in PDFs one at a time; the agent summarizes, cross-references entities, updates the index, and logs every ingest. ## Quick Start Ask the agent to add a Karpathy-style LLM wiki knowledge base to your NanoClaw group and follow the guided setup prompts.
