llm-wiki

A persistent knowledge base prevents fragmented, ephemeral research by automatically cross-referencing and maintaining consistency across markdown notes. Unlike traditional RAG systems, it compounds information over time rather than rediscovering it from scratch.

Yes, this Skill integrates with local or remote editors like Obsidian by managing markdown files directly in your file system. The agent handles the heavy lifting of filing, linking, and linting notes while you focus on high-level research.

You can start by asking the agent to initialize a new wiki in your home directory, which guides you through the initial schema setup. Once initialized, feed it raw sources to automatically synthesize structured markdown entity and concept pages.

Yes, this Skill requires file system access to manage markdown files and integrate with local or remote editors. This allows the agent to directly handle markdown management, filing, and linting within your knowledge base.

A compounding knowledge base synthesizes raw sources into interlinked markdown pages to build persistent recall, whereas a traditional RAG system rediscovers information from scratch. This approach ensures information consistency and cross-references across sessions.

The primary limitation is the strict dependency on file system access and compatible markdown editors like Obsidian for integration. Without this environment, the agent cannot perform the automated filing, linking, and linting required to maintain the wiki.