Back to directory

LLM Knowledge Bases

A Claude Code plugin that compiles raw research material into an LLM-maintained Obsidian wiki, then answers questions from it at three depth levels and keeps the wiki healthy over time.

LLM Knowledge Bases

This Claude Code plugin turns raw source material dropped into a folder, articles, papers, repositories, transcripts, images or datasets, into a structured Obsidian vault of interlinked wiki articles, summaries, backlinks and generated indexes, maintained by Claude rather than edited by hand. The workflow follows five steps: ingest raw documents, compile them into the wiki, query the knowledge base at a chosen depth, output the answer as a markdown report, slides or a chart, and run automated maintenance checks for broken links, orphan pages and inconsistencies.

Queries support three depth levels: quick answers drawn only from wiki indexes and summaries, standard answers that cross-reference the full wiki and supplement it with web search, and deep answers produced by a multi-agent research pipeline running parallel web-search agents. It installs as a single plugin through Claude Code's plugin marketplace command, requires Obsidian, and automatically installs supporting Obsidian plugins during its own initialization command.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.