LLM Wiki

LLM Wiki is a knowledge base tool for AI agents. It compiles topic wikis from source material and supports parallel multi-agent research, thesis-driven investigation, source ingestion, wiki compilation, session memory, feedback curation, topic archiving, inventory tracking, dataset manifests, truth-seeking audits, token-efficient read-only querying, and artifact generation.

Its workflow is built around turning raw sources into cross-referenced articles and then into outputs such as reports, slide decks, study guides, playbooks, implementation plans, timelines, glossaries, and comparisons. The tool can dispatch up to ten agents, search academic, technical, applied, news, and contrarian angles, and run a thesis mode that splits agents across supporting, opposing, mechanistic, meta, and adjacent perspectives to produce a verdict rather than a summary. It also ingests URLs, files, PDFs, inbox drops, Git doc repos, MediaWiki dumps, message archives, and Wayback CDX snapshots. Raw sources remain immutable while articles are synthesized on top, and the system can dedupe, catalog artifacts, track durable follow-up state, index large external data with manifests and profiles, and maintain human-owned schema.md topic guides.

LLM Wiki ships as a Claude Code plugin, an OpenAI Codex plugin, an OpenCode instruction file, or a portable AGENTS.md. It is Obsidian-compatible. The documentation also describes a full workflow and a read-only querying path, with compact file-cited lookup available through wiki-query. Sessions are captured under .sessions/ with redacted events, state JSON, and Markdown digests, and feedback, librarian scoring, audit, lessons, and plan-generation functions are described as part of the same system.

The page refers to it as llm-wiki by nvk. No pricing or license terms are stated in the provided material.