LLM Wiki

llm-wiki is a toolkit for creating knowledge bases that are compiled and maintained by AI agents. Rather than being a standalone app, it ships as a plugin and skill set for agent environments such as Claude Code, OpenAI Codex and OpenCode, plus portable agents, and the resulting wikis are compatible with Obsidian.

The workflow covers capturing rough ideas, researching and shaping them, and promoting approved briefs into delivery projects. It includes parallel multi-agent research, thesis-driven investigation, collector catalogs, session memory, source ingestion, wiki compilation, audits, querying at different depths and artifact generation. Recent releases add lint exit gates, project knowledge exports, cross-runtime packaging and an optional framework for personal specialist review methods.

The project is written in Python and licensed under MIT, with a website and a changelog documenting frequent releases. It runs locally with your own agent subscription or API access. It suits developers and researchers who already work with AI coding agents and want a persistent, linked, reviewable wiki instead of scattered chat transcripts.