A self-healing local knowledge base where a local LLM compounds your documents across four memory tiers ,working, episodic, semantic, and procedural — with bi-temporal facts, automatic contradiction resolution, and scheduled memory maintenance.
Instructions file CodexOpenCode
Instructions for krishddd/llm-wiki, covering agents.md — llm-wiki agent surface, quick start for an agent, tool catalogue (mcp), resources (mcp) and rest endpoints (when used directly without mcp).
Instructions file
Instructions for krishddd/llm-wiki, covering claude.md — llm-wiki v2 schema, models, multi-provider llm fleet (v4, llmwiki/providers.py), adaptive model routing (vibethinker) and domain-specialized stem embeddings (optional).