arturseo-geo/llm-knowledge-base
📦 Open Source Projectarturseo-geo
A standardized schema for building LLM-compatible personal knowledge bases with integrated spaced repetition and agentic workflows.
The llm-knowledge-base repository introduces a formal schema standard designed to transform personal notes into a high-utility dataset for AI agents. At its core is the AGENTS.md specification, which defines how information should be structured to ensure maximum compatibility with LLM context windows and RAG pipelines.
Key features include:
- AGENTS.md Spec: A standardized format for metadata and content organization that helps AI agents navigate personal knowledge graphs effectively.
- Spaced Repetition Layer: An integrated system for active recall, allowing users to turn their knowledge base into a learning engine.
- PKM Integration: Designed for compatibility with tools like Obsidian, ensuring that human-readable markdown remains the primary interface while providing machine-readable structure.
- Agentic Workflows: Templates and examples that demonstrate how to query, summarize, and expand upon personal notes using AI agents, effectively turning a static knowledge base into a dynamic, evolving intelligence partner.
💡Highlights
- ├─AGENTS.md schema standard
- ├─Built-in spaced repetition layer
- └─Obsidian-compatible markdown
🎯For
- ├─Knowledge Management Enthusiasts
- ├─AI Developers
- └─PKM Power Users