agentic-knowledge-graph

Overview ¶

Example: Autonomous Agent with Knowledge Graph Memory

This example demonstrates an Agent that:

1. Stores its observations and tool-call results as nodes in a knowledge graph
2. Connects related observations with typed edges
3. Uses vector search to retrieve semantically relevant context before each action
4. Navigates the graph (BFS/Dijkstra) to find reasoning chains

This is the foundational pattern for memory-augmented autonomous agents. In production you would replace mockEmbed() with a real embedding call (e.g., BGE-M3 via Ollama or OpenAI text-embedding-3-large).