Enables persistent, graph-based memory for Claude Code to retain context and learn from previous sessions.
The Auto-Claude Memory system utilizes Graphiti and an embedded LadybugDB to provide long-term context retention for AI-driven development. It enables Claude to build a project-specific knowledge graph, storing insights as 'episodes' that can be retrieved via semantic search in future sessions. This eliminates the 'forgetting' problem in complex projects, allowing the AI to remember specific patterns, architectural decisions, and previous bug fixes without requiring a Docker setup. It supports a wide range of providers including OpenAI, Anthropic, and local Ollama instances for full privacy and offline use.
Key Features
01MCP tools for manual memory management and automated insight extraction
02Embedded LadybugDB storage with no Docker dependency
03Multi-provider support for LLMs and embeddings (OpenAI, Anthropic, Ollama, etc.)
04Persistent cross-session knowledge graph using Graphiti
05Semantic search for retrieving relevant entities and relationships
060 GitHub stars
Use Cases
01Building a searchable knowledge base of project-specific 'gotchas' and solutions
02Implementing offline AI memory using local Ollama models and embeddings
03Retaining architectural context and patterns across long-term coding projects
