Long Term Memory
Provides long-term memory storage for LLMs, enabling them to retain context across multiple sessions by leveraging semantic search and embeddings.
About
Long Term Memory offers a robust system for LLMs to maintain project context throughout its entire lifecycle. By using the Model Context Protocol (MCP) standard, it enables LLMs to recall past interactions and development decisions, even across sessions. This is achieved through semantic search powered by embeddings, allowing for retrieval of relevant context from historical data to assist with ongoing tasks and decision-making.
Key Features
- Organizes memory by project, ensuring context is maintained within specific projects.
- Supports multiple memory types, including conversations, code, decisions, and references.
- Implements a tagging system for improved memory organization and retrieval.
- Utilizes semantic search with embeddings (Ollama, nomic-embed-text model) to retrieve relevant memories.
- Stores rich metadata, such as implementation status, key decisions, and file modifications.
Use Cases
- Allowing LLMs to access and understand the history of a project during development.
- Providing LLMs with access to relevant code changes, files, and dependencies for context.
- Enabling LLMs to make informed decisions based on previous conversations and architectural choices.
Sponsored