wrannaman/rememberall

Provides semantic search over stored memories using vector embeddings, enabling GPTs and AI agents to retrieve relevant past context through a simple REST API with JWT authentication. Deploys via Docker Compose with end-to-end encryption and self-hosted vector storage, avoiding third-party memory dependencies. Integrates directly into OpenAI's custom GPT builder via the `@rememberall` mention syntax, with optional developer integration for building memory-aware applications.

123 stars. No commits in the last 6 months.