Provides AI agents with persistent, infinite long-term memory across sessions and various tools.
Sponsored
Neural Vault Core is a robust MCP (Model Context Protocol) memory server designed to combat the inherent forgetfulness of AI agents. It addresses the challenges of expensive, finite context windows and the need for a unified memory system across different AI tools. By serving as a single, fast, and searchable memory vault, Neural Vault Core allows any MCP-compatible AI agent—from Claude Code to Cursor and ChatGPT—to store and recall information persistently, ensuring that context is never lost and knowledge accumulates across projects and sessions.
Key Features
01Namespaces to organize memories by project, context, or category.
02Full-text search capabilities using SQLite FTS5 for efficient retrieval.
034 core MCP tools: store, retrieve, search, and delete memories.
04Auto-versioning that keeps the last 5 edits for each memory.
05Permanent storage ensuring memories never expire or auto-delete.
060 GitHub stars
Use Cases
01Prevent AI agents from losing context and forgetting information between sessions.
02Manage and extend AI context without exceeding expensive token limits.
03Provide a single, unified memory vault accessible by any AI tool on a network.
