What problem does it solve?
AI coding assistants and personal agents lose all context between sessions, forcing them to re-ask users for preferences, project conventions, and past technical decisions every time, leading to repetitive work and inconsistent behavior.
Core Features & Use Cases
- Persistent Categorized Memory: Stores user preferences, personal facts, technical skills, project tech stacks, coding conventions, and architectural decisions in a structured, searchable format.
- Multi-User Collaboration: Supports private personal memory pools and shared team knowledge bases, with sensible default visibility per memory category.
- Flexible Access: Exposes memory operations via MCP tools for AI agent integration and a REST API for direct programmatic access, with asynchronous writes and synchronous search.
- Use Case: A developer working on a FastAPI project can have their agent recall the team's snake_case naming convention and their personal preference for 4-space indentation automatically at session start, without manual re-entry.
Quick Start
Use the neuralscape-memory skill to load all known user preferences and project-specific conventions for the current project at the start of a new coding session.