MCPFast / Tools / Engram: Local AI Identity Layer for Claude, Codex, and Cursor

GitHubTool★★★★☆

Engram: Local AI Identity Layer for Claude, Codex, and Cursor

Engram is a local, MCP-compatible AI identity layer that stores who you are, not just what you did, for Claude, Codex, and Cursor.

View on GitHub

Engram: Local AI Identity Layer for Claude, Codex, and Cursor

Engram provides a crucial local identity layer for AI models, specifically designed to enhance the context and memory capabilities of Claude, Codex, and Cursor. Unlike standard AI interactions that often reset context with each query, Engram allows these models to maintain a persistent understanding of your identity and past interactions. This is achieved through a local, MCP-compatible storage mechanism, ensuring that the AI remembers who you are and the nuances of your ongoing projects, rather than just the immediate input. This is particularly valuable for developers building complex AI applications where consistent persona and historical context are paramount.

What Engram Does

Engram acts as a persistent memory for your AI interactions. It stores information about your identity, preferences, and the history of your conversations and projects with supported AI models. This local storage means that the AI can reference past information and maintain a consistent persona across multiple sessions. For developers, this translates to more coherent and contextually aware AI assistants, reducing the need for repetitive explanations and enabling more sophisticated AI workflows.

Key Features

Who Engram is For

Engram is an essential tool for AI developers working with Claude, Codex, or Cursor. It is particularly beneficial for those building applications that require: