ebrain

<div align="center">

<img src="https://img.shields.io/badge/E--Brain-Persistent_Memory_for_AI_Agents-0066ff?style=for-the-badge&logo=rust&logoColor=white" alt="E-Brain">

# E-Brain

### The memory layer that makes AI agents remember everything.

*Across sessions. Across compactions. Across projects.*

[![Built with Rust](https://img.shields.io/badge/Rust-000000?style=flat-square&logo=rust&logoColor=white)](https://www.rust-lang.org/)
[![MCP Compatible](https://img.shields.io/badge/MCP-Compatible-7C3AED?style=flat-square)](https://modelcontextprotocol.io)
[![Status](https://img.shields.io/badge/Status-Active_Development-22c55e?style=flat-square)]()
[![License](https://img.shields.io/badge/License-Proprietary-ef4444?style=flat-square)]()
[![Memories](https://img.shields.io/badge/107K+-Memories_in_Production-f59e0b?style=flat-square)]()

[Blog Post](https://engram.hashnode.dev/memory-crisis-ai-agents) | [Discussions](https://github.com/orion1139/ebrain/discussions) | [Issues](https://github.com/orion1139/ebrain/issues)

---

</div>

## The Problem

Every AI coding agent has the same fatal flaw: **amnesia**.

- Context windows are finite. Your codebase is not.
- Auto-compaction silently destroys architectural decisions, bug history, and project context.
- Every new session starts from zero.
- Existing memory solutions (Mem0, Letta, Zep) are built for chatbot personas — not for the demands of real software engineering.

> If you've watched an AI agent forget a critical decision it made 20 minutes ago because the context window compacted, you understand why E-Brain exists.

## How E-Brain Works

E-Brain is a **server-side memory backend** designed specifically for AI coding agents. It ingests everything the agent sees and does, organizes it across multiple cognitive memory tiers, and serves precisely the right context back — fast enough to be invisible.

**Not a wrapper. Not a prompt hack. A dedicated memory architecture.**

```