> ## Documentation Index
> Fetch the complete documentation index at: https://docs.aegismemory.com/llms.txt
> Use this file to discover all available pages before exploring further.
# What is Aegis Memory?
> The secure context engineering layer for AI agents
# Aegis Memory
**Aegis Memory** is an open-source, self-hostable context engineering layer for production AI agents. It provides content security, integrity verification, and trust hierarchy — plus persistent, semantic memory that survives context window resets.
Get running in 5 minutes
Understand the problem we solve
Learn the fundamentals
Build real-world examples
## The Problem
Every AI agent framework promises memory. Few deliver.
```python theme={null}
# What happens today
agent.chat("I'm John, a Python developer")
# ... context window resets ...
agent.chat("What's my name?") # "I don't know your name"
```
After analyzing 200+ production agent deployments, we found **40-80% fail due to memory coordination issues**—not model capability.
| Problem | Reality |
| --------------------- | ------------------------------------------ |
| Context loss mid-task | Agents forget what they learned |
| No shared state | Multi-agent teams can't coordinate |
| Memory bloat | Everything saved, nothing useful retrieved |
| Session boundaries | Long tasks restart from zero |
## The Solution
Aegis Memory provides:
4-stage pipeline: input validation, PII scanning, injection detection, LLM classification
HMAC-SHA256 tamper detection, OWASP 4-tier trust hierarchy, cryptographic agent binding
ACE patterns — self-improving agents that learn what works (voting, reflections, playbooks)
Scoped memory sharing with structured handoffs and access boundaries
## Quick Example
```python theme={null}
from aegis_memory import AegisClient
client = AegisClient(api_key="your-key")
# Store a memory
client.add(
content="User prefers dark mode and Python",
agent_id="assistant",
user_id="john_123"
)
# Later, retrieve relevant context
memories = client.query(
"What are John's preferences?",
user_id="john_123"
)
# Returns: "User prefers dark mode and Python"
```
## Framework Integrations
Aegis works with your existing agent stack:
Multi-agent teams
Chains & agents
Stateful workflows
## Open Source
Aegis Memory is fully open source under the Apache 2.0 license.
Star us on GitHub