Aegis

What it does

Aegis is a concurrent AI investigation swarm for detecting scams and phishing attacks. Instead of relying on a single sequential analysis, Aegis coordinates multiple specialized agents that independently investigate signals such as suspicious URLs, email content, sender behavior, language/social-engineering indicators, and threat intelligence. Their findings are coordinated through a shared event-driven runtime and combined into a final risk assessment and verdict.

How the agents run concurrently

Aegis uses an event-driven multi-agent architecture built with @mozaik-ai/core. When an investigation starts, multiple specialized agents join the same runtime and begin analyzing the incident concurrently rather than waiting for one agent to finish before starting another. Agents include URL, email, sender, language/social-engineering, and threat-intelligence investigators. They communicate through shared runtime state and semantic events, allowing findings and investigation signals to be coordinated as agents complete their analysis. The runtime then combines the collected results into a final Aegis risk verdict.