Skip to main content

Aegis – Zero-Trust edge proxy with Shannon entropy and Linux RAPL telemetry

Kagi Small Github
17 lis 2026, 14:35Tech
Oryginał
2 min read

AIVO Labs has released Aegis, a unified zero-trust security and environmental governance suite designed for SMBs and mid-market companies, featuring post-quantum cryptography and hardware energy monitoring.

Built with native Go and Rust, Aegis runs as an ultra-frugal transparent consuming less than 200 MB of RAM (excluding local LLM weights) and imposes no per-seat SaaS fees. The platform has been engineered to protect information systems while auditing and reducing environmental footprint under the Ethical Open-Core model.

The suite includes a live hardware benchmark feature that demonstrates its effectiveness against the CVE-2024-4577 (PHP2Shell) vulnerability. Tests on an Intel Core i5-8350U CPU showed 602 blocked requests compared to 126 unshielded exploits when using the Aegis protection.

Architecture & Services

Aegis is structured as an autonomous zero-trust stack with no dependencies on US Big Tech clouds. Its core components include an Ingress Policy Enforcement Point (PEP) & Web Application Firewall (WAF) that provides ultra-low latency protection against OWASP Top 10 threats, virtual patching, and Shannon entropy analysis.

Ultra-low latency (< 1.5 ms) OWASP Top 10 WAF, Virtual Patching, and Shannon Entropy analyzer.

AIVO Labs, developer

The platform features a unified Security Operations Center (SOC) and ESG Cockpit that displays real-time cyber threat maps, live attack logs, instant WAF controls, and hardware energy metrics. It also includes a community portal providing bilingual support (English/French) with one-click access to the SOC Cockpit and documentation.

Deployment & Testing

Aegis can be deployed on Linux systems (Ubuntu 22.04/24.04, Debian 12, Rocky Linux 9) or macOS with standard Docker requirements. The minimum hardware specification is 2 vCPUs, 2 GB RAM, and 20 GB SSD for basic operation, while full functionality with local inference requires 4 vCPUs and 8 GB RAM or compatible GPU.

The platform includes an automated pentest suite that simulates real-world attack vectors, successfully neutralizing automated web scanners, SQL injections, local file inclusion attempts, remote code execution, cross-site scripting, and DNS exfiltration. Legitimate traffic continues to be forwarded with sub-1.5 ms latency.

Enterprise Support

For enterprise environments, Aegis supports Kubernetes deployments with granular per-namespace power attribution via Kepler eBPF. It can be installed with a single Helm command or applied through pure manifests for GitOps workflows. The platform is designed for hassle-free deployment on VPS and container PaaS platforms like Dokploy, Coolify, and Traefik.

Distributed under AGPLv3 licensing, Aegis is fully auditable and modifiable. Developed in France by AIVO Labs, it is immune to the extraterritorial US Cloud Act and built for regulatory compliance including NIS 2, DORA, RGESN 2024, and the European VS Standard (EFRAG 2026).

Z.ai
Article written by AI
Z.ai GLM-4.5-FlashFree
Free AI articles150 of 150
No limit after signing in with Telegram