SS 360 SAS - Open Source Security & AI Framework

Overview

This repository contains a subset of the core engineering utilities used by SS 360 SAS to power our Cybersecurity as a Service (CSaaS) infrastructure. Our goal is to provide the community with high-performance, asynchronous tools for AI orchestration and security observability.

Key Components

• Async-LiteLLM Adapters: Optimized routing for LiteLLM with a focus on low-latency failover.
• Semantic Cache Layer: Integration with Valkey to minimize token consumption and redundant LLM calls.
• Security Telemetry Parsers: Python modules for structured logging compatible with Promtail, Loki, and Grafana.
• Zero-Trust Utility: Scripts for automated Caddy/Cloudflare Tunnel provisioning.

Tech Stack

• Language: Python 3.11+ (Asynchronous/FastAPI)
• Cache: Valkey (High-performance Redis alternative)
• Infrastructure: Docker / Caddy Server
• Observability: Loki / Grafana

Why we are contributing

At SS 360, we believe that AI security must be proactive. By open-sourcing our orchestration layer, we aim to standardize how AI agents interact with sensitive infrastructure while maintaining strict adherence to Zero Trust principles.

Getting Started

Prerequisites

• Python 3.11
• Valkey (or Redis) instance

Installation

pip install -r requirements.txt

Running the Gateway

export VALKEY_HOST=localhost
uvicorn src.main:app --reload

License

Distributed under the MIT License. See LICENSE for more information.