AEGIS
AEGIS is the first OpenEnv RL environment for AI security.
Core trick
Reward is a bounded SIREN-style score from a 42-layer detection cascade, with hard floors and ceilings that block reward hacking.
Tags
Description
AEGIS is the first OpenEnv RL environment for AI security. While other submissions train agents to play games, AEGIS trains AI to defend itself — a 42-layer adversarial defense cascade where an RL agent optimizes detection thresholds to catch prompt injections, jailbreaks, and data exfiltration while minimizing false positives.
What makes it unique: • Novel domain — no RL environment for AI defense exists • Real data — 600+ adversarial prompts, not synthetic benchmarks • Auditable — every step() is cryptographically sealed with POAW (Proof of Agent Work), enabling EU AI Act compliance • Ungameable reward — SIREN-style bounded reward with hard floors/ceilings prevents reward hacking • Patent-backed — 366+ claims protect the cascade (MIT license covers the wrapper only)
AEGIS = Auditable · Energy-efficient · Governance-ready · Integrity-first · Sovereign
Built by OHM (Offline Human Mode) — sovereign AI infrastructure from Austria.
README excerpt
--- title: AEGIS Security RL Environment emoji: 🛡️ colorFrom: blue colorTo: green sdk: static pinned: false --- # AEGIS Security Environment — OpenEnv Hackathon SF 2026 # The first RL environment where AI learns to defend itself. ## �️ A.E.G.I.S. > **Auditable · Energy-efficient · Governance-ready · Integrity-first · Sovereign** | | | | | :---: | -------------------- | ------------------------------------------------------------------- | | **A** | **Auditable** | Every step recorded. POAW cryptographic proof. EU AI Act compliant. | | **E** | **Energy-efficient** | Green AI — NPU-first, 90% token compression, planetary respect. | | **G** | **Governance-ready** | ISO 42001, NIS2, 12-Sigma safety — built-in, not bolted-on. | | **I** | **Integrity-first** | 42-layer cascade, zero trust, the defense never sleeps. | | **S** | **Sovereign** | Self-hosted, your data never leaves, no cloud dependency. | ## �🔒 What is this? **AEGIS Security Environment** is an OpenEnv-compatible RL environment where an AI agent learns to optimize a 42-layer adversarial defense cascade. Unlike gaming environments (chess, Mario), this environment trains AI to **defend itself** against prompt injection, jailbreaks, and data exfiltration attacks. ### Why This Matters Every other team at the hackathon built a game or coding environment. We built the first R