One command.
Your device is secured.
The mona.expert wrapper is the AI Security Layer that runs on your machine. Every LLM call gets inspected for prompt injections (21 patterns), every access is audited via SHA-256 chain, every metric is streamed to your dashboard. No data leaves without your wrapper's consent.
curl -fsSL https://mona.expert/install.sh | bash
The wrapper includes
Injection Guard
21 detection patterns block prompt injections, jailbreaks, and system prompt leaks before they reach the LLM.
mTLS Gateway
Mutual TLS between your wrapper and mona.expert. Certificate-based authentication — no shared secrets.
Audit Chain
Every LLM call is hashed into a SHA-256 chain. Tamper-proof audit trail — verify integrity at any time.
Live Metrics
CPU, memory, disk, network, agent status — streamed to your dashboard every 120 seconds. Real-time visibility.
Agent Runtime
Run AI agents locally with built-in sandboxing, rate limiting, and consent management. No vendor lock-in.
Open Source
Full source available under MIT license. Self-hostable. Community-audited — not by marketing slides.
Install in three commands
From zero to secure in under 60 seconds. No configuration needed.
Install the wrapper
curl -fsSL https://mona.expert/install.sh | bash
One command. The installer checks prerequisites, installs dependencies, and sets up your local wrapper.
Connect to mona.expert
mona-expert connect YOUR_TOKEN
Paste the token from your account. The wrapper performs a secure handshake and establishes an encrypted tunnel.
Start monitoring
npm run sync:watch or npm start
The wrapper starts streaming telemetry to your dashboard. Audit events, system metrics, security checks — live.
Systemanforderungen
Ready to secure your AI agents?
Create your account, get your token, and install the wrapper. No credit card. No commitment. Open source forever under MIT license.