AI Enablement
How Guard's AI works, from Query Mode data questions to Agent Mode and autonomous offensive security agents.
- AI OverviewHow Guard's three AI tiers differ: Query Mode for questions, AI Automation in the background, and Agent Mode for operations.
- Aurelius: The AI OperatorHow Aurelius plans an operation, confirms it with you before running jobs, and delegates work to specialist agents.
- Query ModeAsk the AI Assistant plain-language questions about your assets, risks, ports, and technologies in Query Mode.
- Agent ModeTurn on Agent Mode so Aurelius can run scans and delegate work to specialist agents, for Analyst and Admin users.
- Reconnaissance AgentsHow Asset Analyzer and Aurelian, the Agent Mode recon specialists, analyze one asset in depth and inventory your cloud.
- Scanning AgentsHow Julius, Titus, Trajan, and Vespasian fingerprint LLM services, find secrets, scan CI/CD pipelines, and discover APIs.
- Offensive AgentsHow Brutus and Augustus, the Agent Mode offensive specialists, test credentials and attack LLM applications to prove risk.
- Validation AgentsHow Cato, Priscus, and Seneca validate findings, retest fixed vulnerabilities, and research CVEs in Agent Mode.
- Automated AgentsHow Guard's background AI agents, such as Auto-Triage, review new findings without anyone having to ask.
- Asset Affiliation AgentHow Guard verifies whether discovered assets and DNS records belong to you and flags dangling records at risk of takeover.
- AI Data PrivacyHow Guard decides what data goes into AI prompts and anonymizes finding text before storing it for search.