Setting up your first AI agent

CybAura ships with 11 AI security agents. Each one covers a specific domain of your security programme and is scored for suitability based on your DNA. This guide shows you how to enable and configure your first agent.

1. Open the Agent Marketplace

Go to Agents → Marketplace. Every agent card shows a short description and a suitability score — start with the agents scored highest for your organisation.

2. Enable an agent

Click an agent (the SOC Analyst is a good first choice) and select Enable. Depending on your plan, agents may be included or available as an add-on — see Understanding agent SKUs and billing.

3. Configure scope and autonomy

Each agent has settings that control how it works:

• Schedule — how often the agent runs (e.g. every 15 minutes, hourly, daily).
• Autonomy level — whether the agent acts automatically on low-risk findings or escalates everything for human approval. Agents never take irreversible actions without approval.
• Escalation owner — who receives findings that need a human decision.

4. Review the first run

After the first run, open the agent's detail view to see what it did. Findings appear in your Agent Activity Feed and, where relevant, as escalations with full context and a recommended action.

5. Approve or adjust

For each escalation you can approve, dismiss, or edit the recommended action. The agent learns from your decisions and your feedback improves its future confidence scoring.

Good to know

• You can enable multiple agents — they share context through your DNA and asset graph.
• Disabling an agent stops new runs but preserves its history.
• Every agent action is logged in the audit trail for compliance evidence.

Tip: Start with one or two agents, get comfortable with the escalation flow, then expand. Enabling all 11 at once can be a lot of activity to review on day one.