Vaikora AI Agent Signals for Microsoft Defender for Cloud
by Data443
Ingest Vaikora AI agent signals into Microsoft Defender for Cloud as security alerts via Logic App.
The Vaikora AI Agent Signals connector for Azure Security Center ingests AI agent behavioral signals from Vaikora into Microsoft Defender for Cloud.
The connector polls the Vaikora API every 6 hours for high-risk agent actions, anomalies, and detected threats. It filters signals where severity is high or critical, an anomaly is flagged, a threat is detected, or the Vaikora policy decision is "block". Qualifying signals are written to a custom Log Analytics table (Vaikora_SecurityAlerts_CL) that Defender for Cloud analytic rules continuously monitor.
Each signal carries behavioral anomaly scores across seven risk dimensions: action type risk, agent profile risk, temporal risk, environmental risk, behavioral risk, compliance risk, and data sensitivity risk. Signals also include the Vaikora policy decision (allow, block, or flag) and a tamper-proof audit hash for log integrity verification.
Included analytic rules:
- Vaikora - High Severity Security Alerts — triggers on any high or critical severity agent action in the last 6 hours (Severity: High)
- Vaikora - Anomaly Detection — triggers on behavioral anomaly or threat detection below the high/critical threshold (Severity: Medium)
- Vaikora - Feed Outage Detection — triggers when no signals are received within the expected polling window (Severity: Low)
Prerequisites: Active Vaikora account with API key, Azure Log Analytics workspace (workspace ID and primary key required), Logic App Managed Identity granted Monitoring Metrics Publisher role on the DCR.
Built by Data443 Risk Mitigation, Inc. — data443.com | vaikora.com
