Activity Monitoring
Real-time monitoring and analytics for AI agents and MCP servers. Track verification events, analyze usage patterns, detect anomalies, and gain actionable insights into your agent ecosystem.
What is Activity Monitoring?
AIM's Activity Monitoring provides comprehensive real-time tracking and analytics for all agent operations, verification events, and system interactions. Gain visibility into your AI infrastructure with detailed metrics, trend analysis, and security insights.
Real-Time Visibility
Monitor all agent activities as they happen with live event streams, instant alerts, and real-time dashboards showing current system state.
Historical Analysis
Analyze patterns over time with customizable time ranges, trend detection, and comparative analytics to understand system evolution.
Verification Tracking
Track every cryptographic verification, action approval, and security check with full audit trail and temporal analysis.
Usage Analytics
Understand API usage patterns, agent behavior, resource consumption, and operational trends across your organization.
Anomaly Detection
Automatic identification of unusual patterns, suspicious activities, and potential security threats with ML-powered detection.
Unified Activity Timeline
NEWA comprehensive activity feed that aggregates events from multiple sources into a single, chronological timeline. See all agent registrations, MCP server activities, verification events, and security alerts in one unified view.
Event Types Tracked
- • Agent registrations and verifications
- • Agent suspensions and reactivations
- • MCP server registrations and attestations
- • Capability grants and revocations
- • Verification success/failure events
- • API key operations
Timeline Features
- • Color-coded severity indicators
- • Entity type badges (Agent/MCP/Verify)
- • Trust score display per event
- • Clickable navigation to entity details
- • Real-time updates
- • Automatic deduplication
Intelligent Behavioral Analysis
NEWAdvanced pattern detection that learns from agent behavior to identify anomalies and potential security threats. Uses per-agent baselines rather than fixed thresholds.
Baseline Learning
Learns normal patterns for each agent over time, detecting deviations from established behavior.
Pattern Recognition
Identifies unusual API call patterns, off-hours access, and suspicious activity sequences.
Trust Impact
Automatically adjusts trust scores based on behavioral analysis findings.
Real-Time Monitoring Capabilities
AIM provides several mechanisms for real-time activity monitoring:
Live Dashboards
- • WebSocket-based live updates
- • Auto-refreshing metric cards
- • Real-time event streams
- • Instant alert notifications
Webhook Integration
- • Event-driven notifications
- • Custom webhook endpoints
- • Slack/Teams integration
- • Third-party monitoring tools
Monitoring Best Practices
✅ Recommendations
- • Set up automated alerting for critical metrics
- • Review verification trends weekly
- • Monitor trust score distributions regularly
- • Track API usage to identify bottlenecks
- • Use webhooks for real-time incident response
- • Export data for long-term analysis
⚠️ Watch For
- • Sudden spikes in verification failures
- • Declining trust scores across agents
- • Unusual activity patterns or timing
- • Increasing error rates on specific endpoints
- • Anomalies in agent behavior
- • Policy violation trends
🚀 Next Steps
Ready to start monitoring your AI agent activity?