Agent Analytics MCP Server

Agent Analytics provides comprehensive performance monitoring and insights for AI agents and applications. It tracks user behavior, system performance, and enables data-driven optimization through detailed analytics and experimentation capabilities.

Analytics & Reporting

• Project Management: Create and manage analytics projects with tracking snippets
• Real-time Monitoring: View live visitor counts, active sessions, events per minute, and recent activity
• Historical Analysis: Access cross-project summaries with event totals, trends, and top performers
• Custom Queries: Run flexible ad-hoc analytics with custom metrics, filters, grouping, and date ranges
• Time Series Data: Generate charts and KPIs over custom time periods with period-over-period comparisons

Behavioral Analysis

• Funnel Analysis: Track user drop-off across event sequences (page views → signups → conversions)
• Cohort Retention: Analyze user return patterns over time with session-based retention metrics
• Session Intelligence: View individual session records, duration histograms, and engagement metrics
• Traffic Patterns: Analyze day-of-week and hourly traffic grids with peak detection
• Page Analytics: Monitor entry/exit pages, bounce rates, and pages per session

Experimentation

• A/B Testing: Create, manage, and analyze experiments with detailed results
• Experiment Controls: Update experiment status (pause, resume, complete) and clean up completed tests

Traffic Quality

• Bot Detection: Filter and analyze automated traffic across projects
• Event Discovery: Explore event names and property keys from recent activity
• Data Validation: Monitor dropped events and traffic categorization

The service exposes UI resources for overview dashboards, breakdowns, heatmaps, sessions, insights, pages, funnels, and retention analysis. Event data includes rich properties for segmentation and filtering.