Documentation

📊Activity Monitoring

Monitor your AI operations with streaming consensus, performance analytics, and threshold tracking

← Back to Documentation

📊 Activity Monitoring

Monitor your AI operations with comprehensive activity tracking, streaming consensus pipeline, and performance analytics.

What Makes Our Monitoring Unique?

Unlike basic usage tracking, our activity monitoring provides visibility into every aspect of your AI operations - from streaming consensus stages to cost accumulation and performance analysis.

🔄 Streaming View

Watch 4-stage consensus pipeline progress

📈 Analytics Insights

Usage pattern analysis and thresholds

🚨 Threshold Monitoring

Track usage thresholds and limits

🔄 Consensus Progress Streaming

4-Stage Pipeline Visualization

Monitor your AI consensus requests as they process through all four stages with progress tracking, token rates, and cost accumulation.

# Watch live consensus processing
hive consensus "Explain quantum computing"
🟡 Generator Stage... (25%) - 45 tokens/sec - $0.003
🟠 Refiner Stage... (50%) - 38 tokens/sec - $0.007
🔴 Validator Stage... (75%) - 42 tokens/sec - $0.012
🟢 Curator Stage... (100%) ✓ Complete - $0.018

Metrics Tracked

  • Progress percentage
  • Token generation rate (tokens/second)
  • Response time per stage
  • Cost accumulation
  • Model performance scores

Visual Indicators

  • Color-coded stage progression
  • Animated progress bars
  • Cost tracking
  • Performance trend indicators
  • Success/failure status icons

🏥 Health Monitoring

System Health Checks

Continuous health monitoring every 5 seconds across all system components

# Check system health
hive health --detailed
✅ OpenRouter API: Healthy (1.2s avg)
✅ Database: Healthy (15ms avg)
⚠️ Model GPT-4: Degraded (5.2s avg)
✅ Memory Usage: 67% (Normal)

Component Analysis

Detailed health analysis with automated recommendations

# Monitor specific components
hive health --component models
📊 Success Rate: 98.7% (Target: 95%)
⚡ Avg Response: 2.8s (Target: 3s)
🔄 Uptime: 99.9% (24h)
💡 Recommendation: All healthy

📈 Advanced Performance Analytics

Enterprise Performance Intelligence

Sophisticated performance monitoring with bottleneck detection, optimization recommendations, and comparative analysis across time periods. Features ML-style trend analysis and automated performance optimization suggestions.

2.8s
Avg Response Time
↓ 15% this week
98.7%
Success Rate
↑ 2.1% this week
$12.45
Monthly Spend
49.8% of budget
# Advanced performance analysis with recommendations
hive perf analyze
🔍 Bottleneck Detection: Generator stage (avg 2.1s)
⚡ Optimization Opportunity: 30% reduction possible
💡 Recommendation: Consider faster models for generation
📊 Trend Analysis: Performance improving (+15%)
🎯 Cost Efficiency: $0.018 per consensus (optimal)

🚨 Enterprise-Grade Alert System

Advanced Alerting with Anomaly Detection

Enterprise-grade intelligent monitoring with machine learning-style baseline detection, smart threshold management, and multi-channel notifications.

🤖 Enterprise Intelligence Features

  • • Anomaly Detection: ML-style baseline learning
  • • Smart Throttling: Prevents alert spam
  • • Circuit Breaker Integration: Automatic failover detection
  • • Webhook Integration: External monitoring tools
  • • Slack Notifications: Team alerting ready
  • • Health Endpoints: HTTP monitoring for DevOps

High Response Time

Alert when response time > 5 seconds

Last triggered: Never

API Error Rate

Alert when OpenRouter errors > 10%

Last triggered: 3 days ago

Cost Spike

Alert when spending spikes 3x normal rate

Last triggered: Never

Quality Drop

Alert when model quality drops > 10%

Last triggered: Never

Memory Critical

Alert when memory usage > 90%

Last triggered: Never

Enterprise Integrations

  • 🔍 Prometheus export
  • 📊 CSV/JSON analytics export
  • 🔗 Webhook integration
  • 💬 Slack notifications
  • 🌐 HTTP health endpoints

⌨️ Monitoring Commands

Health & Status

# System health checks
hive health
hive status
# Detailed component analysis
hive health --detailed
hive health --component models

Advanced Performance Analysis

# Comprehensive performance monitoring
hive perf analyze
hive perf bottlenecks
# Time-period comparisons
hive perf compare 7d 1d
hive perf export prometheus

Alert Management

# Alert configuration
hive alerts
hive alerts configure
# Alert history and status
hive alerts list
hive alerts history

Analytics Dashboard

# Interactive dashboards
hive dashboard live
hive dashboard executive
# Live monitoring with auto-refresh
hive dashboard watch
hive dashboard operational

💻 IDE Integration & MCP Tools

Seamless IDE Monitoring

Hive AI monitoring works automatically in your favorite IDE through MCP (Model Context Protocol). No setup required - just ask questions in natural language.

🤖 How It Works

1. Automatic Setup

MCP tools configure automatically when Hive AI is installed

2. Natural Language

Ask Claude/Cursor/Windsurf about your system in plain English

3. Live Data

Get real-time monitoring data directly in your IDE

💬 Natural Language Examples

"How is my AI system performing?"
→ Gets health status, response times, success rates
"Show me my AI usage trends this week"
→ Displays analytics, costs, model performance
"Are there any performance issues?"
→ Analyzes bottlenecks, alerts, optimization tips
"What’s my current AI spending?"
→ Shows costs, budget status, projections

🎯 Supported IDEs

🤖
Claude Code
⚙️
Cursor
🌊
Windsurf
💻
VS Code

🚀 Enterprise-Grade AI Monitoring

Deploy sophisticated monitoring with ML-style anomaly detection, automated performance optimization, circuit breaker integration, and enterprise export capabilities. Complete visibility into your AI operations with intelligent alerting and real-time consensus streaming.

💻 Terminal Commands

# Direct monitoring commands
hive health
hive dashboard live
hive consensus "Test question"

🤖 IDE Integration

Just ask in natural language:
"Show my system health"
"How are my AI models performing?"
"What’s my usage this week?"
Start free trial