# Prometheus Alert Rules for Session-Buddy Phase 4
#
# Place this file in your Prometheus configuration directory and reference it:
#
# prometheus.yml:
# rule_files:
# - "session-buddy-alerts.yaml"
#
# Reload Prometheus: curl -X POST http://localhost:9090/-/reload
groups:
- name: session_buddy_skill_anomalies
interval: 30s
rules:
- alert: SessionBuddyHighAnomalyRate
expr: sum(increase(anomalies_detected_total[5m])) > 5
for: 2m
labels:
severity: warning
component: analytics
annotations:
summary: "High anomaly detection rate ({{ $value }} anomalies in 5m)"
description: "More than 5 anomalies detected in the last 5 minutes. Check skill performance."
runbook_url: "https://docs.session-buddy.dev/runbooks/anomalies"
- alert: SessionBuddySkillPerformanceDrop
expr: skill_completion_rate{skill_name!="test"} < 0.7
for: 5m
labels:
severity: warning
component: analytics
annotations:
summary: "Skill completion rate below 70%: {{ $labels.skill_name }}"
description: "{{ $labels.skill_name }} completion rate is {{ $value | humanizePercentage }} (baseline: 85%+)"
runbook_url: "https://docs.session-buddy.dev/runbooks/performance"
- alert: SessionBuddySkillCriticalFailure
expr: skill_completion_rate{skill_name!="test"} < 0.5
for: 2m
labels:
severity: critical
component: analytics
annotations:
summary: "CRITICAL: Skill failure rate above 50%: {{ $labels.skill_name }}"
description: "{{ $labels.skill_name }} is failing more than half the time. Immediate investigation required."
runbook_url: "https://docs.session-buddy.dev/runbooks/critical-failures"
- name: session_buddy_system_health
interval: 15s
rules:
- alert: SessionBuddyWebSocketServerDown
expr: up{job="session-buddy-websocket"} == 0
for: 1m
labels:
severity: critical
component: infrastructure
annotations:
summary: "WebSocket server is down"
description: "Session-Buddy WebSocket server has been down for more than 1 minute."
runbook_url: "https://docs.session-buddy.dev/runbooks/websocket"
- alert: SessionBuddyDatabasePoolExhausted
expr: (sqlite_connections_active / sqlite_connections_max) > 0.9
for: 2m
labels:
severity: warning
component: database
annotations:
summary: "Database connection pool usage above 90%"
description: "{{ $value | humanizePercentage }} of database connections in use."
runbook_url: "https://docs.session-buddy.dev/runbooks/database-pool"
- alert: SessionBuddyDatabasePoolCritical
expr: (sqlite_connections_active / sqlite_connections_max) > 0.95
for: 30s
labels:
severity: critical
component: database
annotations:
summary: "CRITICAL: Database connection pool exhausted"
description: "Only {{ 100 - mul($value, 100) }}% capacity remaining."
runbook_url: "https://docs.session-buddy.dev/runbooks/database-pool"
- alert: SessionBuddyActiveSessionsZero
expr: active_sessions_total == 0
for: 10m
labels:
severity: info
component: monitoring
annotations:
summary: "No active sessions for 10 minutes"
description: "System may be idle or tracking is broken."
runbook_url: "https://docs.session-buddy.dev/runbooks/sessions"
- name: session_buddy_performance
interval: 1m
rules:
- alert: SessionBuddySlowSkillExecution
expr: histogram_quantile(0.95, rate(skill_duration_seconds_bucket[5m])) > 30
for: 5m
labels:
severity: warning
component: performance
annotations:
summary: "95th percentile execution time above 30s: {{ $labels.skill_name }}"
description: "Skill {{ $labels.skill_name }} p95 duration is {{ $value }}s."
runbook_url: "https://docs.session-buddy.dev/runbooks/slow-skills"
- alert: SessionBuddyVerySlowSkillExecution
expr: histogram_quantile(0.95, rate(skill_duration_seconds_bucket[5m])) > 60
for: 2m
labels:
severity: critical
component: performance
annotations:
summary: "CRITICAL: 95th percentile execution time above 60s"
description: "Skill {{ $labels.skill_name }} p95 duration is {{ $value }}s. Performance severely degraded."
runbook_url: "https://docs.session-buddy.dev/runbooks/slow-skills"
- alert: SessionBuddyHighInvocationRate
expr: rate(skill_invocations_total[1m]) > 100
for: 5m
labels:
severity: warning
component: performance
annotations:
summary: "High skill invocation rate: {{ $labels.skill_name }}"
description: "{{ $labels.skill_name }} invoked {{ $value }}/s for 5 minutes."
runbook_url: "https://docs.session-buddy.dev/runbooks/high-load"
- name: session_buddy_data_quality
interval: 5m
rules:
- alert: SessionBuddyStaleMetricsCache
expr: time() - skill_metrics_cache_last_update > 300
for: 5m
labels:
severity: warning
component: data_quality
annotations:
summary: "Metrics cache not updated for 5 minutes"
description: "Real-time metrics cache may be stale."
runbook_url: "https://docs.session-buddy.dev/runbooks/stale-cache"
- alert: SessionBuddyMissingSkillMetrics
expr: count(skill_completion_rate) < 10
for: 10m
labels:
severity: info
component: data_quality
annotations:
summary: "Less than 10 skills have metrics"
description: "Expected more skills to be tracked. Check data collection."
runbook_url: "https://docs.session-buddy.dev/runbooks/missing-metrics"
- name: session_buddy_integration
interval: 1m
rules:
- alert: SessionBuddyCrackerjackIntegrationFailing
expr: rate(skill_invocations_total{workflow_phase="execution", skill_name=~".*-fail"}[5m]) > 0.1
for: 5m
labels:
severity: warning
component: integration
annotations:
summary: "Crackerjack integration failures detected"
description: "{{ $value | humanize }} failures/sec in Crackerjack workflow."
runbook_url: "https://docs.session-buddy.dev/runbooks/crackerjack"
- alert: SessionBuddyCICDPipelineFailure
expr: rate(skill_invocations_total{workflow_phase="execution", completed="false"}[5m]) > 0.2
for: 3m
labels:
severity: critical
component: integration
annotations:
summary: "CI/CD pipeline failure rate above 20%"
description: "High failure rate in CI/CD skill invocations."
runbook_url: "https://docs.session-buddy.dev/runbooks/cicd"
- name: session_buddy_predictions
interval: 5m
rules:
- alert: SessionBuddyPredictionAccuracyDrop
expr: (skill_prediction_correct / skill_prediction_total) < 0.7
for: 10m
labels:
severity: warning
component: ml
annotations:
summary: "Predictive model accuracy below 70%"
description: "Model accuracy is {{ $value | humanizePercentage }}. Retraining may be needed."
runbook_url: "https://docs.session-buddy.dev/runbooks/predictions"
- alert: SessionBuddyABTestSampleSizeInsufficient
expr: ab_test_samples_total < ab_test_min_sample_size
for: 1h
labels:
severity: info
component: analytics
annotations:
summary: "A/B test sample size insufficient"
description: "Test {{ $labels.test_name }} has only {{ $value }} samples ({{ $labels.min_required }} required)."
runbook_url: "https://docs.session-buddy.dev/runbooks/ab-testing"
