aegislm / observability /__init__.py
ACA050's picture
Upload 57 files
f2c6053 verified
Raw
History Blame Contribute Delete
2.52 kB
"""
AegisLM Observability Module
Provides comprehensive observability features:
- Metrics collection and export (Prometheus-style)
- Health checks (liveness and readiness)
- Alert routing and escalation
- Production readiness auditing
- Dashboard documentation
This module enables production-grade monitoring and operational reliability.
"""
from observability.metrics import (
MetricsRegistry,
get_metrics_registry,
MetricType,
AlertLevel,
Counter,
Gauge,
Histogram,
)
from observability.exporters import (
PrometheusExporter,
get_prometheus_exporter,
export_prometheus_metrics,
)
from observability.alert_router import (
AlertRouter,
AlertHandler,
LogHandler,
AlertEscalationTracker,
Alert,
AlertLevel,
AlertCategory,
get_alert_router,
send_drift_alert,
send_security_alert,
send_robustness_alert,
)
from observability.health_checks import (
LivenessResponse,
ReadinessResponse,
liveness_check,
readiness_check,
check_database_connection,
check_worker_pool,
check_model_loader,
check_policy_config,
check_dataset_registry,
check_secret_manager,
get_uptime_seconds,
)
from observability.readiness_audit import (
ReadinessAuditor,
ReadinessCheck,
ReadinessReport,
get_readiness_auditor,
run_readiness_audit,
SLO_CONFIG,
)
__version__ = "0.1.0"
__all__ = [
# Metrics
"MetricsRegistry",
"get_metrics_registry",
"MetricType",
"AlertLevel",
"Counter",
"Gauge",
"Histogram",
# Exporters
"PrometheusExporter",
"get_prometheus_exporter",
"export_prometheus_metrics",
# Alert Router
"AlertRouter",
"AlertHandler",
"LogHandler",
"AlertEscalationTracker",
"Alert",
"AlertLevel",
"AlertCategory",
"get_alert_router",
"send_drift_alert",
"send_security_alert",
"send_robustness_alert",
# Health Checks
"LivenessResponse",
"ReadinessResponse",
"liveness_check",
"readiness_check",
"check_database_connection",
"check_worker_pool",
"check_model_loader",
"check_policy_config",
"check_dataset_registry",
"check_secret_manager",
"get_uptime_seconds",
# Readiness Audit
"ReadinessAuditor",
"ReadinessCheck",
"ReadinessReport",
"get_readiness_auditor",
"run_readiness_audit",
"SLO_CONFIG",
]