Local Interactive Simulation
MonitorX Alert Evaluation
Explore how latency, errors, and drift signals can be evaluated against explicit model-health policy—without presenting generated telemetry as production data.
View GitHubTelemetry Envelope
Browser-only simulationAlert Evaluation
Adjust telemetry values to inspect MonitorX's illustrative alert policy.
No Firestore database, monitoring backend, or external API is called.
Validation Boundary
Broad component coverage, incomplete suite health.
MonitorX independently implements instrumentation, collection, alerting, storage, API, and dashboard layers. It is locally tested as a reference implementation, not operated against sustained production telemetry.
Types, metrics collection, SDK behavior, resilience helpers, and alert channels pass locally.
The API test module imports a stale storage symbol, preventing a clean full-suite result.
Repository Evidence
Python client, context manager, decorators, buffering, retry, and circuit-breaker behavior.
Inference, drift, resource, threshold, and summary metric models.
Email, Slack, and webhook channels with rate limiting and resolution state.
FastAPI endpoints, InfluxDB adapter, Streamlit dashboard, and export paths.