Skip to content
☰

alive.adapter.mock_alert@1.0.0

Mock alert classification service

Stand in for a monitoring/AI classification service before staging exists.

Boundary

Per-sample classification with assumed sensitivity, false positives, latency and errors.

Assumptions

  • Classifier behavior parameters are assumptions, never measurements.

Limitations

  • No model drift, calibration or subgroup performance.

Known failure regimes

  • Treating assumed accuracy as evidence about a real model.

Parameters

NameDefaultProvenance
sensitivity

Chance an abnormal sample alerts.

0.95 probabilitysynthetic assumption

Assumed candidate behavior.

false_positive_rate

Chance a normal sample alerts.

0.005 probabilitysynthetic assumption

Assumed candidate behavior.

latency_seconds

Response time.

exponential(mean=1) secondssynthetic assumption
error_probability

Request failure rate.

0.005 probabilitysynthetic assumption

Assumed candidate behavior.