Environmental conditions in Laboratory Animal Resource Centers (LARCs), including ammonia, air quality, temperature, humidity, light intensity, and sound level, directly affect animal welfare and experimental reproducibility. Most LARC facilities rely on manual inspection or single-parameter systems that cannot detect threshold exceedances before physiological effects occur. This paper presents a Cyber–Physical System for multi-parameter LARC environmental monitoring that addresses three gaps: an ESP32-based sensor node integrating five sensor types in a device-agnostic, web-configurable design; a per-device, per-sensor configurable policy engine enabling institution-specific threshold compliance without code changes; and a configurable escalating-interval alert dispatcher that consolidates exceedances, lengthens the repeat interval under sustained exceedance, and resets upon new violations, reducing alert fatigue while preserving prompt notification. Evaluation comprised functional testing of 64 cases (40 executed; 62.5% coverage; 87.5% executed pass rate; 54.7% coverage-weighted pass rate), latency analysis (worst case: 10 min), alert-repeat verification (five scenarios passed), API performance tests, and a 32-user perception study suggesting perceived improvements. All five defects were resolved and retested. Monitoring and alerting functions operate as designed under the conditions evaluated; administrative functions were tested only in part. Firmware and calibration data are publicly available.