Unaware fatigue causes inefficiencies and unsafe behaviors. Therefore, an easily-used technique to monitor labor intensity is critical. In this work, the humidity-sensitive material Cs3Cu2I5 was used to monitor breath and grade labor intensity without contact. The Cs3Cu2I5 showed reversible conversion to the CsCu2I3 phase by being exposed to water. EMI-TFSI was added to improve the material's conductivity and electrochemical stability, helping to passivate surface defects and adjust the response sensitivity. Further mechanism explanation was explored by density functional theory calculation and electrochemical impedance spectroscopy. Switching from 40 % RH to 90 % RH, the sensor resistance changed by 161.3 megohms with a relative change ratio of 0.967. Besides, the sensor has a short response/recovery time (e.g., 2.0/0.5 s switching from 40 % RH to 60 % RH), with good repeatability and stability in high humidity. Then, the developed high-performance lead-free sensor was integrated into a smart mask with a neural network algorithm to monitor and forecast breath, and grade labor intensity. The model was built with 240 sample data and 19 constructed features. The accuracy of identifying three intensity levels in external validation was 0.9. Based on the sensing data and the grading model, a real-time labor intensity monitoring and visualization platform was constructed.