ReHealth AI HealthAgent is an open-source multi-agent health intervention simulation project published by ReHealth AI on GitHub. It organizes emotion, behavioral adherence, physiological state and intervention strategies as collaborating agents, providing researchers and developers with an inspectable starting point for repeatable health-behavior simulations.
View source and documentation on GitHub ↗
What the project explores
Health intervention outcomes may depend not only on the intervention itself, but also on factors such as emotion, fatigue, motivation and sustained participation. HealthAgent places these factors in a multi-agent simulation loop so developers can experiment with behavioral assumptions and intervention strategies, then inspect how simulated trajectories change.
The repository describes emotion, adherence, physiology and intervention agents coordinated by an orchestrator. Its simulation loop follows emotion, adherence, physiology, intervention and record. Example scenarios include sleep and cardiovascular simulations.
For research and prototyping
- Explore hypotheses about behavioral adherence and health interventions;
- Prototype intervention workflows for digital health services;
- Study how multi-agent systems can be organized for health-behavior simulation;
- Inspect, reproduce or extend simulation methods using public code.
How it relates to ReHealth Core
HealthAgent is an open-source research and development project focused on simulation. ReHealth Core serves institutional programmes with infrastructure for risk identification, intervention records, outcome attribution and evidence management. They address different questions along the prevention pathway. Simulation outputs should not be treated as clinical research findings, real-world outcomes or settlement evidence.
HealthAgent is currently a simulation framework for research and development, not a medical device or clinical decision system. Simulated outputs do not represent real-patient outcomes and are not diagnosis, treatment advice or evidence of clinical effectiveness. Any research or application requires independent methodological, ethics, safety and compliance review by the user.
Repository details
- Source repository: github.com/csong8904-spec/ReHealthAI-HealthAgent
- License: Apache License 2.0 (refer to the repository's LICENSE file)
- Primary language: Python
- Contributions: review the repository documentation, open an issue or contribute through a pull request.