ROS2 Diagnostics

Create a diagnostic_updater::Updater instance with your node, call setHardwareID, then register callbacks with updater.add(name, callback). Each callback fills a DiagnosticStatusWrapper with summary level and key-value entries.

Use diagnostic_updater::HeaderlessTopicDiagnostic with FrequencyStatusParam specifying min and max frequency, tolerance, and window size. Call tick() each time a message arrives on the monitored topic.

Yes, diagnostic_updater supports both rclpy and rclcpp. The Skill shows a C++ DiagnosticsManager implementation and references the equivalent Python pattern, both publishing standard diagnostic_msgs.

Use the standard levels OK, WARN, ERROR, and STALE from diagnostic_msgs. Map domain health states to these levels in your status callback so tools like rqt_robot_monitor interpret them consistently.

Setting a unique hardware ID via updater.setHardwareID groups all diagnostic entries from one device together in monitoring tools. Without it, statuses from multiple components become hard to attribute to physical hardware.