visualizationlisted

# visualization The visualization-selection umbrella for robium. "I can't see what the robot is doing" is one of the fastest ways a debugging session stalls, and picking the wrong tool for the context (a local-display tool on a headless server, a ROS-native tool for a non-ROS ML pipeline) wastes more time than the debugging itself. This skill decides which of rviz2, Foxglove, or Rerun fits a given context, and states the cross-cutting practice of *what* to visualize at each build stage. It does not teach any one tool's UI, panels, or API — that is each tool's own skill. ## When to use this skill - A viz tool hasn't been chosen yet for the current context (local desktop, headless remote server, or an ML/data pipeline outside ROS). - The trigger phrases in the description: 'visualize', 'see what the robot sees', 'debug visually', 'plot the trajectory', 'dashboard for the robot'. - Deciding whether to look at data live or record it for later inspection, before wiring up either path. - Cross-references — go straight to the tool skill, skipping this one, once the tool is already chosen and the question is tool-specific: - RViz2 panel configuration, displays, or plugins on a local Linux desktop → `rviz2`. - Foxglove layouts, the Foxglove Bridge, or MCAP recording/playback → `foxglove`. - Rerun logging API, blueprints, or timeline views for ML/perception data → `rerun`. - Whether headless/remote is the deployment target at all (as opposed to which viz