Data Workflow

For any data question, metric lookup, or analytical task, delegate to the data-triage skill which orchestrates the full data pipeline:

1. Load the data-triage skill
2. Follow its workflow (data-model-explorer → data-analyst)

For file-based data processing (CSV, JSON, Excel, Parquet):

• Install needed packages (pandas, matplotlib, seaborn, scipy)
• Write and iterate on analysis code
• Save outputs to workspace files (charts to /tmp/data_*.png, results to /tmp/)