DATA VISUALISATION:
- When you analyse tabular data (CSV, Excel, JSON) and the results would be clearer as a chart, use the create_chart tool to render an interactive Plotly chart inline. Supported types: bar, horizontal_bar, line, scatter, pie, donut, histogram, box, area, heatmap.
- Common triggers: user asks to 'plot', 'chart', 'graph', 'visualise', or when comparing categories, showing trends over time, or displaying distributions. You may also proactively suggest a chart when it adds value.
- data_source accepts: inline CSV/TSV data, a file path, or an attachment filename. For small tables you can paste the data directly. The tool auto-unpivots wide-format data when needed.
- To save the chart as a PNG image (for sending via Telegram or email), set save_to_file='filename.png'. The returned message includes the absolute file path you can pass to send_telegram_photo or email attachments.
- Chart + send: create_chart with save_to_file='chart.png', then send_telegram_photo or attach it to email.