create-viz

The create-viz skill helps users generate publication-quality data visualizations using Python. It is designed for turning query results, uploaded datasets, pasted tables, or existing analysis data into clear and effective charts for exploration, reporting, and presentations. The skill guides the visualization process from understanding the data source and analytical goal to selecting an appropriate chart type and producing polished output. It addresses common visualization challenges such as choosing the right chart for a specific data relationship, applying professional formatting, and ensuring charts communicate insights clearly.

The skill supports multiple workflows depending on the data source, including querying connected data warehouses, parsing uploaded CSV or Excel files, and working with DataFrames generated earlier in a conversation. It recommends chart types when none are specified, using best-practice mappings between analytical questions and visualization methods such as line charts for trends, scatter plots for correlations, heatmaps for relationship matrices, and Sankey diagrams for process flows. Visualization generation is performed with Python libraries including matplotlib, seaborn, and plotly, with support for both static publication-ready graphics and interactive charts with hover and zoom functionality. The skill also applies design best practices related to color usage, typography, labeling, layout, and number formatting.

This skill is useful for analysts, data scientists, business intelligence teams, researchers, and professionals preparing reports or presentations. It can be used for executive dashboards, technical analysis, stakeholder reporting, exploratory data analysis, and presentation-ready graphics. Teams that need visually consistent and professionally styled charts can use the skill to streamline visualization workflows while following established data visualization standards.