data-viz-plots

This data visualization skill empowers agents to generate professional, publication-ready scientific graphics directly within a local environment. Designed for researchers, data scientists, and engineers, this skill abstracts the complexities of Python-based plotting by providing standardized, high-performance templates for matplotlib and seaborn. By operating locally, it ensures data privacy and portability, functioning seamlessly across any LLM infrastructure, including Claude, GPT, Gemini, and open-source models like DeepSeek or Qwen.

The skill supports a comprehensive suite of plotting types, ranging from exploratory data analysis (EDA) charts to multi-panel figures for academic journals. It includes robust styling configurations to ensure consistent DPI, font sizes, and aesthetic standards required for high-impact reports, biological research (such as gene expression clustering or UMAP projections), and complex time-series analysis. Users can easily customize color palettes, legend layouts, and grid aesthetics to meet specific presentation requirements.

• Generates publication-quality figures, including scatter plots, line charts, heatmaps, violin plots, box plots, and bar plots with error bars.

• Provides multi-panel figure support for combining disparate data insights into a single cohesive layout using GridSpec.

• Includes pre-configured boilerplate for matplotlib and seaborn, ensuring consistent figure DPI and output quality for papers or presentations.

• Features flexible API for data input, supporting common structures like pandas DataFrames and numpy arrays.

• Supports various export formats, including high-resolution PNG, PDF, and interactive display options.

• Ensure all necessary libraries, including matplotlib, seaborn, pandas, and numpy, are installed in the local Python environment before executing the skill.

• Input data should be cleaned and structured in long-form or matrix formats for best results when using seaborn functions.

• For multi-panel visualizations, define the GridSpec layout early to manage figure space and axes handles effectively.

• Constantly verify file paths for saved plots, as the skill will output files directly to the working directory.

• While the tool is universal, complex layouts may require manual adjustment of margins and whitespace to prevent text overlapping in the final exported figure.