Modifies Matplotlib plots through visual browser-based annotations and AI-powered code updates.
The Visual Plot Annotator is a specialized Claude Code skill designed to bridge the gap between visual data representation and Python source code. It allows users to load Matplotlib-generated images into a browser-based interface where they can draw, label, and group annotations over specific plot elements. Claude then interprets these visual cues—such as a box over a legend or a note on an axis—to identify the corresponding Matplotlib objects in the script. The skill automatically suggests and applies precise code modifications to update styles, labels, or data parameters, providing a streamlined workflow for refining data visualizations without manual coordinate hunting.
주요 기능
01Intelligent mapping of visual bounding boxes to Matplotlib elements like axes, legends, and titles
02Interactive browser-based annotation interface powered by Fabric.js
031 GitHub stars
04Annotation grouping for multi-step plot modifications in a single pass
05Automatic discovery of Python source scripts linked to generated plots
06Safe code application with mandatory preview and confirmation before writing to disk
사용 사례
01Quickly updating plot labels, titles, and legends via direct visual reference
02Iterative visualization debugging by annotating errors directly on the output image
03Fine-tuning aesthetic parameters like font sizes and rotations without searching through code
