Python module

The chemiscope package provides utilities to prepare JSON input files, and to visualize them as an interactive notebook widget. The package supports multiple structure formats including ase.Atoms, stk.BuildingBlock, and MDAnalysis.AtomGroup objects.

• Command line interface
  • chemiscope-input - CLI interface
• chemiscope functions reference
  • Input files
  • Data management
• Notebook widget
  • Supported environments
  • Try it on Google Colab
  • Creating a chemiscope widget
  • Widget properties and methods
  • Headless widget
  • Dataset exploration
• Chemiscope in streamlit
  • Installation
  • Basic Usage
  • API reference
• Chemiscope in sphinx
  • Configuration
  • Chemiscope directive
  • External structure files
  • Viewing a documentation locally
• Use with sphinx-gallery
  • Using widgets in the examples
  • Setting up sphinx-gallery

Tip

If you work with an AI coding assistant, the chemiscope repository ships a reusable reference skill under docs/skills/. It teaches the assistant the chemiscope Python API, the input format, the visualization settings and the custom shapes, so it can help you write visualizations for your own data, including forces, atom-centered properties and custom shapes. The files are plain Markdown that any agent can read; point your tool at docs/skills/ (an AGENTS.md pointer at the repository root does this automatically for tools that support it). For Claude Code the skill is also exposed via .claude/skills/chemiscope-python; copy that folder into your ~/.claude/skills/ (or a project’s .claude/skills/) to use it in your own projects.