Available Data Science Plugins
These plugins extend Claude Code capabilities for data science workflows. Enable when needed.
Code Intelligence
| Plugin | Description | Enable Command |
|---|---|---|
serena | Semantic code analysis, refactoring, symbol navigation | claude --enable-plugin serena@claude-plugins-official |
pyright-lsp | Python type checking and diagnostics | claude --enable-plugin pyright-lsp@claude-plugins-official |
Documentation
| Plugin | Description | Enable Command |
|---|---|---|
context7 | Up-to-date library docs (pandas, numpy, sklearn, etc.) | claude --enable-plugin context7@claude-plugins-official |
Web & Automation
| Plugin | Description | Enable Command |
|---|---|---|
playwright | Web scraping, browser automation, screenshots | claude --enable-plugin playwright@claude-plugins-official |
Workflow
| Plugin | Description | Enable Command |
|---|---|---|
ralph-loop | Self-referential iteration loops | Already enabled |
hookify | Create custom hooks from conversation patterns | Already enabled |
Data Access Skills (Built-in)
These are skills, not plugins - already available via /ds:
| Skill | Description |
|---|---|
/wrds | WRDS (Wharton Research Data Services) queries |
/lseg-data | LSEG Data Library (formerly Refinitiv) |
/gemini-batch | Gemini Batch API for large-scale LLM processing |
/jupytext | Jupyter notebooks as text files |
/marimo | Marimo reactive Python notebooks |
File Format Skills (Bundled)
Office document and PDF skills from Anthropic's official skills repo (bundled via submodule):
| Skill | Use For |
|---|---|
/xlsx | Spreadsheets, formulas, CSV→Excel conversion |
/pdf | PDF extraction, creation, form filling |
/pptx | Presentation creation and editing |
/docx | Word docs, tracked changes, reports |
These skills are available via the shared plugin - no separate installation needed.
When to Enable
- •serena: Understanding complex analysis codebases, refactoring pipelines
- •context7: Access current docs for pandas, scikit-learn, statsmodels, and other libraries
- •playwright: Scrape web data sources, automate data collection
- •pyright-lsp: Type check data pipelines
Usage
Enable a plugin for the current session:
bash
claude --enable-plugin <plugin-name> # Enable a plugin by name and source
Enable a plugin for a project by adding to .claude/settings.json:
json
{
"enabledPlugins": {
"serena@claude-plugins-official": true
}
}