Wikipedia Article Extractor

Overview

The Wikipedia Article Extractor template is a robust browser automation workflow designed to parse and extract structured data from public Wikipedia articles. Operating in a fast, stateless execution mode, this agent navigates directly to a target Wikipedia slug and processes the DOM content using advanced JavaScript extraction routines. No login is required, and the workflow is completely read-only.

What it does

This template automates the retrieval and normalization of Wikipedia encyclopedia entries. Key capabilities include:

• Structured Page Metadata: Extracts the article title, language code, short description, page URL, and last modified timestamp.
• Markdown Conversion: Cleans the primary parser output by stripping out editing sections, navboxes, footnotes, and metadata boxes, and converts the remaining document body into clean Markdown.
• Lead Paragraphs: Collects the initial introductory paragraphs to provide an executive summary of the subject.
• Infobox Extraction: Parses key-value pairs from sidebar infoboxes.
• Section Mapping: Builds an index of all section and subsection headers with their heading levels.
• Asset Gathering: Identifies relevant images with captions and alt tags while filtering out miniature icons.
• References & Links: Collects unique categories, bibliographical references, and internal Wikipedia links.

Inputs

• articleTitle (string): The Wikipedia article slug or title suffix used to build the target URL (https://en.wikipedia.org/wiki/{$articleTitle}). Defaults to Linus_Torvalds.

Output

Returns a comprehensive JSON payload containing:

• title: The primary heading of the article.
• pageUrl: The absolute URL of the extracted page.
• language: The document language attribute.
• description: The short summary description.
• lead: Combined lead paragraphs.
• leadParagraphs: An array of individual lead paragraphs.
• bodyMarkdown: Cleaned Markdown version of the article body.
• infobox: Key-value pairs extracted from the article's sidebar infobox.
• sections: Array of sections with titles and hierarchy levels.
• sectionCount: Total number of sections found.
• images: Array of extracted images with URLs and captions.
• categories: Filtered list of article categories.
• references: Extracted list of citation text strings.
• referenceCount: Total number of references.
• internalLinks: Unique internal Wikipedia links found within the text.
• internalLinkCount: Total count of internal links.
• lastModified: Footer text indicating the last modification date.
• extractedAt: ISO timestamp of when the extraction took place.

How to use

1. Open the template in the Figranium task canvas.
2. Locate the articleTitle variable and replace the default value with the slug of the Wikipedia article you wish to scrape.
3. Run the task in Scrape mode.
4. Review or export the resulting structured JSON payload.

Notes

• This workflow is stateless, executes securely without recording sessions, and relies entirely on public Wikipedia pages.
• Boilerplate elements such as edit links, navigation tables, and references lists are automatically pruned from the main Markdown body for clean readability.