Automatically extracted from the page. Use {{\"Your prompt here\"}} to ask the AI Interpreter to extract or summarize info. Extract content using CSS selectors. S
The Variables skill enables marketers to dynamically extract and insert key pieces of web content into workflows by referencing preset, meta, schema.org, selector, and AI prompt variables. This allows automated capture of article titles, URLs, author names, publication dates, main content, and other structured data directly from pages during data collection or content enrichment. It supports both straightforward data points like `{{title}}` and complex extractions using CSS selectors or AI summarization prompts, making content-driven campaigns and reporting more precise and scalable.
By leveraging these variables, practitioners can streamline content personalization, audit SEO metadata, and populate templates without manual copy-pasting. This reduces friction in multichannel campaigns where consistent, accurate page data is critical for performance tracking and audience targeting.
This skill is designed for performance marketers managing content-heavy SEO and PPC campaigns who need clean, structured data from landing pages to optimize ad copy or meta tags. Growth leads running content audits or scaling content refreshes will find it useful for automating data extraction and reducing manual review cycles. Agency strategists coordinating cross-channel messaging can use Variables to ensure brand consistency by pulling real-time page info into campaign briefs or asset libraries.
Practitioners typically start by identifying which variables are critical for their campaign—like `{{title}}` for headline testing or `{{description}}` for meta description audits. Next, they configure selectors or AI prompt variables to extract more nuanced content segments, such as highlighted text or product details, tailoring the output to channel requirements. During campaign execution, these variables automatically populate templates or feed analytics tools, enabling real-time adjustments based on extracted content signals. Finally, marketers review extraction accuracy and update selectors or prompt logic to handle edge cases or new page layouts.
How do I extract custom content not covered by preset variables? Use selector variables with CSS selectors targeting specific page elements. Can I get structured data like author or publication date? Yes, schema.org and meta variables capture this when present on the page. What if a variable is empty or missing? You can chain fallbacks or provide default values to maintain data consistency.
Attach the Variables skill to a Metaflow agent task that involves content scraping or enrichment to automatically pull structured data from target pages using predefined or custom variables. Expect outputs like article titles, URLs, or highlighted text to be injected into your workflow data, reducing manual extraction effort. Use this skill as a foundation for content-driven decision-making and campaign automation within Metaflow, and explore further integrations with other content and AI analysis capabilities.
For broader context, see our roundup of claude marketing skills, and read common Claude Code content mistakes for related setup guidance.