Brand Identity Extractor
This skill uses the Firecrawl API to gather brand identity data from a provided URL.
Core Workflow
1. Data Gathering
- •Use
firecrawl_scrapewithformats: ["branding", "markdown"]to get initial design tokens and full page content. - •Use
firecrawl_extractwith the schema in branding_schema.md to get structured brand data. - •Search for "logo", "brand", "style", "icons" within the scraped markdown to find additional asset URLs.
2. Manual Analysis & Refinement
- •Inspect the
brandingoutput fromfirecrawl_scrapefor specific CSS values (HEX codes, font-family names). - •Identify "Hero" or "Header" images in the markdown and verify their relevance to the brand.
- •Look for component-specific styling (e.g., button border-radius, shadows, hover animations).
3. Reporting & Artifact Creation
- •IMPORTANT: After extraction, use the structured data to generate a new brand guidelines markdown file.
- •Use the brand_guide_template.md as a starting point.
- •The resulting file should be saved in the project's
directives/ordocs/folder for future reference (e.g.,directives/brand-guidelines-<domain>.md). - •Ensure all image URLs are validated and clickable.
Best Practices
- •Verify Images: Always check if the images pulled are high-resolution and relevant. Favicons and small icons should be categorized separately from hero images.
- •Color Accuracy: If the scraped HEX codes seem off (e.g., pure black/white), double-check the
brandingformat output for more nuanced color tokens. - •Typography: Check if fonts are standard Google Fonts or custom web fonts. This helps in implementation.