What problem does it solve? Multiple blog posts targeting the same keywords compete against each other in search results, splitting rankings and traffic. This Skill identifies which posts overlap on primary keywords and search intent, then recommends whether to merge, differentiate, canonicalize, or leave them alone. ## Core Features & Use Cases - Local Keyword Extraction: Scans .md, .mdx, and .html files to extract primary keywords from titles, H1s, H2s, and meta descriptions, then clusters posts by exact, stem, semantic, and subset matches. - DataForSEO API Mode: Uses Page Intersection and Ranked Keywords endpoints (~$0.01/call) to find shared ranking keywords with search volume and position data for severity scoring. - Severity-Scored Reports: Outputs a summary table and per-cluster recommendations (MERGE, DIFFERENTIATE, CANONICAL, NO ACTION) with Critical/High/Medium/Low severity levels. - Use Case: A content team with 100+ published posts suspects traffic dilution. Run the Skill against the content directory to find that 14 posts all target "best pipe company in India", then follow the merge recommendation to consolidate them into one authoritative post. ## Quick Start Run the blog cannibalization skill on my content directory to find posts competing for the same keywords.
