
Put my 4 years of SEO experience into a claude skill so you don't have to figure it out yourself
I condensed my SEO experience into a Claude Code skill that actually does keyword research and writes articles the right way & open sourced it
Most AI writing tools I came across gave really shallow output. They go straight from keyword to article with no research in between. No competitor analysis, no understanding of what's already ranking, no reason why someone would read your article over the 10 that already exist. The content always feels hollow because there's nothing behind it.
I've been doing SEO long enough to know the research layer is everything. The writing is the easy part. Finding the right keyword, understanding the competitive gap, knowing what angle to take. that's what actually makes content rank
So I put my exact workflow into a Claude Code skill. Three slash commands.
/blog-onboard - scrapes your site, extracts your business profile, domain rating, ICP, brand voice, and finds your direct competitors automatically
/blog-topics - pulls competitor keywords, generates seed phrases based on your ICP pain points, expands them, classifies by funnel stage, clusters into topic groups, scores every keyword by opportunity, picks your first week of articles with titles already generated
/blog-write - scrapes the top ranking articles for your keyword, pulls recent news and expert opinions via Tavily, extracts YouTube insights, does SERP gap analysis to find what the current results are missing, generates a full outline, then writes the article in one shot against that outline
Everything local, no subscription, just your API keys
github.com/maun11/claude-blog-engine
It works but there's room to improve. If you've built anything in this space or have opinions on the research layer specifically I'd like to hear it. PRs welcome.