Jack's SEO MCP turns a keyword into a publish-ready SEO article inside your AI agent. It researches real keyword and SERP data, mines how people actually describe the problem, drafts, then runs quality gates: readability checks, banned AI-slop phrases, and a strict judge that sends failing drafts back for a rewrite. Your agent writes the markdown straight into your repo, where you review it like any other diff. The writing is your agent's, so token costs are tiny.
I built this for me (can you tell by the very original name of the product?)
Every time I wanted one blog / content for one of my other ventures, I did the same loop: pull keywords, check who's ranking and why, dig through Reddit for how people actually describe the problem, write the brief, draft it, then spend an hour sanding off the AI tells before it was fit to publish. Over and over.
So I turned the loop into an MCP server. Now my agent runs it: research, scoring, outline, draft, then a set of quality gates before anything counts as done. Readability checks, banned slop phrases (obvs no em dashes), and a strict judge that grades the draft and sends it back for a rewrite if it fails. What comes out the other end is a markdown file with frontmatter, ready to publish.
The part I actually cared about was the writing. Most AI SEO content reads like AI SEO content. I wanted something that reads like a person who knows the subject.
It's early days, but the venture I built it around is growing off the back of search, so I figured it was worth sharing rather than keeping to myself.
About a dollar an article, everything in. Your agent does the typing and research, matched to your code repo blog structure.
It's live at jacks-seo.com if you want a look. If you give it a go, let me know if it's useful or not.
I built this for me (can you tell by the very original name of the product?)
Every time I wanted one blog / content for one of my other ventures, I did the same loop: pull keywords, check who's ranking and why, dig through Reddit for how people actually describe the problem, write the brief, draft it, then spend an hour sanding off the AI tells before it was fit to publish. Over and over.
So I turned the loop into an MCP server. Now my agent runs it: research, scoring, outline, draft, then a set of quality gates before anything counts as done. Readability checks, banned slop phrases (obvs no em dashes), and a strict judge that grades the draft and sends it back for a rewrite if it fails. What comes out the other end is a markdown file with frontmatter, ready to publish.
The part I actually cared about was the writing. Most AI SEO content reads like AI SEO content. I wanted something that reads like a person who knows the subject.
It's early days, but the venture I built it around is growing off the back of search, so I figured it was worth sharing rather than keeping to myself.
About a dollar an article, everything in. Your agent does the typing and research, matched to your code repo blog structure.
It's live at jacks-seo.com if you want a look. If you give it a go, let me know if it's useful or not.