AI agent for SEO research and page fixes
SEO research with an agent covers a site and its rivals in one loop. Your agent runs managed scrapers for public Ahrefs, Semrush, Moz, Similarweb and Ubersuggest data under a credit cap, reads Google's live results, finds the weakest pages, and drafts fixes for your Webflow or Contentful site that you review first.
The job. Research a site and two rivals with managed SEO scrapers and live Google results, then draft fixes for the weakest pages.
Paste this into Claude Code, Codex or Cursor
set up https://corespeed.io/SKILL.md (read it raw, e.g. with curl), then research example.com and two rivals for US search with the SEO scrapers and Google results, keep each paid call under 20 credits, and draft fixes for the five weakest pages; show me each change before you make it.One prompt.Three steps.
Set up first. Nothing for the research: the scrapers and Google results need no vendor account. To apply fixes, your agent asks you to connect Webflow or Contentful.
Name your site and rivals
Paste the prompt into Claude Code, Codex or Cursor. Set the credit cap for each paid call.
Let it research
Scrapers return public estimates; Google search shows the real results page. Each paid call stays under the cap.
Review each fix
It shows the current title and description beside the new ones before it writes anything.
The tools it calls
web__list_scrapers· Find SEO scrapers by site and recordweb__get_scraper· Check a scraper's input and chargesweb__run_scraper· Run it under your credit capweb__google_search· See Google's results and AI Overviewweb__scrape· Read your page and the pages that outrank itwebflow__data_pages_tool· Read and update page titles and descriptionscontentful__patch_entry· Or edit the entry in Contentfulmemory__remember· Save the plan for next month's run
Every tool and its price: Web search and read, Webflow, Contentful