Cursor's agent is good at writing and bad at knowing what to write about, because it has no data. Give it live keyword, SERP, backlink and AI-visibility tools and that changes.
Live now: the MCP server, the free AI visibility audit and the keyword tools. Cursor reads two configs: ~/.cursor/mcp.json applies everywhere, and a .cursor/mcp.json in the repo applies to that project only, which is the cleaner option when a client's keys should not follow you around.
Cursor reads two configs: ~/.cursor/mcp.json applies everywhere, and a .cursor/mcp.json in the repo applies to that project only, which is the cleaner option when a client's keys should not follow you around.
Paste this in and it reads our skill, then writes its own config in the format Cursor expects.
Read https://omnifamous.com/mcp/skill.md and set up the omnifamous SEO MCP server for this project, then use it to audit my site's search and AI visibility.
~/.cursor/mcp.json
{
"mcpServers": {
"omnifamous": {
"type": "http",
"url": "https://omnifamous.com/api/mcp",
"headers": {
"Authorization": "Bearer sk_your_secret_key"
}
}
}
}.cursor/mcp.json
{
"mcpServers": {
"omnifamous": {
"type": "http",
"url": "https://omnifamous.com/api/mcp",
"headers": {
"Authorization": "Bearer sk_your_secret_key"
}
}
}
}Project config is scoped to one repo. The settings UI config applies everywhere.
Every surface of the engine, as tools your agent can call mid-conversation.
One JSON block in ~/.cursor/mcp.json, or in the repo. Cursor gets the full tool catalog: keywords, SERP, backlinks, audits, AI visibility, drafting and publishing.
Run any prompt against ChatGPT, Gemini, Perplexity, Copilot, Grok or Google AI Overviews and get the answer, the citations and the query fan-out back.
Anonymous AI visibility audit and keyword research at /tools/seo-audit, no login required.
Every tool the MCP server exposes is also a plain HTTP endpoint, so you can script it without an MCP client.
Whichever way you connect, the omnifamous engine does the heavy lifting for your brand.
Audits your AI visibility so you can see which prompts name your brand and which name a competitor.
Researches winnable keywords for your market with real search data behind every pick.
Drafts ranking-ready articles mapped to those keywords, structured to be quoted by AI assistants.
Re-scans your visibility over time so you can prove the trend as content ships.
Four steps. You stay the editor, the engine handles the rest.
Add the server to Cursor. One JSON block in ~/.cursor/mcp.json, or in the repo.
Ask in plain language. The agent picks the tools, so you never call one by hand.
Review the research. Every number comes from a live data provider, not from the model.
Approve the drafts. You stay the editor while the engine handles research and writing.
Everything you need to know about how the engine does the work for you.
Yes. One JSON block in ~/.cursor/mcp.json, or in the repo. Once it is connected, Cursor can research keywords, pull live SERPs, check backlinks, run technical audits, ask ChatGPT and Gemini what they say about your brand, and draft articles against what actually ranks.
Cursor reads two configs: ~/.cursor/mcp.json applies everywhere, and a .cursor/mcp.json in the repo applies to that project only, which is the cleaner option when a client's keys should not follow you around. The exact config block for Cursor is on this page, ready to copy.
Project config (.cursor/mcp.json) is scoped to one repo, which is what you want when you run SEO for several clients. The home directory config (~/.cursor/mcp.json) applies to every window.
Because it will make the numbers up. Language models routinely invent search volumes, difficulty scores and competitor rankings that look plausible and are wrong. Every number our tools return comes from a live data provider or your own workspace.
Reads of your own workspace are free and unmetered: your articles, your calendar, your keyword pool, your visibility history. Anything that creates content or buys live market data spends credits, and every account starts with free credits and no card.
Yes, and that is the part no other SEO MCP server has. You can run any prompt against ChatGPT, Gemini, Perplexity, Copilot, Grok, Claude or Google AI Overviews on demand and get the answer text, every citation with its URL, and the query fan-out the engine actually issued.
Still have questions? Talk to us
Same engine, wherever your agent lives. See how it connects to each runtime.
Connect the server and it works from real keyword, SERP, backlink and AI-visibility data instead of numbers it invented. Start with the free audit, no account and no card.
Run my free auditNo credit card required. Results in under 5 minutes.Tell us about your site and we'll let you in when your spot opens up.