Manifold MCP

Picture asking Claude a single question: "Which keywords do we rank 5 to 15 for that could reach page one?" Instead of a guess, you get back a real list, like "crm for startups" at position 7 with 2,900 searches, then a concrete fix list for your /pricing page: put the keyword in the title, answer "is there a free plan?" because ChatGPT gets it wrong, and link to the page from your four best-ranking posts. That's the workflow Manifold MCP is built around. It isn't another dashboard you log into. It's a hosted MCP server that plugs live marketing data into the agent you already use.

SEO is only one channel. Manifold exposes tools like seo_get_ranked_keywords, aeo_run_ai_answers, reddit_search_posts and linkedin_search_posts, and your agent picks which ones to call. You can check whether ChatGPT, Claude and Gemini recommend you for buyer questions and which pages they cite. You can find TikTok, YouTube and Instagram creators in your niche with transcripts of their top videos, pull a competitor's live Meta and Google ads with how long each has run, surface LinkedIn and X posts from people complaining about the problem you solve (plus their verified work email), and find Reddit threads asking for what you make so your agent can draft a reply. Setup takes about a minute, with no API keys, and it works with Claude, ChatGPT, Cursor and other MCP apps.

Pricing runs on one credit balance across every tool. You get 250 free credits once with no card, then Pro starts at $20/month for 2,500 credits and Max at $250/month for 42,000. Your agent sees the credit cost of a call before making it, which is a nice touch for a usage-based model. The real question is whether you want your marketing research to live inside a chat window. If you already spend your day in Claude or Cursor, this is one of the more practical MCP servers to add.