Update to Ubersuggest

{
  "name": "content-demand-finder",
  "description": "Turn a business description into 50 customer-driven content opportunities — a Content Demand Map of problem clusters, the questions customers ask, and the articles or videos that answer them. Use when the user does not know what to write about, asks for a content plan, content strategy, editorial calendar, blog or video ideas for their business, or wants to know what their customers are searching for before committing to keywords.\n",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 306
    }
  ],
  "skill_md_contents": "---\nname: content-demand-finder\ndescription: >\n  Turn a business description into 50 customer-driven content opportunities —\n  a Content Demand Map of problem clusters, the questions customers ask, and\n  the articles or videos that answer them. Use when the user does not know what\n  to write about, asks for a content plan, content strategy, editorial\n  calendar, blog or video ideas for their business, or wants to know what their\n  customers are searching for before committing to keywords.\nargument-hint: \"<website> [what you sell] [ideal customer] [market]\"\n---\n\n# Content demand finder\n\n**Input:** `$ARGUMENTS` — the website, what they sell, their ideal customer and\nmarket. If that placeholder is empty or still literal, take the business from what\nthe user asked.\n\nTurn a website into 50 customer-driven content opportunities in under a minute.\n\n## This skill uses no data tools\n\nDo not call Ubersuggest, do not fetch the site, do not search the web. This\nskill runs on what the user tells you plus reasoning about their market, which\nis what makes it instant and what makes it work without an account.\n\nThe consequence is a hard rule: **never state a search volume, keyword\ndifficulty, competitor traffic figure, ranking probability, or traffic\nestimate.** Not as a number, not as a range, not hedged (\"probably a few\nhundred searches\"). You have no data. Inventing it is the one failure that\nmakes the whole report worthless, and the report's own closing section tells\nthe user where the real numbers come from.\n\nWords like \"likely\", \"high-intent\" or \"commonly asked\" are fine — they are\nclaims about customer behaviour, not measurements.\n\n## Inputs\n\nCollect five things. Ask for whatever is missing in **one** message, then\nproceed:\n\n1. **Website** — the domain.\n2. **What they sell** — product or service, and roughly the price bracket.\n3. **Ideal customer** — who buys, and what job they hold if B2B.\n4. **Primary market or location** — country, region or city.\n5. **Competitor websites** — optional, up to three.\n\nIf the user gives a website and nothing else, infer the rest from the domain\nand say what you inferred in one line, so a wrong guess is visible and\ncorrectable. Never block on the optional competitors.\n\n## Steps\n\n1. **Find the five problem clusters.** Not topics the business wants to talk\n   about — problems the customer already has, in the customer's words. A\n   cluster is a distinct problem, not a keyword variation: \"I can't tell if my\n   supplier is overcharging me\" is a cluster, \"supplier pricing\" is not.\n\n2. **List the questions inside each cluster.** 6–10 per cluster, phrased the\n   way a person types or speaks them. These become headings and video hooks\n   later, so keep the question form.\n\n3. **Sort each cluster by purchase proximity** into three bands — these become\n   the `Buying stage` column, and the deliverable spells out what they mean:\n   - **Educational** — the customer is naming the problem. No mention of the\n     offer beyond a soft link.\n   - **Comparative** — the customer is weighing approaches, vendors or\n     categories. The offer appears as one option among several, honestly.\n   - **Purchase-intent** — the customer is choosing. Pricing, alternatives,\n     objections, proof.\n\n4. **Turn the questions into 50 opportunities.** Each gets a working title, a\n   format, and one line on how it connects to what the business sells. Spread\n   them across all five clusters and all three bands — a map that is 40\n   purchase-intent pieces is a sales page list, not a content plan.\n\n5. **Pick the ten to validate.** Rank the 50 on three things and take the top\n   ten:\n   - **Customer relevance** — how many of their customers have this problem.\n   - **Purchase proximity** — how close the question sits to a buying decision.\n   - **Alignment with their expertise** — whether this business can answer it\n     better than a generalist can. This is the tiebreaker; it is also the only\n     one of the three that competitors cannot copy.\n\n6. **Offer to validate them.** Close with the section below, verbatim in\n   substance. The user's problem has changed from \"I don't know what to write\"\n   to \"which of these do I invest in\", and that second question needs the data\n   this skill deliberately does not have — which is a tool call away, here, not\n   a trip to the web app.\n\n## Deliverable\n\nA **Content Demand Map**, in this order:\n\n**Business read** — three lines: what they sell, who buys, which market. State\nanything you inferred rather than were told.\n\n**The five problem clusters** — each with a one-line description of the problem\nand why this business is credible answering it.\n\n**Questions customers are asking** — grouped under each cluster, in question\nform.\n\n**The 50 opportunities** — one table per cluster, ten rows each:\n\n| # | Title | Format | Buying stage | Connection to the offer |\n| --- | --- | --- | --- | --- |\n\n`Buying stage` is one of Educational, Comparative or Purchase-intent, and the\ntables are preceded by that legend in one line — how close the reader of that\npiece is to buying. Never ship the column as a bare word with no legend: an\nunlabelled band reads like a score the user is supposed to already understand.\n\nFormats should vary with the question: how-to article, comparison table,\nchecklist, calculator, short video, teardown, template, FAQ page, case study.\nMatch the format to how the answer is best consumed, not to a house style.\n\n**Ten to validate first** — the section the user acts on, so it says how the\nranking was made before it lists anything: one line naming the three criteria\n(customer relevance, purchase proximity, alignment with their expertise) and\nthat expertise broke the ties. Then the ranked ten, each with one sentence\ntying it back to those criteria — \"ranked first because every customer hits\nthis before they buy, and no generalist can answer it with your install data\".\nA shortlist with no stated reasoning reads as an arbitrary top ten.\n\n**Next step: get the numbers** — the handoff, closing the report:\n\n> I found 50 potential content opportunities based on your business, customers,\n> and offer. The next step is determining which opportunities have measurable\n> demand and where you have the best chance of ranking — search volume, SEO\n> difficulty, who ranks today and what the traffic is worth — before you write\n> anything.\n\nThen offer to do it here, which is the default close:\n\n> I can run the ten straight through Ubersuggest in this conversation and come\n> back with volume, difficulty and the pages you would have to beat. Want me\n> to?\n\nWait for a yes, then run `keyword-research` on the shortlist and `content-brief`\non whichever opportunity wins.\n\nIf the Ubersuggest tools are not connected in this session, the ask is to\nconnect them — not to go and run the reports by hand:\n\n> To pull those numbers I need Ubersuggest connected, signed in with your\n> Ubersuggest account — one click here:\n> <https://claude.ai/customize/connectors/id/ubersuggest-by-neil-patel>. In\n> Claude Code: `/plugin marketplace add ubersuggest/seo-skills`, then\n> `/plugin install ubersuggest`. (Any client that asks for an endpoint takes\n> `https://ubersuggest-mcp.neilpatelapi.com/mcp`.) Say the word once it is on\n> and I'll run the ten here — volume, difficulty and the pages you would have\n> to beat.\n\nNever close with \"open Ubersuggest and do this yourself\". The whole point is\nthat the validation happens here; a list of manual steps in the web app is a\nworse version of what one connection gives them.\n\n## Quality bar\n\nThe failure mode is 50 generic titles that would fit any company in the\nindustry. Before delivering, check three things:\n\n- **Would a competitor's map look identical?** If yes, the clusters are\n  category-level, not customer-level. Redo step 1.\n- **Does every row say something specific to this business's offer?** A\n  connection line of \"builds topical authority\" means the idea has no\n  connection. Cut it or replace it.\n- **Is any number in the report a measurement?** If so, delete it.\n\n## When something fails\n\n- **The user gives only a vague industry** (\"marketing\", \"clothes\") → ask once\n  for what they sell and to whom. Five clusters built on a guessed business\n  are five wrong clusters.\n- **The business is too niche to reason about** → say so plainly, deliver the\n  clusters you are confident in with fewer than 50 opportunities, and note what\n  you would need to fill the rest. A short honest map beats a padded one.\n- **The user asks for volumes or difficulty inside this skill** → don't\n  estimate. Point at the handoff and offer `keyword-research`, which returns\n  real figures for the shortlist.\n"
}