agentleFS
eduardmureduardmur/searcherries-claude-plugin/skills/improve-ai-visibility/SKILL.md

improve-ai-visibility

Skill0 starsChanged 42 days agoMITView on GitHub

Use when the user wants AI assistants and AI search to recommend their brand more often, asks why competitors appear in AI answers instead of them, or asks which pages of their website to change based on their Searcherries data. Reads the Searcherries tools and turns the evidence into concrete page changes.

The file

We split this skill into the rules it gives an agent

10 rules. About 0 are common, 5 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If the user names a brand or domain, call get-action-plan with project set to it and audience set to marketing (use seo when the user asks …

    line 12
  2. Very specific:

    If you can read the website's code, look for the site's domain in the repository first (site config, package.json homepage, CNAME, sitemap, …

    line 13
  3. Very specific:

    If several projects match, or none, call list-projects and ask the user which one to use.

    line 14
  4. Very specific:

    Use the project_id from step 1.

    line 18
  5. Very specific:

    Search demand: get-search-performance with dataset: "opportunities" (high impressions, low CTR, long-tail queries) and dataset: …

    line 23
  6. Distinctive:

    If a source is not connected, say so and continue with what is available.

    line 25
  7. Distinctive:

    add an FAQ section for question-shaped queries that already bring impressions.

    line 33
  8. Distinctive:

    If you can work with the website's code, find the source file behind each URL (routes, page components, Markdown or MDX content, …

    line 38
  9. Distinctive:

    Do not invent numbers, rankings or competitors, and do not promise that a change will win a recommendation.

    line 43
  10. Distinctive:

    Keep claims on the site true.

    line 45

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Improve AI visibility from Searcherries data

Searcherries tracks how AI platforms answer the customer questions a brand cares about, which competitors and pages they name, and the site's Google Search Console, Bing Webmaster Tools and GA4 AI-traffic data. The Searcherries tools return that data read-only. If the tools are not available, tell the user to connect Searcherries first and stop.

1. Find the project

  • If the user names a brand or domain, call get-action-plan with project set to it and audience set to marketing (use seo when the user asks about search traffic). It returns dated priorities and the project_id.
  • If you can read the website's code, look for the site's domain in the repository first (site config, package.json homepage, CNAME, sitemap, canonical URLs).
  • If several projects match, or none, call list-projects and ask the user which one to use. Never guess.

2. Collect the evidence

Use the project_id from step 1.

  • Gaps in AI answers: get-ai-answers with prominence_max: 1 lists the answers where the brand is missing or only mentioned in passing. Read the full text of at most five of them (include_answer_text: true, limit: 5).
  • Who wins instead: get-competitors (default ranking) shows the brands the platforms name; dataset: "detail" with competitor explains one brand.
  • Which sources the answers rely on: get-citations with prompt_id for a gap question shows the cited URLs; own_site_only: true shows which of the site's own pages are already cited.
  • Search demand: get-search-performance with dataset: "opportunities" (high impressions, low CTR, long-tail queries) and dataset: "ai_question_queries" (question-shaped searches). Use source: "bing" for Bing.

If a source is not connected, say so and continue with what is available. Missing data is unknown, not zero.

3. Turn the evidence into page changes

For each relevant page, propose specific changes tied to the evidence, for example:

  • answer the tracked question directly near the top of the page, in the words customers use;
  • state plainly what the product is, who it is for and how it compares, covering the facts the AI answers give for competitors;
  • add an FAQ section for question-shaped queries that already bring impressions;
  • link to the page from related pages so it is easier to find.

Flag gaps that need a new page rather than an edit.

If you can work with the website's code, find the source file behind each URL (routes, page components, Markdown or MDX content, templates), present the plan first, and edit files only after the user agrees. Otherwise, list the changes by page URL.

Ground rules

  • Searcherries data is read-only and dated. Quote the period of every figure you use.
  • Do not invent numbers, rankings or competitors, and do not promise that a change will win a recommendation.
  • Visits are not revenue. Do not present AI traffic as sales.
  • Keep claims on the site true. Do not add statements the user cannot back up.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.