agentleFS
SpytrendSpytrend/spytrend-mcp/skills/spytrend-advertiser-overview/SKILL.md

spytrend-advertiser-overview

Skill30 starsChanged 3 months agoMITView on GitHub

Rank the top advertiser pages in a chosen vertical and country and profile each in depth with SpyTrend MCP data: current winning ads, freshly launched ads, creatives and destination pages, from a database of 1B+ tracked Meta ads. Use when the user wants to see who the biggest advertisers in a market are and what they run, profile market leaders' current ads, check which pages have the most active ads, or catch fresh launches from top pages. Triggers: "spy on competitor ads", "top advertisers in a niche or country", "biggest advertisers", "longest-running ads", "whose ads dominate this geo"; RU «кто крупнейшие рекламодатели и что они крутят», «топ рекламных страниц рынка», «спай конкурента»; ES «espiar anuncios de la competencia»; PT «anúncios dos concorrentes»; JA «競合の広告». Not for a whole-vertical map (use spytrend- vertical-intel), one website (use spytrend-website-review) or a text search (use spytrend-keyword-search).

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 2 are distinctive, and 7 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 7

    Very specific

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

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

    Read references/spytrend-contract.md before making any data call.

    line 21
  2. Distinctive:

    Default profile count: five.

    line 26
  3. Distinctive:

    Default output: chat; default media: off.

    line 27
  4. Very specific:

    Call get_usage, state the profile count and maximum spend, and get media confirmation separately.

    line 31
  5. Very specific:

    Call get_trends with dimension=advertisers, the selected AI category, country, min_geo_share=0.3, explicit limit and the current-volume …

    line 38
  6. Very specific:

    When an exact-filter state is returned, use filter_match.exact_active only if filter_match.exact_available is true.

    line 44
  7. Very specific:

    Call search_ads with page_id set to the advertiser page id, the selected categories and countries, status_today=active, sort_by=date, …

    line 58
  8. Very specific:

    Use advertiser card data for exact page counters and destination coverage when the card carries no card_status field: a complete card does …

    line 68
  9. Very specific:

    Use references/report-template.md or the packaged renderer with references/report-template.html.

    line 85

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

SpyTrend advertiser overview

Read references/spytrend-contract.md before making any data call.

Inputs

  • Required: vertical and ISO-2 country.
  • Default profile count: five; allow three or ten when the user chooses depth.
  • Default output: chat; default media: off.

Preflight

Call get_usage, state the profile count and maximum spend, and get media confirmation separately. If the account does not own the requested vertical, show the returned access state instead of calling an unfiltered market and pretending it answers the original question.

Workflow

  1. Call get_trends with dimension=advertisers, the selected AI category, country, min_geo_share=0.3, explicit limit and the current-volume sort. The geo-share gate limits the ranking to pages for which the selected country is a dominant geography; a bare country filter is only a present-in match and ranks global pages with a marginal local slice first. State the threshold beside the leaderboard. Preserve server order.
  2. When an exact-filter state is returned, use filter_match.exact_active only if filter_match.exact_available is true. Otherwise label the page as a server-ranked candidate and do not claim exact filtered activity.
  3. For recently discovered advertiser activity, make a separate get_trends call using sort_by=launched_14d. This field currently counts ads first discovered by SpyTrend in that window; label it exactly that way, never as Facebook launches. Never derive this table by re-sorting the bounded current leaderboard.
  4. For each admitted advertiser, call search_creatives with its page id and category, without country, ordered by current activity with a small explicit limit. Call the result an exact advertiser slice only when the response's slice/sort status says so; otherwise label it a bounded candidate sample. Country-specific examples come from the ad query below, not by weakening the advertiser creative slice.
  5. Call search_ads with page_id set to the advertiser page id, the selected categories and countries, status_today=active, sort_by=date, dedupe=true and a small explicit limit. A first-discovered date stays labeled as first discovered unless the response supplies the platform launch date. Pass the page id in page_id, never in advertiser_id: that parameter takes the internal advertiser UUID returned by search_advertisers, and a page id placed there is ignored without an error — the call then answers for the whole unfiltered market. If you hold the UUID instead, pass it in advertiser_id. Before reporting, verify the binding: every returned row must carry the identifier you filtered on.
  6. Use advertiser card data for exact page counters and destination coverage when the card carries no card_status field: a complete card does not emit it. When card_status is building, the card is a placeholder and its counters are unavailable, not zero; say so instead of reporting numbers. Category counters describe the labeled subset and must be presented with coverage.
  7. With media confirmed, fetch only the final selected ids. Prefer the lower-cost representative-ad media path when one placement example is sufficient, and state that choice.

Report

Show two separately sourced leaderboards (current activity and recently first discovered ads), profiled-versus-returned coverage, current creative winners, fresh ads, destinations, exact/unavailable states and spend. Do not present a bounded candidate pool as the complete market.

Use references/report-template.md or the packaged renderer with references/report-template.html. Chat must contain the full result when file creation is unavailable.

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.