agentleFS
SpytrendSpytrend/spytrend-mcp/skills/spytrend-website-review/SKILL.md

spytrend-website-review

Skill30 starsChanged 3 months agoMITView on GitHub

Review one website or store domain with SpyTrend MCP data: monthly traffic estimate, advertising footprint over time, which advertiser pages run ads to it and its leading creatives, from a database of 1B+ tracked Meta ads. Use when the user gives a domain or URL and wants its traffic, whether and since when it runs ads, who sends paid traffic to it, or how its ad volume changed month by month. Triggers: "store spy", "does this site run ads", "how much traffic does this store get", "check this domain: traffic, ads, creatives", "competitor website traffic"; RU «крутит ли домен рекламу», «сколько трафика у сайта», «спай магазина». Use this instead of guessing about a specific site's ads or traffic. Not for discovering new stores (use spytrend-shop-discovery), whole niches (use spytrend-vertical-intel) or the operation behind the ads (use spytrend-webmaster-overview).

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 20
  2. Very specific:

    Call get_usage. State explicit seed limits, profile cap, media count and worst-case spend.

    line 26
  3. Distinctive:

    Default media to off. Do not download or inspect the website itself.

    line 28
  4. Very specific:

    Call search_shops with the domain, optional selected country and a small explicit limit.

    line 32
  5. Very specific:

    Call get_shop for the primary returned surface.

    line 35
  6. Very specific:

    Call search_webmasters by exact domain with optional selected countries and a small limit.

    line 45
  7. Very specific:

    Call one bounded search_ads seed page per selected surface with landing_domain_exact=true and optional selected country.

    line 48
  8. Distinctive:

    Keep Markdown and HTML sections equivalent: website snapshot, traffic period and quality, advertising volume, advertisers, networks, …

    line 72
  9. Very specific:

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

    line 77

≈ “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 website advertising review

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

Normalize and preflight

  1. Accept one domain, remove scheme, path and leading www, and preserve the user's exact input as context. Reject credentials and IP-literal targets.
  2. Call get_usage. State explicit seed limits, profile cap, media count and worst-case spend.
  3. Default media to off. Do not download or inspect the website itself.

Workflow

  1. Call search_shops with the domain, optional selected country and a small explicit limit. Match the exact host or registrable root deliberately and state which scope was selected.
  2. Call get_shop for the primary returned surface. Extract only bounded public fields. Traffic history is a monthly estimate for its stated month.
  3. Label ads_monthly as ads first discovered by SpyTrend. It is not a Facebook launch series. Mark the open current month incomplete.
  4. For a closed launch window, call search_ads with landing_domain, landing_domain_exact=true, optional selected country, concrete dates and a one-row limit, then use pagination.total only with its date-axis and total status. If launch-period analytics are unavailable, say so; do not substitute the discovery series.
  5. Call search_webmasters by exact domain with optional selected countries and a small limit; present matches as observed advertising-network attribution, not ownership.
  6. Call one bounded search_ads seed page per selected surface with landing_domain_exact=true and optional selected country. Use its total for volume and its returned rows only to discover advertiser candidates. Cap advertiser cards, call get_advertiser for only those unique ids, and report processed-versus-seen coverage. A card without a card_status field is complete; card_status=building means the card is a placeholder whose counters are unavailable, not zero.
  7. For website-scoped creative reuse, call search_ads again with the exact domain/country/date filters and sort_by=most_reused_creative, using a small selection: on a high-volume domain this ranking exceeds the analytics deadline and returns a timeout instead of rows. Always pair it with a concrete date window, and add the selected country when the domain is large; widen only if the bounded call returned rows. Report the bound you used. Also treat the two reuse counters as separate measurements: when same_creative_ads_in_selection exceeds same_creative_ads_lifetime, report both as returned and do not present the pair as a subset relation. explicit limit. Its representative ads and selection reuse counts are the website-scoped ranking. Do not substitute a country-global search_creatives page or re-sort a bounded lifetime list.
  8. With confirmation, retrieve media once for the selected representative ad ids and reuse it in every report section.

Report

Keep Markdown and HTML sections equivalent: website snapshot, traffic period and quality, advertising volume, advertisers, networks, creative leaders, coverage, period-axis notes and spend. A different landing domain is a returned fact, not proof that businesses are unrelated.

Use references/report-template.md or the packaged renderer with references/report-template.html; otherwise provide the full report in chat.

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.