Find Shopify stores using an app and running Meta ads.

Cost guard: before each paid Spytrend call, run the free `estimate_cost` with the same tool and arguments, estimate the cost and stay within the user's stated budget; if the user has not stated a budget, tell them the estimate before the first paid call. If pricing consent is required, show the returned terms and stop until the customer explicitly agrees; never accept on the customer's behalf. If a required call does not fit the stated budget, make no paid call and return `Budget limit prevents required research`; omit an optional call that does not fit and disclose reduced coverage. When opening ads with `get_ad`, skip multilang ads (`is_multilang`) and never call `get_media` after `get_ad`. Retry once only when `retryable=true`, after `retry_after_seconds`. Read `params.ignored` after every search; an ignored required filter never counts as applied. Cost notices are not part of the final customer report.
Use the connected Spytrend MCP to find Shopify merchants with the exact detected app relationship **Klaviyo** and a verified active Meta ad in **US**.

Execute every research stage in this prompt in one turn. Apart from mandatory pre-call cost notices, do not emit intermediate responses and do not stop after scope validation. Stop early only for invalid input or unavailable required evidence. Return only the final customer report.

1. Validate the exact Shopify Apps catalog value and country. Preserve the app value exactly. Detection is an observed technology relationship, not proof of installation completeness, active use, configuration or performance. Country applies to the later ad-delivery gate, not store traffic.
2. Run one bounded store discovery with platform Shopify, exact `shopify_app`, `has_ads=true`, `exclude_infrastructure=true`, sort_by=visits, sort_order=desc and `limit=10`. Do not add country or minimum-ad filters and do not broaden or rerun discovery. Normalize domains, remove infrastructure/non-merchants and preserve order.
3. Use the exact app-filtered discovery rows directly as compact store/app evidence. Normalize and deduplicate exact registrable domains. Only when a row lacks the domain, platform or app evidence, check that store once with search_shops(q=<exact domain>, platform=Shopify, limit=5) and accept only the exact normalized-domain row. Never call `get_shop`. Missing app evidence is Not confirmed, not absent.
4. In returned order, run one exact-domain Meta search per store sequentially: landing_domain=<exact domain>, countries=[`US`], status_today=`active`, dedupe=true, limit=2. Open at most one representative ad per store with `get_ad`. Include a store only when the opened ad confirms active status, US delivery, a matching merchant offer and the exact first-party destination. Continue until five stores are verified or the candidates run out. Report reduced coverage rather than exceed the stated budget.
5. Required discovery/verification failure makes the run Partial; completed zero is Complete. Candidate failure excludes that candidate. Retry only on explicit `retry_after_seconds` guidance.

Write a table-free report under 300 words with 3–5 verified stores (fewer only when fewer passed). Show store/offer, exact detected app relationship, selected-country active-ad evidence, first-party route, one reason to inspect and High/Medium/Low confidence. Add `What stands out`, `Inspect first`, `Evidence & limits`, and useful unresolved evidence only when material.

Label merchant destinations `Website` or `Landing page`. Use `Evidence` only when MCP explicitly identifies a stable first-party research resource link. A merchant destination, image CDN, preview or media URL is never `Evidence`. Write the country as a full name. Hide service name, internal labels, app slugs, routes, tools, calls, raw IDs, UTC timing, elapsed time, tokens, retries, pagination and run summaries.

End exactly:

`Technology detection and Meta ad activity are observable relationships. They do not prove active app use, conversions, revenue, ROAS or profitability.`

Only after `Show supporting record`, use retained responses for non-sensitive route and verification details without rerunning research.