App discovery

When to use

• The user wants a list of apps for a use case, category or feature ("habit trackers with subscriptions", "AI photo editors on Android").
• The user asks which apps in a space are growing now, or which are new.
• As the first step of competitor, market or keyword work when no app list exists yet.

Not for one known app: use get_app_detail directly, or the competitor-analysis skill.

Inputs

Ask for these only when they matter and are missing:

Tools

• search_apps — search and filter apps; sort by installs, daily_installs, monthly_installs, monthly_revenue, released, updated, ratings_count or reviews_count.
• get_app_detail — full profile of one app, including ratingScore.
• get_app_historicals — daily time series (installs, revenue, ratings, reviews, score).

Workflow

1. search_apps with the main query, sort: "installs", limit: 20. Without search_in, the query matches titles and short descriptions (the App Store subtitle, the Google Play summary), which suits most feature searches. For a broader search that also reads the long description, pass search_in: ["title", "short_description", "description"].
2. search_apps with the same filters and sort: "daily_installs": the apps growing now.
3. Optional: sort: "released" with released_after (YYYY-MM-DD) to find new entrants.
4. Remove off-topic rows (a query word can match an unrelated title). Keep 5–10 apps.
5. get_app_detail for each kept app: category, ratingScore, ratingsCount, price, last update.
6. For the 3 most interesting apps, get_app_historicals with metrics: ["installs", "revenue"] (the last 90 days by default) to confirm the trend.

Use pagination.nextCursor only when the first page is not enough. Each page is one call against the quota.

Rows follow sort, not text relevance. To look up one known app by name, use get_app_detail; a name that exists on both stores needs store.

Output

1. A table: app, store, developer, installs, daily installs, revenue (estimate), ratingScore, release date, link (asospyUrl).
2. "Growing now": the apps with the highest daily installs relative to total installs, and their 90-day trend.
3. Three short takeaways (for example, a crowded top, a subscription-heavy category, a recent entrant growing fast).

Limitations

• iOS installs and all revenue are ASOSpy estimates. Revenue can be null (no estimate); do not read null as zero.
• Small apps show installs as text, "< 100" (Google Play) or "< 1k" (App Store). Say so; do not turn it into a number. get_app_historicals keeps these days as numbers: quote a day under meta.metrics.installs.floor the same way.
• search_apps has no country filter; titles are the US listing.
• min_installs/max_installs do not work with store: "ios". Use daily or monthly installs instead.
• Removed apps are hidden unless include_removed: true.