recipes / audit-app-store-listing

Audit your App Store listing with Claude in one prompt

One prompt in, a full audit out: what your metadata covers today, which keywords you are missing, and a redraft that fixes both — with the math checked before you touch App Store Connect. Works in Claude Code, Claude.ai, Cursor, or any MCP client with openaso connected.

the prompt

Audit the App Store listing for <your app — name or App Store id> in the US.

1. What does our visible metadata cover today, and what is wasted?
2. Which keywords should we target but don't? Score them by demand
   vs. how strong the apps ranking there are.
3. Propose a new title, subtitle, and keyword field — then verify
   every target phrase is actually covered by the combined word pool.

what your agent does

1. 01

   audit

   One call per storefront gathers the live listing, its visible word pool, and autocomplete-discovered keyword candidates already scored for demand and competition.

2. 02

   rank

   Spot-checks where the app actually stands on the shortlisted keywords — live per-storefront positions, top 200 deep.

3. 03

   Drafts replacement metadata against the model the server teaches: title, subtitle, and hidden keyword field index as one combined word pool, so a word repeated across fields is wasted budget.

4. 04

   coverage

   Verifies the draft locally: character budgets, duplicate words, keyword-field waste, and whether each target phrase can be assembled from the pool.

what you get back

A gap list with demand and competition scores, a redrafted title (30 characters), subtitle (30), and keyword field (100), and a per-phrase coverage table proving the draft targets what it claims. The agent also passes along the honest caveats — cross-field phrases tend to rank weaker, so verify with rank after shipping.

{
  "phrase": "habit streak tracker",
  "covered": true,
  "via": { "habit": "title", "streak": "subtitle", "tracker": "title" }
}

make it yours

keep going