AGENTSOURCE

The Shelf / App Store / App Store Keyword Engine

App Store

App Store Keyword Engine

Your agent rewrites your App Store name, subtitle and keywords from the apps actually winning those searches.

The job: your agent picks your App Store words from what the store is showing today, not from guesswork.

Most keyword advice fails the same way. Somebody guesses what the competition is going after, crams the guesses into a subtitle, and then wonders why nobody ever finds the app. This skill makes the agent go and look. For every search term you're weighing up, it pulls the apps really sitting at the top of that result list — straight off Apple's own free public pages, using the trick that stops it grabbing the wrong app's subtitle off the page, a mistake we made and wrote down. Then it counts which words those apps use, lays the count out in a table, and rewrites your app name, subtitle and keyword field one character at a time.

What's on the tag:

  • The first check of all: does your own app come up when someone searches its name? If it doesn't, everything after that changes, and the skill tells you which of three causes you've got.
  • The wrong-neighborhood check. The most expensive mistake in App Store search is one word in your name that files you under the wrong category, against apps you cannot beat. One documented run found a productivity app buried among journaling apps with six-figure review counts, because of a single word in its name. The check finds that in minutes.
  • Rules for filling in each field: write keywords in the singular ("category" also covers "categories", but not the other way round), never use the same word twice across two fields, and usually lead your app name with the search word instead of your brand — plus how to check whether your category is one of the ones that runs the other way. And how to work three or four of your search phrases into the first 250 characters of the description, as normal sentences and not a word list — that's the part people see before they tap More.
  • A rewrite for eight more countries, using the local word for your category that most foreign apps never bother with. The German and Japanese ones alone are why a translated subtitle is the cheapest ranking win on the shelf.
  • A rollout plan that knows where the sharp edges are: which fields change live without waiting on review, and why starting a new version with no build attached leaves you a draft you can never delete. That one was learned the hard way.

Who it's issued to: indie iOS developers and small studios whose apps never show up in search, and anyone who wants the every-few-months keyword pass to be one prompt instead of an afternoon of juggling browser tabs.

Why not a free download off some list? Free skill files are copied patterns nobody ever ran. This one shipped real App Store listings, and every warning in it — the four ways the page pull goes quietly wrong, the draft you can't delete, the risk of renaming an app while it's in review — was paid for on a live app. It's kept up to date: when Apple moves a limit or changes a screen, the file gets fixed and your locker gets the new copy.

Runs on Apple's free public data. No keyword subscription needed.

FIELD REPORT real output, not a promise

From a documented run of this workflow on a voice-notes app (Productivity category) that didn't rank in the top 20 for its own brand name. Excerpted and anonymized from the skill's own worked example — this is the Step 0 diagnosis, the Step 2.5 intent-poisoning gate, and the resulting gap-table actions, as produced.

Step 0 — Brand-defense check

Query: iTunes Search for the exact brand token, then the full live name. App absent from the top 20 on both.

Diagnosis (two root causes, combined):

  1. Velocity — 0 reviews, 11 days since launch. Apple's index weights install/review velocity; metadata cannot fix this directly. Path: post-accomplishment rating prompt + organic installs + time.
  2. Wrong-category framing in the Name — the live Name carried the token Journal, filing the app into Health & Fitness / Lifestyle search results. The app is a Productivity app.

Step 2.5 — Intent-poisoning gate

Search term Top-3 category App's category Verdict
voice journal Health & Fitness Productivity DROP
ai journal Health & Fitness Productivity DROP
audio diary Lifestyle / Health Productivity DROP
voice memo ai Productivity Productivity KEEP
ai voice notes Productivity Productivity KEEP
ai note taker Productivity Productivity KEEP

The journal/diary cluster — the terms the app was actually indexed for — resolves to a different category, where the top-3 results are journaling incumbents with six-figure review counts. No subtitle cleverness competes there. The Productivity-side terms are the right neighborhood.

Resulting actions

  • Name: remove the Journal token; rename the tagline portion to a category-matched head term ("AI Voice Note Taker" pattern) at the next binary bump — never mid-review.
  • Indexed package: keep all journal tokens out of Name, Subtitle, and Keywords; rebuild around the three KEEP terms so cross-field combinations (voice + note + ai + memo) form the discoverable phrases.
  • Set expectations before deploy: because root cause #1 is velocity, the metadata fix is necessary but not sufficient — flagged to the owner before anything shipped rather than after.

That is the shape of every run: a live-data verdict table, an explicit action per token, and the honesty about what metadata can and cannot fix.

SERVICE RECORD living gear — updated as the factory learns

v1.1.0 — 2026-08-25

Field-tested updates from a recent multi-app ASO cycle. New: a free Apple-autocomplete search-volume probe (with the two silent instrument traps that make it read "no demand" when it's just misconfigured); a "measure what the current Subtitle already earns before replacing it" step that saved three top-4 rankings on one run; the intent-poisoning gate now warns that a clean category match can still be the wrong industry — with the disambiguating-token test that proves it; subagent-delegation rules for the data pull (including the fabricated-competitor-row trap); and a "shipped vs. done" audit that catches finished packages that never reached Apple. Corrected: the subtitle scrape's four silent failure modes (Apple changed the page markup — the old JSON-field method now reads every subtitle as empty), and the "visual search page" browser check is retired — the route 404s for everyone; ad-slot presence is not obtainable from the public web. Keyword-first Name ordering is now explicitly a verify-against-your-category heuristic, not a law.

v1.0.0 — 2026-07-17

First issue. Ported from the factory's internal skill: sanitized for general use, methodology intact, field report captured from a real run.

Every update ships free to owners — your locker always serves the latest version.

QUESTIONS

Do I need to pay for a keyword tool?

No. The whole thing runs on Apple's free public data — the same search results, app pages and charts the App Store shows anyone. If you already pay for a keyword tool, the skill says where its numbers fit in and which checks it still can't do for you.

Does it update my listing on App Store Connect for me?

It hands you finished text, ready to paste, plus the order to change things in. If your agent has your App Store Connect key, the skill walks that safe order — including the stop that keeps you from creating a draft version you can never delete.

Will this work outside finance and productivity apps?

Yes. The steps don't care what your app does — the finance ones are just worked examples. The other-countries section has a general procedure you follow for any kind of app.

What agents does it run on?

Written for Claude Code, and it works in claude.ai Projects and Codex setups too. It's a plain text file — any agent that can read instructions can follow it.