Metadata optimization
ASO & app-market research skills for AI agents — Claude Code, Cursor, and any MCP client. Install: npx skills add getappniche/aso-skills
npx -y skills add getappniche/aso-skills --skill metadata-optimizationAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 12 days oldThe repository was created 12 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Draft optimized App Store title, subtitle, keyword field, and description variants with exact character counts. Use when the user wants to rewrite store metadata, improve ASO, or prepare a metadata A/B test.
SKILL.md
2.5 KB, as published. Nobody here has run it
Metadata Optimization
Write store metadata that a reviewer would approve and a search algorithm would reward.
Requirements
Works standalone. With the GetAppNiche MCP connected, ground the draft in data first:
get_app_detail for the app's current metadata and get_keyword_difficulty /
batch_keyword_difficulty (10 credits per keyword) for the terms you're about to
spend title characters on. If a keyword-research skill run already produced a
"target now" list in this conversation, reuse it — don't re-score.
Hard limits (Apple App Store)
- Title: 30 chars · Subtitle: 30 chars · Keyword field: 100 chars
- Keyword field: comma-separated, no spaces after commas, no duplicates of words already in title/subtitle (they're already indexed), no plurals of included singulars, no competitor brand names.
Google Play: title 30, short description 80, long description 4000 — keywords are extracted from natural description text instead of a keyword field.
Workflow
- Inventory. Current title/subtitle/keywords (from
get_app_detailor the user), the target keyword list, and the one differentiator the app owns. - Allocate characters by value. Highest-traffic viable keyword goes in the title; next tier in the subtitle; the rest into the keyword field. Brand name earns its spot only if people actually search for it.
- Draft 3 variants with different bets: (A) keyword-max, (B) benefit-led,
(C) balanced. Show
(N/30)style character counts on every line — count before you claim; overlong metadata is an automatic fail. - Description (on request): first 2–3 lines carry the hook (that's what shows before "more"), then feature bullets, social proof, closing CTA. For Google Play, weave target phrases into natural sentences at roughly 2–3 mentions each — never keyword-stuff.
- Explain the trade-offs between variants in 2–3 sentences and pick a default.
Output
Three labeled variants with per-line character counts, a filled keyword field, the recommendation, and a short "what we deliberately left out" note.
Guardrails
- Never exceed a character limit, even by one.
- Metadata changes ship with app releases on Apple — bundle recommendations accordingly; don't promise instant ranking effects.