Audit app store keywords for gaps
asc-aso-auditskillsetup L2★796
rorkai/app-store-connect-cli-skills ↗Causal-lift measurements
aso-audit+16pp vs no-skill baselinewith-skill 84% · baseline 68%
Measured by running the task with and without this artifact, K=5, graded by deterministic checks — no LLM judging.
What it does
Audit app store metadata for keyword waste and ASO gaps
Best for
App marketers publishing to Apple App Store who want to validate metadata efficiency (keyword waste) and field utilization before submitting new versions.
Inputs
- · Metadata directory structure (app-info/{locale}.json, version/{version}/{locale}.json)
- · App name (from export or asc apps info list)
- · Target locale (default: en-US)
Outputs
- · Keyword waste findings (tokens in subtitle + keywords, wastes budget)
- · Underutilized fields (keywords < 90 chars, subtitle < 20 chars)
- · Missing field errors (empty subtitle, keywords, description, whatsNew)
- · Bad separator findings (spaces after commas, semicolons, pipes)
- · Cross-locale gaps (keywords identical to primary locale = not localized)
Requires
- · asc metadata pull (prerequisite to run first)
- · Astro MCP (optional, for gap analysis phase 2)
Preconditions
- · Metadata pulled via asc metadata pull and normalized into canonical layout
- · App name available (from export or asc apps info list)
Failure modes
- · Doesn't normalize tokenization for CJK (requires character-level split, not whitespace)
- · Fuzzy matching misses prefix-stripped Arabic variants (ال stripping missed)
- · Flags name field empty when it's not in the export JSON at all
Trust signals
- · 5-phase audit structure (keyword waste, underutilized fields, missing fields, bad separators, cross-locale gaps)
- · CJK tokenization rules documented (split by CJK punctuation, not whitespace)
- · Arabic prefix stripping for ال variant detection
- · Field minimum utilization thresholds (keywords 90/100 chars, subtitle 20/30 chars)