Built for HackOnVibe · DoraHacks

LaunchCopilot

Paste your app's store listing. Get a graded launch.

An ASO grader + a validated launch kit — for the solo builder who just shipped.

$ npx aso-lint --score   → 34/100  ·  rewrite → 93/100
The launch that wasn't

Six downloads.
Four were friends.

Not because the app was bad — because nobody could find it.

The problem

Maya shipped on a Friday.
By Sunday the store had buried her.

A solo dev in Nairobi. Four months of nights and weekends. Six downloads by Sunday — because her title used 12 of the App Store's 30 characters, her keyword field wasted a third of its budget on words Apple ignores, and she'd never heard of any of the rules.

TITLE
12 / 30 characters used — 18 wasted
KEYWORDS
⅓ of the budget on terms Apple ignores
DESCRIPTION
First 255 chars never say what it does

Launch marketing is a specialist discipline. Solo builders can't buy it.

The solution

LaunchCopilot performs the launch — not advice about it.

Paste a store link. One flow runs — Grade, Generate, Validate — and hands back a launch kit you can post today.

28-rule ASO grade A/B rewrites PH · social · community · press
launchcopilot.edycu.dev
LaunchCopilot — paste a store link, auto-fill, and grade
How it works

The AI is graded by the same judge as the human.

01 · GRADE
28-rule lint

Deterministic. No AI — pure encoded store rules. Scores 0–100 with per-field fixes.

→
02 · GENERATE
5 parallel writers

Claude builds an AppProfile, then schema-constrained generators draft the kit.

→
03 · VALIDATE ↺
Score ≥ 90 or regenerate

The rewrite is fed back through the same lint engine. Below 90 → auto-repair with findings injected (max 2).

↺

Then a persona panel — synthesized from your own listing — critiques every artifact in character; weak ones regenerate once before you ever see them.

Live demo

Paste a link → graded → rewritten.

launchcopilot.edycu.dev
Auto-fill from a store link, then grade to C 76/100
1

Paste an App Store / Play link — every field fills, then grade.

launchcopilot.edycu.dev
Deterministic scorecard — the validated rewrite, re-graded to 93/100
2

Validated rewrite (34→93) + Product Hunt, social, community, press.

▶ Watch the full 60-second run — youtu.be/2if0sEFKe8c

What's under the hood
THE MOAT
Copy that's provably store-legal

A deterministic 28-rule engine grades the AI's output on every generation — not vibes, encoded App Store & Play rules. Rewrites that don't pass don't ship.

60-community dataset

Never suggests a sub that bans self-promo; knows Self-Promo Saturday.

Persona-panel QA

A focus group synthesized from your listing critiques every artifact.

Free API + CLI · no key

POST /api/analyze · npx aso-lint · 120+ tests · DEMO_MODE.

Why us

Everyone reports the score.
Nobody writes the fix.

ASO SUITES · $50–500/mo
Report numbers

Dashboards, no copy. You still write it.

AGENCIES · $2–5k
Write copy

Unaffordable at indie scale.

RAW CHATGPT · free
Breaks the rules

Writes copy that silently violates store limits.

LaunchCopilot is the only one that writes the copy and proves it's store-legal — because the same engine that grades you grades the AI.

Shipped for HackOnVibe
34 → 93

A demo listing, graded, rewritten, and re-validated by the same engine — automatically.

Free · 3 kits/day Pro · $12/mo Studio · $39/mo
Deterministic engine

120+ unit tests in CI

Benchmark

reproducible · sub-ms p95

Free API + CLI

/api/analyze · npx aso-lint

Runs with no key

DEMO_MODE=1

Live app + landing + auto-fill from any App Store / Google Play link
The ask

Grade your own app.

No signup, no API key. Paste any App Store or Google Play link and watch it score — then rewrite itself past 90.

api   POST /api/analyze

scan → grade your listing

Launch marketing, encoded.

01 / 10
← → navigate · ESC overview · P presenter · C contrast