docs(clone-app): add stack/effort/infra/report reference guides
This commit is contained in:
fatih.bulut
parent
10eb027efa
commit
f88da4d2d9
|
|
@ -0,0 +1,34 @@
|
|||
# Effort Estimation Guide (Phase 5) — AI-Assisted Units
|
||||
|
||||
Estimate in **AI Sprints**, NOT human days. 1 AI Sprint = one focused Claude
|
||||
Code session producing a reviewable increment (~2-4h human review time).
|
||||
|
||||
## Method
|
||||
1. Build the feature list from RE output:
|
||||
- Screens = Activity + Fragment count (dedupe obvious base classes).
|
||||
- API surface = endpoint count from find-api-calls.
|
||||
- Integrations = third-party SDKs (auth, payment, maps, analytics, push, chat).
|
||||
- Backend = REST/GraphQL/WebSocket presence + first-party host count.
|
||||
2. Assign each feature a complexity and sprint range using the table below.
|
||||
3. Sum to a range (min–max sprints). Never give a single false-precise number.
|
||||
|
||||
## Reference sprint costs
|
||||
| Feature class | Low | Typical | High | Notes |
|
||||
|---|---|---|---|---|
|
||||
| Project scaffold + CI | 0.5 | 1 | 1 | nav, theming, state mgmt setup |
|
||||
| Auth (email/social) | 1 | 1.5 | 2 | +1 if custom backend auth |
|
||||
| Simple list/detail screen | 0.3 | 0.5 | 1 | per screen, AI-fast |
|
||||
| Complex interactive screen | 1 | 2 | 3 | maps, editors, realtime |
|
||||
| API integration layer | 1 | 2 | 3 | scales with endpoint count |
|
||||
| Each major SDK integration | 0.5 | 1 | 2 | payment > maps > analytics |
|
||||
| Custom backend (CRUD) | 2 | 4 | 8 | scales with entity + endpoint count |
|
||||
| Realtime (WebSocket/push) | 1 | 2 | 4 | |
|
||||
| Offline/sync | 2 | 3 | 5 | |
|
||||
| Polish/QA/store submission | 1 | 2 | 3 | |
|
||||
|
||||
## Obfuscation caveat
|
||||
If RE flagged heavy R8/Flutter (feature list incomplete), add a **+20-40%
|
||||
uncertainty band** and state it explicitly in the report.
|
||||
|
||||
## Output
|
||||
A table: Category | Complexity | Sprints (min–max), then a TOTAL row.
|
||||
|
|
@ -0,0 +1,24 @@
|
|||
# Infrastructure Cost Guide (Phase 5)
|
||||
|
||||
Estimate **monthly USD** at three scales: MVP (<1k MAU), Growth (~50k MAU),
|
||||
Scale (~500k MAU). Use round heuristics; mark assumptions.
|
||||
|
||||
## Components
|
||||
| Component | MVP | Growth | Scale | Notes |
|
||||
|---|---|---|---|---|
|
||||
| App hosting/backend | $0-20 | $50-200 | $500-2k | Railway/Render/Fly → AWS/GCP |
|
||||
| Managed DB (Postgres) | $0-15 | $50-150 | $300-1k | Supabase/Neon/RDS |
|
||||
| Object storage/CDN | $0-5 | $20-80 | $200-800 | S3+CloudFront/Cloudflare |
|
||||
| Auth (managed) | $0 | $0-100 | $200-500 | Supabase/Auth0 tiers |
|
||||
| Push notifications | $0 | $0-50 | $50-300 | FCM free; OneSignal paid tiers |
|
||||
| Third-party APIs | varies | varies | varies | maps, SMS, payments % |
|
||||
| Monitoring | $0 | $20-50 | $100-300 | Sentry/Datadog |
|
||||
|
||||
## Third-party cost flags from RE
|
||||
- Maps SDK → Google Maps billing after free tier.
|
||||
- Payment SDK → % per transaction (Stripe ~2.9%+30¢).
|
||||
- SMS/OTP → per-message (Twilio).
|
||||
List each detected paid dependency explicitly.
|
||||
|
||||
## Output
|
||||
A 3-column (MVP/Growth/Scale) monthly cost table + a one-line "biggest cost driver".
|
||||
|
|
@ -0,0 +1,51 @@
|
|||
# Clone Feasibility Report — {APP_TITLE}
|
||||
|
||||
**Package:** {PACKAGE}
|
||||
**Date:** {DATE}
|
||||
**Analyzed by:** clone-app skill
|
||||
|
||||
## 1. App Overview
|
||||
- Title / developer / category / installs / rating ({RATING} from {RATING_COUNT} reviews)
|
||||
- iOS App Store presence: {YES/NO + link}
|
||||
- One-paragraph description of what the app does.
|
||||
|
||||
## 2. Tech Stack (Detected)
|
||||
- Mobile framework: {framework} (RE fingerprint marker: {marker})
|
||||
- HTTP stack: {Retrofit/Ktor/...}
|
||||
- Backend signals: {first-party hosts, REST/GraphQL/WS}
|
||||
- Notable SDKs: {list}
|
||||
- Obfuscation level: {low/med/high} — analysis completeness caveat if high.
|
||||
|
||||
## 3. Recommended Clone Stack
|
||||
- Selected by user: {stack}
|
||||
- Rationale: {1-2 lines}
|
||||
|
||||
## 4. Feature List (from APK)
|
||||
- Screens: {n}
|
||||
- API endpoints: {n} (key ones listed)
|
||||
- Integrations: {list}
|
||||
- Backend required: {yes/no + why}
|
||||
|
||||
## 5. Effort Estimate (AI-Assisted)
|
||||
{effort table from effort-estimation-guide}
|
||||
**Total: {min}–{max} AI Sprints** (1 sprint ≈ one focused Claude Code session).
|
||||
Uncertainty band: {±%} due to {reason}.
|
||||
|
||||
## 6. Infrastructure Cost Estimate (monthly)
|
||||
{MVP/Growth/Scale table from infra-cost-guide}
|
||||
Biggest cost driver: {item}.
|
||||
|
||||
## 7. Market Analysis
|
||||
- Current app metrics: installs {x}, rating {y}, review velocity {if known}.
|
||||
- Competitor landscape: {2-4 named competitors}.
|
||||
- Target market size: {rough estimate + basis}.
|
||||
- Differentiation opportunities: {list}.
|
||||
|
||||
## 8. Viability Verdict
|
||||
**{GO / CONDITIONAL GO / NO GO}**
|
||||
- Key risks: {list}
|
||||
- Key opportunities: {list}
|
||||
- Recommendation rationale: {2-3 sentences tying effort+cost vs market}.
|
||||
|
||||
## 9. Next Step
|
||||
{Link to implementation plan if user proceeds, else "report only".}
|
||||
|
|
@ -0,0 +1,23 @@
|
|||
# Stack Recommendation Guide (Phase 4)
|
||||
|
||||
Goal: present the user 2-3 concrete stack options for building the clone, then
|
||||
let them pick. The pick locks effort + cost math in Phase 5.
|
||||
|
||||
## Inputs you have
|
||||
- Detected original framework (from RE fingerprint): Native Kotlin/Java, Flutter, React Native, etc.
|
||||
- Detected HTTP stack + backend signals (Retrofit/Ktor/Apollo/GraphQL/WebSocket).
|
||||
- Feature surface (screen count, SDKs, permissions).
|
||||
|
||||
## How to choose the 2-3 options
|
||||
Always include:
|
||||
1. **Fastest for AI-assisted dev** — default to **Flutter** (single codebase, strong AI codegen, fast UI) unless the app is heavily native-platform-dependent.
|
||||
2. **JS-ecosystem option** — **React Native + Expo** when the team is JS-leaning or web reuse matters.
|
||||
3. **Match-the-original** — only when the original's nativeness is essential (deep platform APIs, AR, heavy native SDKs). Note the higher effort.
|
||||
|
||||
## Backend
|
||||
- If RE shows first-party API hosts → a backend is required. Recommend **Node/TS (NestJS or Express)** or **Supabase** (fastest with AI) unless GraphQL detected → recommend **Apollo Server** or **Hasura**.
|
||||
- If only third-party hosts (Firebase, payment SDKs) → likely **BaaS / no custom backend**; note it.
|
||||
|
||||
## Output format to the user
|
||||
Present a short table: Option | Mobile stack | Backend | Why | Relative effort (Low/Med/High).
|
||||
Then ask: "Which stack should I base the effort + cost estimate on?"
|
||||
Loading…
Reference in New Issue