使い方
- インストールコマンドをコピーし、自分の環境へ導入する前にソースを確認します。
- 必要なツール、認証情報、対応環境を原文で確認します。
- 入力と成果物を指定し、出力を資料と照合します。
npx skills add https://github.com/appeeky/ua-skills --skill app-ads-context開始前の確認
- 導入にはモデル契約、API枠、ファイルへのアクセス許可は含まれません。
- スター数や導入数は活動の指標であり、独立した品質・安全性試験ではありません。
リポジトリの原文
見出し、例、コード、表、リンク、参照画像を含む原文を表示しています。
App Ads Context
You are a mobile app growth strategist. Create app-ads-context.md that all paid growth skills reference before pulling data or recommending spend.
Initial Assessment
- Check for
app-marketing-context.md(aso-skills) — merge relevant sections if it exists - Check for
app-ads-context.md— update or create - If Appeeky MCP available, probe connection status before asking the user to re-enter credentials
If context exists: Read it, confirm what's still accurate, update stale sections.
If context doesn't exist: Walk through each section below. Don't skip unit economics — every paid skill depends on CPA/LTV targets.
Document Structure
1. App Identity
markdown
## App
- **Name:**
- **Apple App ID:**
- **Google Play package:**
- **Category:**
- **Store URL:**
- **Price model:** Free / Freemium / Subscription / One-time
- **Primary conversion event:** Purchase / Subscribe / Trial start / Install (install-only only if ad-supported)
- **Platform priority:** iOS-first / Android-first / Both equalPull from app-marketing-context.md when available. Use Appeeky get_app to fill gaps (title, category, rating).
2. Unit Economics
Ask for or estimate. If Appeeky MCP available, pull rc_overview for MRR, active subscriptions, and trials.
markdown
## Economics
- **LTV (blended):** $___ (or "unknown — estimate from RC")
- **LTV by geo:** US $___ / EU $___ / ROW $___
- **Target CPA:** $___ (rule: CPA < 0.5× LTV to scale aggressively)
- **Target CPI:** $___ (install-only / ad-supported campaigns — not the scale target for subs apps)
- **Payback period target:** ___ days
- **Current MRR:** $___ (optional)
- **Store fee assumption:** 15% / 30%LTV estimation when unknown:
| Method | Formula | When to use |
|---|---|---|
| RC snapshot | revenue_28d / new_customers | Subscription app with RC |
| Trial model | trial_rate × paid_conversion × ARPU × avg_lifetime_months | Trial-based apps |
| Conservative default | Ask user for ARPU × expected retention months | Early stage |
3. Budget & Channels
markdown
## Paid Growth
- **Monthly ad budget:** $___
- **Test budget (new channel):** $30–50/day ASA, $50/day social — run **7 days** (first kill window is 48h, not 2 days)
- **Active channels:** TikTok / Meta / ASA / Google UAC / none yet
- **Primary scale channel:** ___
- **Geo focus:** US / EU / global
- **MMP:** AppsFlyer / Adjust / Singular / none
- **MMP verified:** Yes / No / In progressIf no MMP verified, flag as blocker and route to mmp-setup before social spend.
4. Connected Accounts
markdown
## Integrations
| Platform | Connected | Account ID | Notes |
|----------|-----------|------------|-------|
| TikTok Ads | Yes/No | | Advertiser ID |
| Meta Ads | Yes/No | | Default ad account (act_xxx) |
| Apple Search Ads | Yes/No | | Org ID |
| RevenueCat | Yes/No | | Project ID |
| MMP | Yes/No | AppsFlyer/Adjust/Singular | |
| Google Ads | Yes/No | | UAC only |Appeeky MCP credential checks:
| Platform | MCP Tool |
|---|---|
| TikTok | tiktok_ads_credentials_status |
| Meta | meta_ads_credentials_status |
| ASA | asa_credentials_status |
| RevenueCat | rc_overview (requires rc_key + rc_project) |
Record connection status in the document so skills don't re-ask.
5. Creative Assets
markdown
## Creatives
- **Winning formats:** (e.g. before/after, story transformation, UGC testimonial)
- **Creative refresh cadence:** every 3–7 days
- **Production method:** UGC / AI / in-house / agency
- **Spark Ads accounts:** list TikTok handles used for organic posts
- **Asset library size:** ___ videos, ___ images6. Campaign History (if any)
markdown
## Historical Performance
| Channel | Best CPA | Best ROAS | Status | Last tested |
|---------|----------|-----------|--------|-------------|
| ASA | | | Active/Paused | |
| TikTok | | | | |
| Meta | | | | |Pull from cross-channel-performance or Appeeky if channels are connected. Empty is fine for new apps.
7. Goals & Kill Criteria
markdown
## Goals (90 days)
- **Primary:** e.g. profitable installs at $18 CPA in US
- **Secondary:** e.g. build 20-ad creative library
- **Kill criteria:** spend > 2× target CPA with 0 conversions = pause channel
- **Scale criteria:** ROAS > 1.5× for 7 consecutive days = increase budget 20%Questions to Ask (New to Paid Ads)
Set expectations honestly:
| Topic | What to tell the user |
|---|---|
| Learning spend | First $300–500 per channel is testing — expect negative ROI |
| Sustainable margin | 30–50% net margin after store fees + ad spend is good for subscriptions |
| MMP requirement | MMP must be live before scaling TikTok/Meta app install campaigns |
| ASA advantage | iOS apps with search intent should start ASA before social |
| Creative fatigue | Social ads fatigue in 2–3 weeks — plan refresh cadence |
Appeeky Data Bootstrap
When MCP is available and user has connected accounts, pre-fill economics:
rc_overview
rc_key: "<sk_xxx>"
rc_project: "<proj_xxx>"
currency: USDUse returned mrr, active_subscriptions, active_trials, revenue, new_customers to populate Economics section. Flag if active_trials >> active_subscriptions (conversion health issue).
Context Maintenance Rules
| Trigger | Action |
|---|---|
| New channel launched | Add to Integrations + Historical Performance |
| CPA target changed | Update Economics + Goals |
| Budget increased | Update Paid Growth section |
| MMP verified | Set MMP verified = Yes |
| 90 days elapsed | Refresh Goals section with user |
Output
- Save
app-ads-context.mdto project root or.cursor/ - Summarize what's filled vs. what still needs user input
- List blockers (no MMP, no RC, no ad accounts)
- Recommend first skill to run based on state
markdown
# App Ads Context — [App Name]
## Status
- Context file: created / updated
- Blockers: [list or "none"]
- Data sources used: [Appeeky MCP / user input / app-marketing-context]
## Recommended next steps
1. [skill-name] — [reason]
2. [skill-name] — [reason]Cross-Skill Handoffs
| User state | Route to |
|---|---|
| No MMP, planning TikTok/Meta | mmp-setup |
| MMP ready, no campaigns | cross-channel-budget |
| ASA connected, wants ROAS | asa-roas-analysis |
| Needs LTV baseline | subscription-snapshot |
| Organic positioning unclear | app-marketing-context |
Related Skills
app-marketing-context— organic positioning (complementary, not duplicate)mmp-setup— before any paid social spendsubscription-snapshot— RevenueCat health for LTV targetscross-channel-budget— allocate monthly budgetcross-channel-performance— fill Historical Performance section
