JPeetz/agent-skills / skills/business-strategy/app-discovery-scrutiny
app-discovery-scrutiny
VerifiedUse this skill when the user wants to evaluate a mobile app idea for commercial viability — whether it's a clone opportunity, a market gap, or a new launch. Produces a harsh "Build / Pivot / Kill" verdict backed by competitive analysis and revenue modeling. Use when the user says things like "Is this app idea worth building?" "Should I clone this?" "Analyze this niche" "Evaluate this competitor" even if they don't use the word "scrutiny" or "dominance report." Mobile apps only — do not activate for CLI tools, SDKs, or SaaS.
| File | Kind | Size |
|---|---|---|
| CHANGELOG.md | markdown | 579 B |
| SKILL.md | skill md | 3.3 KB |
| evals/evals.json | config | 2.2 KB |
| references/corrections.md | markdown | 1.5 KB |
| scripts/score.pyexecutable | script | 2.2 KB |
references/corrections.md
View on GitHub# Scrutiny Corrections Log
_Real-world failures that shaped this skill. AgentForge test runs 2026-05-24._
## Correction 1: CLI Dev Tools Mistaken for Mobile Apps
**Problem:** First test run picked Forge (CLI guardrail framework, 676 HN points) as winner. Joerg: "And that's a mobile app?"
**Root cause:** Platform-agnostic scoring. HN Show HN overweights dev tools.
**Fix applied:** Added hard scope gate — mobile apps only. Auto-reject: CLI tools, SDKs, libraries, desktop-only, Chrome extensions without mobile companion, infrastructure/dev tools. Added scoring note: "Mobile apps only. If a candidate isn't clearly a mobile app, eliminate before scoring."
## Correction 2: Kill-Floor Kills Mobile Apps on SEO
**Problem:** ShadowCat (QR file transfer) and Pablo (Chrome extension) killed on SEO even though their real discovery channel was Chrome Web Store / direct virality.
**Fix applied:** SEO → "SEO/ASO Opportunity." Score the HIGHER of ASO (for mobile apps) or SEO (for web apps). Never kill a mobile app on web SEO when App Store optimization is its real channel.
## Correction 3: Research-Without-Verification
**Problem:** First reports used inference-based competitor data because SearXNG was offline. Data was plausible but unverified.
**Fix applied:** Tavily web_search is now primary. Research phase must use live web_search/web_fetch. If research fails, flag it explicitly rather than inferring silently. "No winner today" is valid if data quality prevents scoring.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |