Research &
prototype notes.
Where the examples come from, what this prototype does, and what still needs verification.
How the example library was selected
The 20 examples are an educational selection spanning popular mobile use cases and useful learning scenarios. They are not a ranked top 20, a market forecast, or a promise of commercial success. Apple and Google use different category labels; download charts, app counts, spending, and engagement measure different things. AI is a product type that can span multiple store categories.
The library adapts these ideas to browser-first planning, with Basic, Moderate, and Complex starting scopes. It does not reproduce the listed commercial apps or imply their endorsement.
- Apple: US top apps of 2025 ↗ — examples of popular apps, rather than a combined platform category ranking.
- Sensor Tower: State of Mobile 2026 ↗ — cross-platform trends based on its own methodology and taxonomy.
- AppBrain: Google Play category statistics ↗ — category inventory and download-threshold statistics; these are not the same as total category downloads.
These sources may update. The library is a curated snapshot, not a live ranking feed. Loaded example projects remain independent copies.
What works in this prototype
Local project saving, Basic and full planning paths, option guidance, glossary search, 20 three-tier worked examples, review, full/compact prompt exports, JSON backup/import, recoverable Delete and Trash, the user manual, and reading modes.
No AI service is called by the planner. The exported prompt is a planning input for an external tool you choose. Referenced screenshots and files must be supplied separately.
Important limits
Project data stays in this browser and site origin. Application login and cloud backup/sync are not implemented. Hosting sign-in does not back up projects. Browser data may be cleared or lost; retain JSON backups. Guaranteed offline startup, automatic app generation, attachment upload, and shared editing are not included.
Consistency checks cover selected rules, not every possible contradiction. A reviewed section or recorded answer is not proof of completeness, accessibility, security, legal compliance, or production readiness. Examples use fictional data and simulated external actions until real integrations and their rules are confirmed.
Validation and maintenance
Automated checks cover planning/export semantics, navigation, project recovery helpers, example variants, import compatibility, and selected documentation requirements. Browser checks cover selected user journeys and visual changes. A complete real-device, screen-reader, security, or legal audit has not been completed.
The app and manual use matching release versions. Manual lists and counts derive from the catalog; other instructions require review when behavior changes. A repository maintenance checklist and changelog guide each release. This is a development process, not unattended monitoring.
AI results and responsibility
AI results vary and are not guaranteed. Independently review and test outputs before relying on them. The app’s footer includes the full AI results & disclaimer notice. Its liability wording is a working draft for qualified legal review; it cannot guarantee elimination of all liability.