High-fidelity extraction mode for clone-app: skips feasibility phases
(effort/cost/verdict), deepens RE to full Tier-2 + in-app logic + nav
graph + inferred backend design, adds a fidelity build-spec variant.
Phase A (static deep) only; dynamic analysis deferred to a follow-on.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add design for pixel-perfect clone capture: APK resource design tokens +
Play screenshots, Unity IL2CPP/Mono RE + AssetRipper game assets, and a
standalone Clone Build Spec artifact feeding writing-plans.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
AppBrain is Cloudflare-blocked (403) for free scraping. play.google.com's
own /store/apps/top + /store/apps/category HTML works (validated live 46/46).
Rework Task 1, fetch-play-charts.py CLI/output, and all downstream prose.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Approved spec for market-research v2: AppBrain Play charts + Play
resolver/saturation scripts + best-effort Google Trends + curated numeric
sources, numeric source-cited scoring, 1-2 verified Play links per candidate,
survivor-only enrichment for efficiency.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Six-task TDD plan for the market-research plugin: scaffold + marketplace
entry, fetch-charts.py (App Store RSS), history.py (non-repeat memory),
reference rubrics, 6-phase SKILL.md with clone-app handoff, test aggregator.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Umbrella spec for two new plugins (market-research, hermes-build) plus a
one-branch clone-app Phase 7 edit, chaining discover -> analyze -> build.
Decomposes into three implementation plans built in order.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Phase 2 rewrite: invoke android-reverse-engineering skill in an isolated
subagent (fallback to direct scripts), extract Tier-2 payloads for
auth/payment/core flows, persist a digest contract to $WORK/.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
fatih.bulut