Review a subscription paywall against the actual trigger, offer terms, user journey, and purchase path before you ship or test it.
A paywall can look polished and still ask at the wrong moment, misstate a trial, or test five ideas at once. This skill starts with the configured offer and the user's trigger. It separates facts from assumptions, flags release blockers, preserves the real transaction terms in rewrites, and returns one experiment your team can instrument.
Use it for trial starts, trial endings, feature gates, usage limits, introductory offers, and plan upgrades. The review does not promise a lift or declare legal compliance. Missing evidence stays visible as an unknown.
Input
Review this annual-plan paywall. App Store Connect says $39.99 per year after a seven-day trial. The screenshot, purchase states, and restore flow are attached.
Output
A frozen offer-fidelity table, evidence map, release blockers, prioritized screen changes, a term-faithful rewrite, and verification steps for purchase and restore states.
Input
This paywall appears when a free user exports a second project. We track exposure, purchase start, and purchase completion. Review it and choose one variable to test.
Output
A trigger-aware conversion brief with facts separated from assumptions, the strongest supported finding, and one experiment with control, treatment, primary metric, guardrails, and event gaps.
Provide a paywall screenshot, exported copy, or component code. Add the trigger, user journey, configured plans, prices, trial and renewal terms, and any funnel events you have. Ask the agent to review the screen, preserve the verified offer in any rewrite, and return one bounded experiment.
A screenshot, screen recording, exported copy, or relevant component code.
Known plan names, prices, billing periods, trial terms, renewal behavior, and paid entitlements.
The trigger, the user's immediate goal, what they have experienced, and what happens after dismissal.
SKILL.md; evidence-first review workflow; paywall review rubric; conversion brief template; one-variable experiment format
No reviews yet.
Input
Rewrite this upgrade screen, but keep the monthly and annual products exactly as configured. We do not have testimonials or baseline conversion data.
Output
Reordered screen content and final copy that preserves the supplied transaction terms, uses no invented proof, and leaves unknown sizing inputs explicit.
Automated scan passed. It found 1 finding; highest severity info.
Scanned Jul 13, 2026.
Automated scanning reduces risk but cannot guarantee a package is safe for every environment.
