Marketing
Featured
$9.99Apple Search Ads
Manages Apple Search Ads campaigns through asa-cli, including account discovery, keywords, negatives, reports, and campaign changes.
apple-search-adsasa-cli
Capabase
Turn subscriber outcomes, cancellation reasons, billing states, and product-use evidence into a prioritized app retention plan.
A churn dashboard can tell you that subscribers left without telling you what to fix. This skill starts with the denominator and the source systems. It separates voluntary exits from billing loss, keeps observed facts apart from hypotheses, and traces each priority back to evidence.
Use it to review a period-start subscriber cohort, cancellation exports, usage data, support themes, or an existing retention proposal. The package includes a deterministic CSV profiler, a cohort schema, and a decision-ready plan template. It does not invent benchmarks, promise a lift, or treat offer acceptance as proof of durable retention.
Input
Analyze this Q2 period-start cohort for our fitness app. It includes subscriber outcome, plan, tenure, acquisition channel, cancellation reason, core actions, and revenue cents. Tell us what to fix first without treating small segments as causal proof.
Output
A validated cohort profile, denominator-bound rates, ranked possible causes with confidence, one prioritized intervention, required measurement, and limitations for small or incomplete segments.
Input
Our dashboard combines cancellations, failed renewals, grace-period recoveries, and expired entitlements. Review these exports and produce a plan for the next release.
Output
Provide a period-start subscriber cohort CSV or a bounded set of cancellation, billing, product-use, and support evidence. State the app, review period, billing source of truth, and decision you need to make. Ask the agent to profile the evidence, rank likely causes, produce an intervention backlog, and define one bounded experiment or the data repair needed first.
A period-start cohort CSV, cancellation export, billing report, product-use data, support themes, or an existing retention proposal.
The app, platform, plans, review period, billing source of truth, and retention decision the team must make.
Required only for the bundled deterministic cohort profiler. No third-party packages are needed.
SKILL.md; deterministic subscriber-cohort profiler; cohort input schema; evidence-backed retention plan template
No reviews yet.
A reconciled state map, coverage gaps, separate voluntary and involuntary loss findings, a prioritized intervention backlog, and source-of-truth checks before release.
Input
The team wants to offer 30% off to everyone who taps cancel. We have cancellation reasons and 90-day re-churn data but no experiment assignment event. Check the plan and tell us what can be tested safely.
Output
An evidence review that rejects unsupported blanket targeting, identifies the eligible segment and missing assignment event, and specifies one bounded experiment with outcome and harm guardrails.
Updated Jul 13, 2026