15 commands available

Content, growth,
and product intel
for RevenueCat

An AI agent that pulls live data from RevenueCat's GitHub repos and blog, then generates actionable content, experiment designs, and SDK health reports.

rcagent — zsh
$
15
CLI Commands
5
SDK Repos Monitored
3
Content Pillars
50+
Issues Analyzed
C Module 01

Create content that
developers actually read

Blog posts, social threads, changelogs, and competitor comparison pages. Grounded in RevenueCat's actual SDK APIs, content pillars, and live blog trends.

Content Ideas AI + RSS

$ rcagent content ideas --pillar growth

Pulls current blog posts via RSS, cross-references with known pain points, and generates original ideas with SEO keywords and impact ratings.

Blog Posts 2000+ words

$ rcagent content blog "Paywall A/B Testing"

Full draft with real SDK code examples (Swift, Kotlin, React Native), TL;DR, structured headers, and CTAs to RevenueCat docs.

Social Posts Multi-platform

$ rcagent content social "topic" --platform twitter

Platform-native content for Twitter threads, LinkedIn, Reddit, and dev.to. Tone-aware: educational, provocative, casual, or data-driven.

Comparison Pages SEO

$ rcagent content compare --competitor Adapty

Honest competitor comparison pages with feature tables, migration guides, decision trees, and community sentiment. Built for /compare/ SEO.

output/content/ideas.md Sample Output

1. The Web-to-App Funnel Playbook

AttributeValue
PillarGrowth
FormatTactical guide + architecture diagrams
TargetIndie devs seeking a second acquisition channel
Key angleWeb paywall as first-touch monetization via RC Web Billing + Stripe
SEO keywordsweb to app funnel, subscription app acquisition, web paywall
ImpactHIGH — high-intent, low competition

2. The 72-Hour Window: Win-Back Automation

Maps RevenueCat lifecycle webhooks (CANCELLATION, BILLING_ISSUE, EXPIRATION) to intervention strategies. Shows webhook → Braze integration with Liquid templates, plus a lightweight n8n alternative for indie devs.

3. Subscription Pricing in 2026: What A/B Data Says

Contrarian hook: higher prices don't reliably hurt conversion — but presentation almost always does. Covers trial length, price anchoring, lifetime-as-decoy pattern with RC Experiments setup.

G Module 02

Design experiments
that move revenue

Hypothesis-driven experiment briefs with ICE scoring, sample size calculations, kill criteria, and variant designs. Powered by RevenueCat's Experiments and Offerings APIs.

Experiment Design Full Brief

$ rcagent growth experiment "onboarding"

Complete experiment brief: hypothesis, ICE score, target audience, implementation plan, metrics with MDE calculations, risks, and A/B variants.

Experiment Backlog Prioritized

$ rcagent growth backlog --focus activation

Ranked backlog of 10 experiments sorted by ICE score. Each with hypothesis, funnel stage, effort estimate, and expected impact.

Competitive Analysis Strategic

$ rcagent growth competitive --competitor Adapty

Positioning gaps, switching triggers, content opportunities, and prioritized actions with ICE scores vs. Adapty, Qonversion, Superwall.

Paywall Experiments A/B Test

$ rcagent growth paywall --metric trial_start_rate

Paywall A/B test designs using RC Experiments + Offerings. Variant layouts, copy changes, pricing anchoring, with implementation steps.

output/growth/experiment-onboarding.md Sample Output

Experiment: "First Purchase in 15 Minutes"

DimensionScoreRationale
Impact8Directly improves north-star activation metric
Confidence8Based on onboarding friction literature
Ease6Backend provisioning + dashboard UI work
ICE Total7.3High-priority — build queue immediately

Metrics & Success Criteria

Primary: 7-day activation rate (target: +7.5pp, from 25% → 32.5%)

Sample size: n ≈ 680 per variant • Runtime: 21 days minimum

Kill criteria: SRM >5% divergence, 14-day real purchase < control by 3pp

Variants

CONTROL Current empty-state dashboard with static docs links

VARIANT A "Launchpad Lite" — personalization quiz + guided checklist + inline code snippets

VARIANT B "Launchpad Full" — everything in A + in-browser Purchase Simulator with celebration moment

F Module 03

Surface product intel
from real signals

Pulls open issues from 5 SDK repos, analyzes GitHub discussions, and synthesizes severity-graded health reports with actionable recommendations.

SDK Health Report Live Data

$ rcagent feedback health

Pulls 50+ issues across purchases-ios, purchases-android, react-native-purchases, purchases-flutter, purchases-kmp. Severity-graded P0-P3.

DX Audit Scored

$ rcagent feedback dx-audit --platform iOS

Scores developer experience 1-10 across 5 dimensions: getting started, API usability, advanced features, maintenance burden, community.

Feature Gaps vs Competitor

$ rcagent feedback gaps --competitor Adapty

Feature parity analysis, quality gaps, developer preference drivers, and recommended investments. Each gap rated by severity and strategic importance.

Sentiment Report Community

$ rcagent feedback sentiment

Synthesizes developer sentiment from most-discussed issues. Churn risk factors, competitive threats, underserved segments, and prioritized actions.

output/feedback/sdk-health.md Sample Output

SDK Health Dashboard

RepositoryStatusCritical Issues
purchases-iosNeeds AttentionDeadlock on restore (100% repro)
purchases-androidNeeds AttentionANRs on TextureVideoView
react-native-purchasesCriticalNew Arch broken + active regression
purchases-flutterCriticalGDPR violation + restore broken
purchases-kmpNeeds AttentionSeveral P1s, growing pains

Top Issues by Impact

P0 RN #1158 — PaywallFooterContainerView broken with New Architecture (14 months open, 25 comments)

P0 Flutter #1639 — Stripe.js auto-injected at startup, GDPR violation for all EU-facing apps

P0 iOS #5729 — Deadlock with RC Paywalls Restore (100% reproduction rate)

P0 RN #1644 — Paywall empty after 9.10.3 → 9.10.5 upgrade (hot regression)

P0 Android #3142 — Multiple ANRs on TextureVideoView (Play Store vitals risk)

> How It Works

Live data in, intelligence out

The agent pulls fresh data from public APIs, injects it into domain-specific system prompts, and routes through your Claude subscription.

Data Sources

GitHub Issues • Releases
Blog RSS • Discussions

→

rcagent

Domain prompts
Research module • CLI

→

Claude

Your subscription
via claude -p

$ Reference

All 15 commands

Every command auto-saves output to output/ with timestamps.

rcagent --help CLI Reference
CommandDescription
content ideasGenerate content ideas from blog trends + pain points
content blogWrite a full blog post with SDK code examples
content socialTwitter threads, LinkedIn, Reddit, dev.to posts
content compareHonest competitor comparison pages for SEO
content changelogDeveloper-friendly changelog from GitHub releases
growth experimentFull experiment brief with ICE scoring and variants
growth backlogPrioritized backlog of experiments by ICE score
growth competitiveCompetitive landscape analysis + strategic actions
growth paywallPaywall A/B test design with RC Experiments
growth funnelAcquisition funnel audit with improvement recs
feedback healthSDK health report from 50+ GitHub issues
feedback dx-auditDeveloper experience audit scored 1-10
feedback gapsFeature gap analysis vs a competitor
feedback sentimentCommunity sentiment synthesis
dashboardFull daily briefing: health + ideas + experiments

Start your first briefing

Clone the repo and run the dashboard in under a minute.

$ git clone && pip install -e . && rcagent dashboard