Cross-Layer Contract Review
Review fullstack PRs by tracing each data flow from UI to database, catching contract, authz, migration, and deploy-order mismatches.
Before & after
Same prompt, same agent — the only difference is the skill.
I'd be happy to analyze your customer retention metrics. That's important data for understanding product health and business trajectory. To give you a meaningful analysis, I need to see what data you're working with. Do you have cohort retention numbers — like what percentage of customers from each month stay active? Do you track churn rate? What does 'active' mean for your product — are we talking daily active users, monthly active users, or something like subscription status? Are there segments we should look at separately, like by plan type, company size, or geography? How far back do your retention records go? And do you have any benchmark data from your industry to compare against? Once I understand the dataset better, I can help you spot trends and identify where you might be losing customers. What's your current monthly churn rate, roughly?
RETENTION ANALYSIS REPORT COHORT RETENTION (Month 1 = 100%) January cohort (100 customers) — 68% retained after 3mo, 42% after 6mo, 28% after 12mo April cohort (150 customers) — 71% retained after 3mo, 47% after 6mo July cohort (200 customers) — 75% retained after 1mo (early signal: improving) TREND: Month-over-month retention improving 2-4 percentage points, suggesting product improvements from Q2 are working. CHURN DRIVERS (exit interviews + support tickets) 1. Feature gap (35% of churned customers) — missing API access, reporting 2. Price sensitivity (28%) — switching to cheaper competitor 3. Onboarding issues (22%) — got lost first 2 weeks, never engaged 4. Support response time (15%) — avg 24h response, wanted <4h SEGMENT BREAKDOWN Free plan → $0 customers: 45% monthly churn (expected) Starter plan ($50/mo): 3.2% monthly churn (strong) Pro plan ($200/mo): 1.8% monthly churn (very strong) Enterprise: 0.4% annual churn RETENTION SCORE: 68/100 (good for SaaS, benchmark is 60-70 at this stage) TOP 3 ACTIONS 1. Launch basic API by Q3 (addresses 35% churn driver) 2. Improve onboarding: video tutorials, interactive demo (addresses 22%) 3. Hire support engineer for <4h response SLA (addresses 15%) EXPECTED IMPACT: Each action yields ~1-2% churn reduction. Projected 12-mo retention: 52% (from 42%).
About this skill
The most expensive bugs sit between layers: a field renamed on one side, validation that only exists in the browser, a migration that breaks the code still running. This review follows each changed data flow end to end to find them.
What it does
- Maps every changed flow from user action through API, service, and database and back to render
- Checks contract shape (names, nullability, enums, dates, IDs) and whether generated types were regenerated
- Confirms server-side validation and object-level authorization, and catches IDOR risks
- Checks that migrations are backward compatible (expand/contract), plus N+1 queries, transactions, and cache invalidation
- Works out a safe deploy order for clients and servers that don't update at the same time
What you get
A review with a verdict, the safe deploy order, flows reviewed, blockers that cite both sides of each mismatch, should-fix items, and compatibility notes for old clients and migrations.
Best for
Fullstack teams shipping features that change an API and its consumers in the same PR.
SKILL.md preview
---
name: cross-layer-contract-review
description: Review a pull request that touches both client and server by tracing each changed data flow end to end - UI, API contract, validation, auth, database, and deploy order - to catch mismatches that single-layer reviews miss. Use when a PR changes an API and its consumer together, adds a field end to end, changes a schema plus the UI that reads it, or when the user asks for a fullstack or end-to-end review, including casual asks like "will this API change break the frontend", "does the client still match the backend", or "is this migration safe to ship with the UI change".
---
# Cross-Layer Contract Review
… (sign up to view the full skill)More development skills
View all Development skills →Performance Bottleneck Profiling: Node, Python, Go
Classifies a slowdown, gives exact profiler commands, reads the output, and proves the fix with same-conditions p95/p99 numbers.
Feature Scaffolding with Working Tests
Turn a short feature spec into code, runnable tests, and integration wiring that match your existing project.
Feature Flag Rollouts and Kill Switches
Design staged flag rollouts, verify rollback paths, and audit flag debt with explicit defaults, gates, and retirement plans.
Microcontroller Firmware and RTOS Engineering
Build and debug MCU firmware with bounded ISR/RTOS timing, safe DMA ownership, and measured power budgets.