maxjcia@github ~/youth-sports-platform $ git log --stat

Engineering readout of a solo, AI-driven build

Youth-sports training platform — Expo mobile app, Next.js dashboard, Supabase backend. One operator driving Claude Code, 104 days, part-time. Every number below is measured from git + the GitHub Actions API, not estimated.

stack TypeScript mobile Expo SDK 54 · RN · Expo Router web Next.js 15 App Router backend Supabase · Postgres · Edge Fns mono Turborepo · pnpm window 2026-03-24 → 07-06
01

Throughput

104 days · part-time
996
merged PRs
9.6/day · 67/week
998 / 2,459
commits on main / all branches
+496K
lines added (ex lockfiles/gen)
−118K
removed · ~24% churn was rework
02

Test architecture

bars scaled to case count
Integration · vitest
1,190
Unit · mobile RN
705
Unit · shared/pkgs
612
Unit · web
552
E2E · Playwright
314
pgTAP DB asserts
3,150
Maestro flows (mobile)
45
~7,300 automated checks 9,285 expect() assertions 478 TS test files · 219 pgTAP files ~0.8:1 test:source (44% of all code is test)
03

CI pipeline

11,663 runs · 24 workflows · ~112/day
Stryker mutation tests CodeQL SAST Argos visual regression PR · tests + typecheck + E2E Preview gate (Vercel) Mobile E2E · Maestro Nightly QA agent · Playwright Nightly QA agent · Appium Reuse-decision tripwire Config-change verification Cloudinary real-API smoke Deploy · staging + prod +12 more
04

Backend & security surface

RLS on every table

Database

Migrations append-only ledger79
RLS policies434
SECURITY DEFINER RPCs329
Edge Functions (Deno)48

Commit intent

fix431
feat329
test / docs87 / 95
fix : feat ratio1.31
05

How it was kept honest

Org isolation runs on signed JWT claims read through a single validated accessor; row-level security enforces it at the database. Migrations are an append-only ledger — non-destructive, backward-compatible, ON DELETE RESTRICT on the core foreign keys, no hard-deletes of user data.

Every change went through a three-review gate before merge, plus a cross-model consult (Codex) on strategy before code. That cross-model pass is not ceremony: it repeatedly caught real bugs a single model's review missed — an access-control slip, a permission check that misfired on legitimate users, a data-freshness race. The 1.31 fix-to-feat ratio is the visible cost of that discipline: more commits correcting shipped code than adding new.

06

On the AI

The AI wrote most of the code. No point pretending otherwise, and any developer reading the diff would see it. What didn't automate was the part that makes generated code shippable: the architecture calls, the review gates, the adversarial verification, the security model, and the refactoring passes that deleted and rebuilt a quarter of the line-churn.

The honest claim isn't "one person out-typed a team." It's that a single operator, with the right harness around the model, held a real production codebase to a bar you can audit — and the git history is the audit.