AI Chaos Index: 23/100
STABLE
After full Cap & Grow remediation: 3 domains extracted, CI/CD pipeline active, tests across all domains. The same app — now structurally ready to scale and safe to hire for.
Total delta: -64 points. From Critical to Stable. Same codebase, same functionality.
The Complete Transformation
| Metric | Chaos | Rescue | Clean |
|---|---|---|---|
| ACI Score | 87 | 58 | 23 |
| Total files | 8 | 32 | 48 |
| Largest frontend file | 1,084 LOC | 174 LOC | 537 LOC |
| Domains extracted | 0 | 1 | 3 |
| Test files | 0 | 2 | 6 |
| Test coverage | 0% | ~15-20% | ~40-50% |
| CI/CD Pipeline | None | None | Lint + Test + Build |
| Architecture docs | None | Partial | Full (slice specs) |
| .env.example | Missing | Missing | Present |
| Legacy code | — | Isolated | Removed |
Root Cause Severity Breakdown
3 fully extracted domains (billing, usage, user). Clean FastAPI entry point with domain routing.
Clear domain boundaries. shared/ for cross-cutting concerns. Own schemas per domain.
README, consistent naming, slice specs, .env.example. Near-perfect structural hygiene.
6 test files across all 3 domains. ~40-50% coverage. Frontend components still untested.
CI/CD pipeline with lint, test, build. Runs on push and PR. Broken code cannot merge.
Remaining Findings
A score of 23 is Stable — not perfect. Here's what's left.
Frontend Components Lack Tests
All 3 UI components (BillingWidget, UsageChart, ProfileCard) have no test files. Backend logic is well-tested but the frontend layer has zero coverage.
Risk: UI regressions will not be caught before deployment. As the frontend grows, this gap becomes costly.
Page.tsx Still Contains Orchestration Logic (537 LOC)
Individual domains are well-extracted, but main page.tsx still contains orchestration and layout logic that could be further decomposed.
Risk: As more domains are added, page.tsx will grow. Consider extracting a DashboardLayout component.
No Pre-commit Hooks
CI catches issues on push, but no pre-commit hooks catch lint errors locally before committing.
Risk: Minor: developers may push code that fails CI. Easy to add with husky.
Stable ≠ zero risk. A score of 23 means controlled complexity — not perfection. The codebase is architecturally ready to scale, safe to hire for, and safe to hand off. Every remaining issue is minor, well-documented, and addressable as part of normal development.
Download the Sample Report
See the full Quick Scan report for a well-structured codebase.
Download PDF Report (ACI 23)PDF • 5 pages • AI Chaos Index Quick Scan Report
Want this for your app?
The journey from 87 to 23 starts with a single scan. Know your score in 24 hours.