Technical Debt Assessment Framework

Monolith with no clear bounded contexts, spaghetti dependencies, no service boundaries possible without full rewrite.

Some modularisation but significant circular dependencies. Major features require changes across 5+ components.

Mostly clean architecture with identifiable service boundaries. Occasional cross-cutting concerns.

Clear bounded contexts, minimal coupling. New features added in under 2 components in most cases.

Exemplary architecture. SOLID principles applied consistently. Easy to decompose further if needed.

Widespread code duplication, deeply nested logic, long functions (500+ lines), no consistent naming.

Significant duplication, mixed naming conventions, average function length over 100 lines.

Some duplication in hotspots. Most functions under 50 lines. Naming mostly consistent.

Occasional minor duplication. Good naming. Cyclomatic complexity within acceptable bounds.

DRY throughout, excellent naming, linting enforced in CI, code review culture strong.

Under 10% coverage. No automated testing enforced. Deploys require manual verification.

10-30% coverage, unit tests only, no integration or end-to-end tests.

30-60% coverage with unit and some integration tests. Critical paths tested.

60-80% coverage with full testing pyramid. E2E tests on critical user journeys.

80%+ coverage, TDD culture, mutation testing, full CI enforcement.

Multiple CVEs outstanding, dependencies 3+ major versions behind, unmaintained packages in critical paths.

Dependencies 1-2 major versions behind, some known vulnerabilities, no automated update process.

Mostly current dependencies. Dependabot or equivalent in place. Patches applied within 30 days.

Dependencies current, minor version updates applied regularly, no known CVEs.

Automated dependency updates with rollback. Zero known CVEs. Minimal dependency surface area.

No documentation. Critical knowledge in heads of 1-2 engineers. Bus factor of 1.

Some README files, but architecture and decision rationale undocumented. Onboarding takes 3+ months.

Core systems documented. ADRs present for some major decisions. Onboarding takes 4-6 weeks.

Comprehensive runbooks, ADRs, and API docs. New engineers productive within 2-3 weeks.

Self-documenting code, living documentation, automated API docs, thorough onboarding guide.

Manual deployments, no infrastructure as code, environments differ, frequent deploy-caused incidents.

Partial automation, some IaC, environments partially consistent, deploys risky.

Most infrastructure in IaC, automated deployments with rollback capability.

Full IaC, immutable infrastructure, blue-green or canary deploys, monitoring-driven rollback.

GitOps, policy as code, self-healing infrastructure, full observability stack.