Code quality stack — Dependabot + biome + CodeRabbit + Sonarcloud

decision Sat Jun 20 2026 00:00:00 GMT+0000 (Coordinated Universal Time) decisionsprocesscode-qualitycioss

Decision

Adopt a 4-layer code-quality stack across every repo in the chirag127/oriz family:

Dependabot — automated dependency / security update PRs (GitHub-native).
biome — lint + format on every commit and in CI (already in place).
CodeRabbit — AI code review on every pull request (free for OSS).
Sonarcloud — deeper static analysis (SAST, code smells, complexity, duplication, coverage) on merge to main (free for OSS).

Each layer catches a class of issues the others miss:

Dependabot = supply chain (CVEs)
biome = style + obvious bugs (deterministic, fast)
CodeRabbit = logic + design + security smells (LLM-grade review of intent)
Sonarcloud = data-flow + cyclomatic complexity + duplication (whole-codebase static analysis)

All four are free forever for the family's public OSS repos, fitting the no-paid-tier rule.

Each repo's .github/dependabot.yml enables weekly version updates for npm (security updates are on by default GitHub-wide).
Each repo's .github/workflows/ci.yml continues to run biome locally; this rule keeps repos working independently per repos-work-independently.
The CodeRabbit GitHub App is installed at the org level, no per-repo setup needed.
Sonarcloud is wired through a .github/workflows/sonar.yml per repo plus a Sonarcloud project per repo. Adds one more PR check.
PR flow: open ? biome runs ? CodeRabbit comments ? reviewer merges ? Sonarcloud analyses ? quality gate visible on dashboard.
Dependabot's PRs flow through the same checks. CodeRabbit reviews dependency PRs the same way it reviews human PRs.