feat: normalize candidate application lifecycle

[seonghobae]

Buyer-visible gap

Orgmetra currently conflates candidate identity with opening-specific recruiting lifecycle through a candidate-level application status. This lane adds tenant-qualified durable application identity, bitemporal Job/optional-Position scope versions, and bitemporal application-stage history while keeping employer terminal outcomes in governed decision boundaries.

Current domain contract

candidate_application_record is the immutable durable anchor; candidate_application_record_version carries bitemporal Job/optional-Position scope; candidate_application_stage_record carries non-terminal recruiting stages. Same-tenant Candidate/Job/Position integrity and Position→Job scope are enforced. Raw stages exclude generic closed, bare withdrawn, and hired so governed terminal-decision/withdrawal evidence cannot be bypassed. Forced RLS is exercised with NOBYPASSRLS regressions. No dedicated-writer dependency repository or cross-service application-table SQL is introduced.

Repair lineage

A CodeRabbit finding correctly identified that docs/ERD.md used the too-strong candidate_application_record ||--|{ candidate_application_record_version cardinality although the authoritative schema permits a durable application anchor with zero scope versions. Regression 10335192ec910c1daa23ce88fd475f18cba81253 pinned the optional ||--o{ edge; 29691f265d100361e2d5a70024e02672384c2da9 repaired the ERD; f27136f5febea21fb9b3cc7f22d4b564484a52be resealed the exact changed workflow/ERD artifacts. Those valid deltas remain preserved.

Live authority

• exact head: 04c60a6d485c1af32973959c37c9133ae928f59b
• PR recorded base snapshot: develop@9e3e4847510e1e612b48474ba42b177b8ed824df
• current protected truth: develop@eb9757f8649aaad026a9865508d9aad50c1a7a4f
• live state: open · Draft · mergeable=false

The older body text that called this PR non-draft/mergeable is superseded. Its prior exact-source hosted snapshot included broad technical success but terminal OpenCode/Strix failures. Protected develop has also advanced materially, so neither that old check rollup nor mechanical mergeability can authorize integration now.

Stack discipline

PR #67 is a dependency-first child and remains on this parent's old branch snapshot. Process #66 first: non-force adopt the then-current protected develop, preserve the normalized-application delta, reconcile migrations/ERD/provenance with protected truth, reseal exact changed artifacts, and rerun every applicable exact-current-head local/central gate. Only after normal protected integration should #67 adopt protected truth and reacquire its own withdrawal evidence. The separate unresolved ADR-manifest provenance dependency owned by #51 must likewise be revalidated from protected truth rather than copied from a mutable branch.

Do not force-push, destructively rebase, self-approve, use routine administrator bypass, weaken a gate, transfer parent/predecessor evidence, or mutate a dedicated-writer dependency.

[coderabbitai]

📝 Walkthrough

Walkthrough

후보자 지원을 불변 앵커, bitemporal 범위 버전, 단계 이력으로 분리했다. 테넌트 격리와 변경·삭제·TRUNCATE 보호를 적용했다. PostgreSQL 계약 테스트와 CI 저장소 검증을 갱신했다.

Changes

후보자 지원 핵심

Layer / File(s)	Summary
지원서 앵커 및 범위 버전 모델 database/migrations/0014_candidate_application_core.sql, docs/DATA_MODEL.md, docs/ERD.md, docs/adr/0026-normalized-candidate-application.md	불변 candidate_application_record와 candidate_application_record_version을 분리했다. 범위 무결성, bitemporal 중복 방지, 단계 연결, RLS, 변경 및 TRUNCATE 보호를 추가했다.
지원서 계약 및 보안 회귀 테스트 tests/test_candidate_application_postgres.sh, tests/test_candidate_application_decision_boundary_postgres.sh, tests/test_candidate_application_rls_postgres.sh	지원서 정정, 앵커 보존, 단계 계보, 참조 무결성, 중복 기간, RLS 및 TRUNCATE 차단을 검증한다.
품질 워크플로 및 저장소 검증 연결 .github/workflows/candidate-application-quality.yml, scripts/foundation-contract-core.mjs, tests/foundation-contract.test.mjs, tests/validate_repository.py, manifest.json	PostgreSQL 16.14에서 계약 테스트를 실행한다. 필수 아티팩트, 데이터베이스 객체, 마이그레이션 fixture, 매니페스트, 정확한 checkout 및 작업 트리 상태를 검증한다.
설계 근거 및 추적성 문서 docs/doctoring/candidate-application-references.md, docs/traceability/candidate-application-core.md, docs/adr/README.md, CHANGELOG.md	정규화 모델의 표준 근거, 설계 결정, 구현 추적성, ADR 인덱스 및 변경 로그를 갱신했다.

Estimated code review effort: 4 (Complex) | ~45 minutes

Merge Risk: 🔵 Low · up to 83918

The PR normalizes candidate application identity and lifecycle history. Current checks are green, but adding a uniqueness constraint to existing position data during deployment may briefly block reads and writes, so merge is low risk with owner awareness of the acceptable migration lock duration; the remaining documentation and test-maintenance issues are non-blocking.

Sequence Diagram(s)

sequenceDiagram
  participant PullRequest
  participant GitHubActions
  participant PostgreSQL
  participant ContractTests
  PullRequest->>GitHubActions: trigger workflow
  GitHubActions->>PostgreSQL: start PostgreSQL 16.14
  GitHubActions->>ContractTests: run ordered PostgreSQL contracts
  ContractTests->>PostgreSQL: apply migration and validate protections
  PostgreSQL-->>ContractTests: return contract results
  ContractTests-->>GitHubActions: report tests and worktree status

Loading

Suggested reviewers: cursoragent

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 66.67% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 12 functions across 6 files. (9 skipped: …	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	제목은 후보자 지원 라이프사이클 정규화라는 이 PR의 주요 변경 사항을 정확하고 간결하게 설명합니다.

Full details: Docstring Coverage

Explanation

Docstring coverage is 66.67% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 12 functions across 6 files. (9 skipped: 9 unsupported.)

• Fix all pre-merge checks with AI

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat/normalized-candidate-application

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[devin-ai-integration]

Devin Review found 1 new potential issue.