# Changelog

## Score

- CAI 59 → 65 (+6.7)
- Rubric changed (rubric-2026.10.5 → rubric-2026.11.2) — scores are not directly comparable.

## Lenses

- Code Health 93 → 94 (+0.9)
- Architecture 78 → 77 (-1.3)
- Maturity 53 → 85 (+31.8)
- Readiness 47 → 47 (+0.0)
- Security 76 → 78 (+2.0)
- Domain Modelling 100 → 100 (+0.0)
- Performance 100 → 100 (+0.0)

## Resolved (35)

- ADR has no Consequences section (docs/history/archive/025-oss-preparation-strategy.md)
- DocLinkChecker._check_link (cognitive 18) (tools/check_doc_links.py)
- Documentation: contradicts the code (docs/history/archive/governor-review-log/README.md)
- Duplicated block (10 lines × 2) (.antigravity/hooks/completion_gate.py)
- Duplicated block (12 lines × 2) (.antigravity/hooks/verify_first.py)
- Duplicated block (14–16 lines × 2) (.antigravity/hooks/post-tool-format.py)
- Duplicated block (19 lines × 2) (.antigravity/hooks/verify_first.py)
- Duplicated block (28–31 lines × 2) (.antigravity/hooks/stop-sync-reminder.py)
- Duplicated block (31–46 lines × 2) (.antigravity/hooks/stop-sync-reminder.py)
- Duplicated block (37–40 lines × 2) (.antigravity/hooks/stop-sync-reminder.py)
- FileTooLong: tools/governor_state_doctor.py (tools/governor_state_doctor.py)
- High CVE: [GHSA redacted] (uv.lock)
- High CVE: [GHSA redacted] (uv.lock)
- Hotspot: src/_core/config.py (src/_core/config.py)
- Hotspot: src/_core/infrastructure/persistence/nosql/dynamodb/base_dynamo_repository.py (src/_core/infrastructure/persistence/nosql/dynamodb/base_dynamo_repository.py)
- Inconsistent naming convention for ID-based lookups. The repository layer uses 'select' (SQL-centric) while the service layer uses 'get'. Furthermore, the pluralization is inconsistent: 'select_datas_by_ids' vs 'get_datas_by_data_ids'. The singular forms also differ: 'select_data_by_id' vs 'get_data_by_data_id'.
- Inconsistent naming convention for creation operations. The repository layer uses 'insert' while the service layer uses 'create'. Additionally, the pluralization suffix is inconsistent: 'insert_datas' vs 'create_datas' (both use 'datas', which is non-standard English, but the verb change is the primary inconsistency).
- Inconsistent naming for delete operations. Similar to update, the generic layer uses 'delete_data_by_data_id' while the DynamoDB service uses 'delete_item'.
- Inconsistent naming for update operations. The generic repository and service use 'update_data_by_data_id', while the DynamoDB-specific service uses 'update_item'. This forces consumers to know the underlying storage mechanism to use the correct method name.
- Inconsistent return types and method names for paginated listing. The repository returns a simple list, while services return a tuple of (list, PaginationInfo). The method names are 'select_datas' vs 'get_datas'. The AdminCrudServiceProtocol uses 'Any' for the second tuple element, breaking type safety compared to BaseService.
- …and 15 more

## New (19)

- Critical CVE: [GHSA redacted] (uv.lock)
- Critical CVE: [GHSA redacted] (uv.lock)
- Duplicated block (15 lines × 2) (.antigravity/hooks/stop-sync-reminder.py)
- Duplicated block (18–20 lines × 2) (.antigravity/hooks/post-tool-format.py)
- Duplicated block (24 lines × 2) (.antigravity/hooks/stop-sync-reminder.py)
- Duplicated block (30–33 lines × 2) (.antigravity/hooks/stop-sync-reminder.py)
- Duplicated block (53–56 lines × 2) (.antigravity/hooks/stop-sync-reminder.py)
- High CVE: [GHSA redacted] (uv.lock)
- High CVE: [GHSA redacted] (uv.lock)
- High CVE: [GHSA redacted] (uv.lock)
- High CVE: [GHSA redacted] (uv.lock)
- High CVE: [GHSA redacted] (uv.lock)
- Hotspot: .agents/shared/work_ledger.py (.agents/shared/work_ledger.py)
- Medium CVE: [GHSA redacted] (uv.lock)
- Members sharing a duplicated core (4 members, 50+ identical tokens) (.antigravity/hooks/stop-sync-reminder.py)
- Repeated repair: .agents/shared/governor/paths.py (.agents/shared/governor/paths.py)
- completion_gate.governor_changing_segment (cognitive 22) (.claude/hooks/completion_gate.py)
- completion_gate.governor_changing_segment (cyclomatic 21) (.claude/hooks/completion_gate.py)
- user-prompt-submit.main (cognitive 16) (.codex/hooks/user-prompt-submit.py)

## Changes since last survey

- 74 commits — 44 feature/other, 30 fixes

## By area

- (repo) — 29 commits
- docs/ai — 15 commits
- .agents/shared — 8 commits
- (root) — 6 commits
- .claude/rules — 5 commits
- tests/unit — 5 commits
- .claude/hooks — 2 commits
- .codex/hooks — 1 commit
- docs/history — 1 commit
- tests/conftest.py — 1 commit
- tests/integration — 1 commit

## Notable commits

- fix: Merge pull request #428 from Mr-DooSun/fix/425-harness-state-root
- fix: Merge pull request #429 from Mr-DooSun/fix/ledger-write-race
- fix: Merge pull request #435 from Mr-DooSun/fix/413-codex-project-doc-tail
- fix: chore(stack): carry review fixes from PR #435 into this unit
- fix: chore(stack): carry review fixes from PR #436 into this unit
- fix: chore(stack): carry review fixes from PR #437 into this unit
- fix: chore(stack): carry review fixes from PR #438 into this unit
- fix: chore(stack): carry review fixes from PR #439 into this unit
- fix: chore(stack): carry review fixes from PR #440 into this unit
- fix: chore(stack): carry review fixes from PR #441 into this unit
- fix: chore(stack): carry review fixes from PR #442 into this unit
- fix: chore(stack): carry review fixes from PR #442 into this unit
- fix: chore(stack): carry review fixes from PR #443 into this unit
- fix: chore(stack): carry review fixes from PR #443 into this unit
- fix: chore(stack): carry review fixes from PR #444 into this unit
- fix: chore(stack): carry review fixes from PR #444 into this unit
- fix: fix(harness): close context-budget review gaps (#416)
- fix: fix(harness): keep list-nested fences visible to the context budget (#416)
- fix: fix(harness): keep the full shared rules visible to Codex (#413)
- fix: fix(harness): reject grouped shell operators in the verify classifier (#414)
- …and 54 more
