### FIN-REVIEW-001-A-008
  status=in_arbeit rel=current sev=nice_to_have topic=json-multiindex-roundtrip
  area=io_backends
  track=FTR-REVIEW-001-MULTIHEADER-READBACK-P5; FTR-MULTIHEADER-P2
  summary: JSON read/write is not symmetric for MultiIndex frames.
  current_assessment: Still open and now tracked by FTR-REVIEW-001-MULTIHEADER-READBACK-P5 as follow-up to FTR-MULTIHEADER-P2. Current JSON write supports nested MultiIndex output, but read_multi still ignores header_levels and does not rebuild symmetric MultiIndex readback.

### FIN-REVIEW-001-A-009
  status=in_arbeit rel=current sev=nice_to_have topic=csv-multiheader-read
  area=io_backends
  track=FTR-REVIEW-001-MULTIHEADER-READBACK-P5; FTR-MULTIHEADER-P2
  summary: CSVBackend.read_multi ignores header_levels and loses real multi-header structure.
  current_assessment: Still open and linked to the FTR-MULTIHEADER-P2 lineage. Current CSV read_multi still reads header=0 and constructs a padded fake MultiIndex instead of honoring header_levels for real multi-header CSV directory input.

### FIN-REVIEW-001-D-004
  status=in_arbeit rel=current sev=should_fix topic=module-size
  area=domain
  track=FTR-REVIEW-001-MODULARIZATION-P4A
  summary: Several domain modules exceed the practical read-window size for humans and agents.
  current_assessment: Still open and tracked by FTR-REVIEW-001-MODULARIZATION-P4A. A current LOC check still finds 6 src modules >500 lines and 11 >440 lines, including domain/core modules, so this is not closed; later package-split methodology is useful precedent but not closure evidence.

### FIN-REVIEW-001-D-006
  status=in_arbeit rel=current sev=should_fix topic=transformation-helper-shape
  area=domain/transformations
  track=
  summary: transformations/helpers.py mixes closure-factory style with unrelated helper responsibilities.
  current_assessment: Still open. No dedicated FTR found. REVIEW-002 lists transformations/helpers.py paradigm mix as unchanged; later package/domain reviews still treat transformations/helpers.py as a structural anomaly and identify a future split candidate C003.

### FIN-REVIEW-001-D-008
  status=in_arbeit rel=current sev=should_fix topic=module-mislabel
  area=domain/validations
  track=FTR-REVIEW-001-META-BOUNDARY-CLEANUP-P5
  summary: validations/validate_columns.py adds constraints rather than validating columns.
  current_assessment: Still open and now tracked by FTR-REVIEW-001-META-BOUNDARY-CLEANUP-P5. Current validate_columns.py still exposes add_validations as a metadata-producing constraint writer, so the naming/ownership finding is not closed.

### FIN-REVIEW-001-D-009
  status=in_arbeit rel=current sev=should_fix topic=compat-residue
  area=core
  track=FTR-DOMAIN-CONSOLIDATION; FTR-FK-HELPERS-POLICY-DRIVEN-PRIMITIVES-P5; FK Helper Consolidation Closure
  summary: Legacy FK-helper compatibility branches and a broad exception handler remain in core.
  current_assessment: Partially addressed by FK-helper v1 retirement and policy-driven primitives, but the original compat-residue evidence is not fully closed: core/fk.py still carries legacy dict input handling and core/refs.py still has a broad except. Keep open for detailed review.

### FIN-REVIEW-001-D-010
  status=in_arbeit rel=current sev=should_fix topic=package-surface
  area=domain
  track=FTR-REVIEW-001-MODULARIZATION-P4A
  summary: Domain and rendering package surfaces are empty or inconsistent.
  current_assessment: Partially mitigated for newer split packages, but REVIEW-002 says the original empty/inconsistent package surfaces remain. Keep current/open.

### FIN-REVIEW-001-D-011
  status=in_arbeit rel=current sev=should_fix topic=domain-io
  area=domain
  track=FTR-REVIEW-001-META-BOUNDARY-CLEANUP-P5
  summary: Domain code performs file I/O and directly depends on pyyaml without a guard.
  current_assessment: Still open and now tracked by FTR-REVIEW-001-META-BOUNDARY-CLEANUP-P5. Current domain code still imports yaml and performs file I/O in structured_yaml.py, yaml_overrides.py, and related domain helpers; this is not closed by comment alone.

### FIN-REVIEW-001-D-012
  status=in_arbeit rel=current sev=nice_to_have topic=module-headers
  area=domain / core
  track=FTR-REVIEW-001-MODULE-HEADERS-P5
  summary: Domain and core modules lack a consistent module-header convention.
  current_assessment: Tracked by the module-headers FTR, but no implementation/closure found. REVIEW-002 lists D-012/IR-014 as unchanged.

### FIN-REVIEW-001-D-013
  status=in_arbeit rel=current sev=nice_to_have topic=carrier-shim
  area=domain
  track=FTR-REVIEW-001-META-BOUNDARY-CLEANUP-P5
  summary: meta_bootstrap and validate_columns still contain a dual-interface Frames shim.
  current_assessment: Still open and now tracked by FTR-REVIEW-001-META-BOUNDARY-CLEANUP-P5. meta_bootstrap.py and validate_columns.py still carry the dual-interface convention, including frames.meta and frames["_meta"] handling.

### FIN-REVIEW-001-IR-009
  status=in_arbeit rel=current sev=nice_to_have topic=rendering-public-surface
  area=rendering
  track=FTR-REVIEW-001-MODULARIZATION-P4A
  summary: rendering/__init__.py is empty, forcing consumers to know deep paths.
  current_assessment: Tracked with C7 modularization, but no direct closure found. REVIEW-002 says newer package surfaces improved while original rendering/domain package surfaces remained incomplete.

### FIN-REVIEW-001-IR-011
  status=in_arbeit rel=current sev=nice_to_have topic=fk-step-naming
  area=pipeline
  track=FTR-DOMAIN-CONSOLIDATION; FTR-PIPELINE-STEP-NAMING-P4
  summary: FK step factory names do not match visible registry keys.
  current_assessment: Public YAML step names were normalized by FTR-PIPELINE-STEP-NAMING-P4, but the original internal factory-name mismatch remains visible: make_apply_fks_step, make_drop_helpers_step, and make_check_fk_helpers_step still differ from registry keys. Keep open under FTR-DOMAIN-CONSOLIDATION.

### FIN-REVIEW-001-IR-013
  status=in_arbeit rel=current sev=nice_to_have topic=exported-at-placeholder
  area=rendering
  track=FTR-REVIEW-001-META-BOUNDARY-CLEANUP-P5
  summary: MetaPass reserves exported_at but no package path fills it.
  current_assessment: Still open and now tracked by FTR-REVIEW-001-META-BOUNDARY-CLEANUP-P5. MetaPass still reserves exported_at with an empty default and no package path fills it; later payload work treats it as workbook identity but does not close the ownership decision.

### FIN-REVIEW-001-IR-014
  status=in_arbeit rel=current sev=nice_to_have topic=module-headers
  area=rendering / pipeline
  track=FTR-REVIEW-001-MODULE-HEADERS-P5
  summary: Rendering and pipeline modules lack a consistent module-header convention.
  current_assessment: Tracked by the module-headers FTR, but no implementation/closure found. REVIEW-002 lists D-012/IR-014 as unchanged.

### FIN-REVIEW-002-P1-003
  status=open rel=current sev=medium topic=guard-inventory
  area=tests / architecture docs
  track=FTR-GUARDRAIL-AUDIT-P5
  summary: Architecture guard inventory no longer enumerates the live guard suite.
  current_assessment: Still open and tracked by FTR-GUARDRAIL-AUDIT-P5. REVIEW-003 independently confirms the architecture guard inventory still lags the live suite; this is follow-up evidence, not closure evidence.

### FIN-REVIEW-002-P1-005
  status=open rel=current sev=medium topic=architecture-doc-lag
  area=architecture docs
  track=FTR-CONCEPT-DOCS-CLEANUP-P5
  summary: ARCH-LAYERS still describes stale modules and pass order after the architecture moved on.
  current_assessment: Partially mitigated by later edits to architectural_layers, but still open as architecture/concept documentation lag. REVIEW-003 carries the remaining glossary/doc adoption work under FTR-CONCEPT-DOCS-CLEANUP-P5 and does not treat it as a 0.2 blocker.

### FIN-REVIEW-002-P1-006
  status=accepted rel=partial sev=low topic=tiny-helper-duplication
  area=domain / transformations
  track=
  summary: Recurring tiny-helper duplication exists across several modules, but is a watch item rather than a defect.
  current_assessment: Still accepted as watch-only. REVIEW-003 confirms tiny-helper duplication remains, but classifies it as trivia-level under-consolidation rather than a defect or active FTR.

### FIN-REVIEW-002-P2-002
  status=accepted rel=partial sev=low topic=predicate-duplication
  area=domain
  track=
  summary: reference_validations and join_views duplicate a small where/condition predicate vocabulary.
  current_assessment: Partially mitigated by commit 05891da extracting shared where predicates for frame extraction, join views, and tabular views. reference_validations still has a separate predicate path, so this remains a watch item rather than closed evidence.

### FIN-REVIEW-002-P3-002
  status=open rel=current sev=medium topic=workbook-option-vocabulary
  area=rendering / io_backends
  track=
  summary: Workbook option vocabulary is triplicated and already divergent across write and read paths.
  current_assessment: Still open. Current code still has composer _WORKBOOK_OPTION_KEYS plus separate XLSX/ODS OPTION_HINT_KEYS; REVIEW-003 confirms adapter-boundary constants remain under-owned. No dedicated FTR found.

### FIN-REVIEW-003-S1-F-A02-3
  status=open rel=current sev=nice_to_have topic=finding-frame-doc-clarity
  area=domain finding docs
  track=FTR-CONCEPT-DOCS-CLEANUP-P5
  summary: Finding-frame duplication has become documentation clarity work.
  current_assessment: Still open as documentation clarity, not runtime debt. The model consolidation is done, but concept docs still need to explain the two remaining finding shapes.

### FIN-REVIEW-003-S1-W-2
  status=accepted rel=partial sev=low topic=ods-parity-third-backend
  area=io_backends
  track=
  summary: ODS adapter parity has only XLSX as its comparison backend.
  current_assessment: Accepted watch-only. No current action; real-use or a third backend would be needed to turn this into a concrete FTR.

### FIN-REVIEW-003-S2-F-0-2-2-A2
  status=open rel=current sev=nice_to_have topic=examples-discoverability
  area=README / examples
  track=
  summary: After root examples removal, package-internal examples are less discoverable from a repo browse.
  current_assessment: Still open as low-priority discoverability polish. No top-level examples are tracked; the remaining question is whether README should point to package-internal examples.

### FIN-REVIEW-003-S2-W-0-2-2-1
  status=accepted rel=partial sev=low topic=cli-shape-summary
  area=CLI / user guide
  track=
  summary: CLI config/steps/profile shapes lack a compact documentation summary.
  current_assessment: Accepted watch-only. The user guide now has a minimal sheets-run bridge, but no detailed precedence table; no evidence this is currently causing user friction.

### FIN-REVIEW-003-S2-W-0-2-2-2
  status=accepted rel=partial sev=low topic=io-input-shape-asymmetry
  area=pipeline config
  track=
  summary: Runtime io.input and dataclass io.inputs shapes are asymmetric.
  current_assessment: Accepted watch-only. Root stale configs that exposed the plural shape are gone; remaining plural shape is internal dataclass surface unless future docs leak it.

### FIN-REVIEW-003-S2-W-0-2-2-3
  status=accepted rel=partial sev=low topic=demo-local-binding-wording
  area=demo onboarding
  track=
  summary: Demo setup-lib-local assumes a sibling checkout before documenting PyPI fallback.
  current_assessment: Accepted watch-only. REVIEW-003 Step 3 says the demo note is sufficient today; observe real-use feedback rather than pre-emptively rewriting.

### FIN-REVIEW-003-S3-F-RR-A02-2
  status=open rel=current sev=nice_to_have topic=parameter-documentation-placeholders
  area=user guide transformation catalog
  track=
  summary: Full parameter documentation for composites and supporting steps remains after-0.2 polish.
  current_assessment: Still open. Current catalog still contains Full parameter documentation planned placeholders.

### FIN-REVIEW-003-S3-F-RR-A02-4
  status=accepted rel=partial sev=low topic=public-python-import-surface
  area=public API docs
  track=
  summary: Public Python import surface should be formalized only if consumers depend on it.
  current_assessment: Accepted watch-only. Current release notes inten