From dcecf7e649585242dd85a23665699650d8040c24 Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Thu, 27 Aug 2026 04:10:32 +0900
Subject: [PATCH 01/17] feat(ontology): import pinned O*NET occupation ratings

---
 CHANGELOG.md                                  |   5 +-
 ...net-occupation-rating-observation-store.md |  11 +-
 docs/product-requirements.md                  |  11 +-
 docs/product-technical-gap-baseline.md        |   2 +-
 scripts/import_onet_ratings.py                | 529 ++++++++++++++++++
 tests/test_import_onet_ratings.py             | 181 ++++++
 tests/test_schema.py                          |  48 ++
 7 files changed, 780 insertions(+), 7 deletions(-)
 create mode 100644 scripts/import_onet_ratings.py
 create mode 100644 tests/test_import_onet_ratings.py

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 45fb3200e..2067e64d2 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -12,7 +12,10 @@ All notable changes to this project are documented here. Format follows
   contract with normalized release, source-table, scale, occupation, element,
   and observation tables. Exact release/source LIST partitions fail closed;
   nullable categories remain idempotent identities, and no source value is
-  promoted to a local weight or person score (ADR 0257).
+  promoted to a local weight or person score. A pinned CSV importer validates
+  rating and scale-reference digests, row counts, source identities, scale
+  bounds, uncertainty, flags, and update dates before immutable transactional
+  insertion (ADR 0257).
 - All eight O*NET 31.0 published linkage tables now contribute 1,417 directed
   Ability/Essential Skill/Transferable Skill/Work Style relations to Work
   Activities and Work Context. Every direct relation has an exact reified
diff --git a/docs/adr/0257-onet-occupation-rating-observation-store.md b/docs/adr/0257-onet-occupation-rating-observation-store.md
index bcdd3e358..b34093eaf 100644
--- a/docs/adr/0257-onet-occupation-rating-observation-store.md
+++ b/docs/adr/0257-onet-occupation-rating-observation-store.md
@@ -45,13 +45,22 @@ also violate LineageWeave's externalized-compute boundary.
 8. This store is immutable source evidence. Any later aggregation, comparison,
    temporal model, multilevel model, or occupational recommendation belongs to
    TEPP/fast-mlsirm or another owning Rust service and must cite these rows.
+9. `scripts/import_onet_ratings.py` accepts one caller-pinned official CSV and
+   Scales Reference file. It verifies both artifact SHA-256 values and row
+   counts, exact scale names and bounds, reference-name consistency, finite
+   decimals, Y/N/blank flags, optional Category and Not Relevant columns,
+   MM/YYYY source dates, and observation-key uniqueness before opening the
+   target connection. A transaction-scoped advisory lock serializes one
+   release's partition DDL.
 
 ## Consequences
 
 LineageWeave can import the complete public O*NET rating corpus without
 manufacturing semantics or embedding large production datasets in git.
 Release/source partitions localize hot imports and permit exact detach/archive
-operations. A separate API/UI ADR is still required before exposing ratings.
+operations. The same pinned artifact is idempotent; a reused release, source,
+scale, occupation, or element identity with different source metadata fails
+closed. A separate API/UI ADR is still required before exposing ratings.
 
 ## References
 
diff --git a/docs/product-requirements.md b/docs/product-requirements.md
index fa51c711f..9e8473d8f 100644
--- a/docs/product-requirements.md
+++ b/docs/product-requirements.md
@@ -136,10 +136,13 @@ persistence and UI remain unavailable until their separate ADR acceptance.
 - Preserve decimals and missingness exactly. No local aggregation,
   normalization, person inference, or psychometric estimation is permitted.
 
-Acceptance: the replay-safe migration creates the normalized store; PostgreSQL
-integration proves missing partitions fail closed, null-category UPSERT is
-idempotent, and invalid uncertainty is rejected. Corpus import, API, UI, and
-derived modeling remain unavailable until separate accepted delivery records.
+Acceptance: the replay-safe migration creates the normalized store; the pinned
+CSV importer validates both rating and scale-reference digests and row counts,
+reference identity, source scale, uncertainty, flags, and dates before
+persistence; PostgreSQL integration proves missing partitions fail closed and
+repeated null-category UPSERT is idempotent.
+API, UI, and derived modeling remain unavailable until separate accepted
+delivery records.
 
 ### PRD-FR-3 — Bounded ontology exploration
 
diff --git a/docs/product-technical-gap-baseline.md b/docs/product-technical-gap-baseline.md
index 1a19a8bd2..29cf50b38 100644
--- a/docs/product-technical-gap-baseline.md
+++ b/docs/product-technical-gap-baseline.md
@@ -26,7 +26,7 @@ explicit unavailable state, not a reason to infer mappings from labels.
 | Classification depth | ADR 0252's candidate branch imports all 1,447 official 2018 SOC nodes with source-declared hierarchy, pinned XLSX/CSV digests, a deterministic Turtle renderer, and fail-closed read model | Pass exact-head review/checks and protected merge; add ISCO/ESCO crosswalks only where the publishing authority supplies them |
 | Construct granularity | ADR 0255's candidate branch publishes all 3,006 O*NET 31.0 Content Model Reference concepts with exact IDs, names, descriptions, and source-defined parents; it imports no ratings or person assertions | Pass exact-head review/checks and protected merge; import separately released occupation-element observations only through the provenance contract below |
 | Construct-to-work relations | ADR 0256's candidate branch imports all 1,417 official Ability/Essential Skill/Transferable Skill/Work Style linkages to Work Activities and Work Context with assertion-level source provenance and no invented weights | Pass exact-head review/checks and protected merge; keep these published relevance links distinct from occupation ratings and causal claims |
-| Occupation-to-construct relations | ADR 0257 defines a candidate 3NF, release/source-partitioned immutable observation store preserving value, optional category, sample/error/CI, suppression, relevance, update date, source digest, and domain provenance; no corpus is imported | Pass exact-head review/checks and protected merge; add a deterministic source importer for pinned public artifacts; never invent or locally normalize a weight |
+| Occupation-to-construct relations | ADR 0257 defines a candidate 3NF, release/source-partitioned immutable observation store and deterministic pinned-CSV importer preserving value, optional category, sample/error/CI, suppression, relevance, update date, source digest, and domain provenance. The official O*NET 31.0 Abilities file (94,640 rows, 910 occupations, 52 elements; SHA-256 `7e9cd79791ce6014e1d26d0a449ae5b1e7aa7ef52d39b3934c3bb8d438104b88`) and all 33 Scales Reference rows (SHA-256 `bcba23858ce21ecaacbde303a8993e35d46724b4afb8c9ec2b10e04f42adcfc9`) imported into a throwaway local PostgreSQL database with all 94,640 observations, 55 suppression flags, 7,572 not-relevant flags, and source dates from December 2004 through August 2026; every scale retained `scales_reference` artifact provenance, the database was dropped afterward, and no corpus is committed or claimed deployed | Pass exact-head review/checks and protected merge; validate and import every selected official rating artifact through an authorized runtime, returning only aggregate evidence; never invent or locally normalize a weight |
 | Job-family and job-series semantics | No authoritative employer-specific job architecture is present | Define an organization-neutral import contract that preserves the authorized source hierarchy and distinguishes standard occupation codes from employer job families/series; no label-based binding |
 | Temporal and multilevel interpretation | Static vocabulary only; no person-level inference is asserted | Version valid and transaction time, preserve occupation/organization/unit nesting and multiple membership, and require TEPP or the owning Rust psychometric service before any calibrated temporal or multilevel result |
 | Product consumption | The read model has no persisted semantic-layer consumer or authenticated UI evidence | Add a provenance-bearing API and accessible ontology exploration flow, then verify synthetic Storybook edge states plus authenticated aggregate runtime evidence without exposing identifying records |
diff --git a/scripts/import_onet_ratings.py b/scripts/import_onet_ratings.py
new file mode 100644
index 000000000..a09cbe4bc
--- /dev/null
+++ b/scripts/import_onet_ratings.py
@@ -0,0 +1,529 @@
+"""Validate and import one official O*NET occupation-rating CSV artifact."""
+
+from __future__ import annotations
+
+import argparse
+import asyncio
+import csv
+import hashlib
+import json
+import re
+from dataclasses import dataclass
+from datetime import UTC, date, datetime
+from decimal import Decimal, InvalidOperation
+from pathlib import Path
+from urllib.parse import urlsplit
+
+import asyncpg
+
+_RATING_FIELDS = {
+    "O*NET-SOC Code",
+    "Title",
+    "Element ID",
+    "Element Name",
+    "Scale ID",
+    "Scale Name",
+    "Data Value",
+    "N",
+    "Standard Error",
+    "Lower CI Bound",
+    "Upper CI Bound",
+    "Recommend Suppress",
+    "Date",
+    "Domain Source",
+}
+_SCALE_FIELDS = {"Scale ID", "Scale Name", "Minimum", "Maximum"}
+_SOURCE_CODE = re.compile(r"^[a-z][a-z0-9_]{0,62}$")
+_SCALES_SOURCE_CODE = "scales_reference"
+_SHA256 = re.compile(r"^[0-9a-fA-F]{64}$")
+
+
+@dataclass(frozen=True)
+class ScaleDefinition:
+    """One exact O*NET scale identity and its declared numeric bounds."""
+
+    scale_id: str
+    scale_name: str
+    minimum_value: Decimal
+    maximum_value: Decimal
+
+
+@dataclass(frozen=True)
+class RatingObservation:
+    """One validated O*NET occupation-to-element source observation."""
+
+    onetsoc_code: str
+    occupation_title: str
+    element_id: str
+    element_name: str
+    scale_id: str
+    category_value: int | None
+    data_value: Decimal
+    sample_size: int | None
+    standard_error: Decimal | None
+    lower_ci_bound: Decimal | None
+    upper_ci_bound: Decimal | None
+    recommend_suppress: bool | None
+    not_relevant: bool | None
+    source_updated_date: date
+    domain_source_code: str
+
+
+def _decimal(value: str, field: str, *, optional: bool = False) -> Decimal | None:
+    """Parse one finite source decimal while preserving an honest blank."""
+    text = value.strip()
+    if optional and not text:
+        return None
+    try:
+        parsed = Decimal(text)
+    except InvalidOperation as exc:
+        raise ValueError(f"invalid {field}: {value!r}") from exc
+    if not parsed.is_finite():
+        raise ValueError(f"invalid {field}: {value!r}")
+    return parsed
+
+
+def _integer(value: str, field: str, *, optional: bool = False) -> int | None:
+    """Parse one source integer while preserving an honest blank."""
+    text = value.strip()
+    if optional and not text:
+        return None
+    try:
+        return int(text)
+    except ValueError as exc:
+        raise ValueError(f"invalid {field}: {value!r}") from exc
+
+
+def _flag(value: str, field: str) -> bool | None:
+    """Parse the official Y/N/blank tri-state flag vocabulary."""
+    text = value.strip()
+    if not text:
+        return None
+    if text == "Y":
+        return True
+    if text == "N":
+        return False
+    raise ValueError(f"invalid {field} flag: {value!r}")
+
+
+def _updated_month(value: str, today: date) -> date:
+    """Parse O*NET MM/YYYY source dates and reject a future release month."""
+    try:
+        month_text, year_text = value.strip().split("/")
+        parsed = date(int(year_text), int(month_text), 1)
+    except (TypeError, ValueError) as exc:
+        raise ValueError(f"invalid source update date: {value!r}") from exc
+    if parsed > today.replace(day=1):
+        raise ValueError(f"future source update date: {value!r}")
+    return parsed
+
+
+def _rows(path: Path, required: set[str]) -> list[dict[str, str]]:
+    """Read one UTF-8 CSV only when its authoritative columns are present."""
+    with path.open(encoding="utf-8-sig", newline="") as handle:
+        reader = csv.DictReader(handle)
+        fields = set(reader.fieldnames or ())
+        missing = sorted(required - fields)
+        if missing:
+            raise ValueError(f"missing CSV columns: {', '.join(missing)}")
+        return [dict(row) for row in reader]
+
+
+def read_scale_file(path: Path) -> dict[str, ScaleDefinition]:
+    """Return exact scale definitions from the official Scales Reference CSV."""
+    scales: dict[str, ScaleDefinition] = {}
+    for row in _rows(path, _SCALE_FIELDS):
+        scale_id = row["Scale ID"].strip()
+        definition = ScaleDefinition(
+            scale_id=scale_id,
+            scale_name=row["Scale Name"].strip(),
+            minimum_value=_decimal(row["Minimum"], "scale minimum"),  # type: ignore[arg-type]
+            maximum_value=_decimal(row["Maximum"], "scale maximum"),  # type: ignore[arg-type]
+        )
+        if not scale_id or not definition.scale_name:
+            raise ValueError("empty scale identity")
+        if definition.minimum_value > definition.maximum_value:
+            raise ValueError(f"inverted scale bounds: {scale_id}")
+        if scale_id in scales and scales[scale_id] != definition:
+            raise ValueError(f"conflicting scale definition: {scale_id}")
+        scales[scale_id] = definition
+    if not scales:
+        raise ValueError("scale file has no rows")
+    return scales
+
+
+def read_rating_file(
+    path: Path,
+    scales: dict[str, ScaleDefinition],
+    *,
+    today: date | None = None,
+) -> list[RatingObservation]:
+    """Validate an official rating CSV and return exact source observations."""
+    observed_today = today or datetime.now(UTC).date()
+    occupations: dict[str, str] = {}
+    elements: dict[str, str] = {}
+    observations: list[RatingObservation] = []
+    identities: set[tuple[str, str, str, int | None]] = set()
+    for row in _rows(path, _RATING_FIELDS):
+        onetsoc_code = row["O*NET-SOC Code"].strip()
+        occupation_title = row["Title"].strip()
+        element_id = row["Element ID"].strip()
+        element_name = row["Element Name"].strip()
+        scale_id = row["Scale ID"].strip()
+        scale = scales.get(scale_id)
+        if scale is None or row["Scale Name"].strip() != scale.scale_name:
+            raise ValueError(f"unknown or conflicting scale identity: {scale_id}")
+        if (
+            onetsoc_code in occupations
+            and occupations[onetsoc_code] != occupation_title
+        ):
+            raise ValueError(f"conflicting occupation title: {onetsoc_code}")
+        if element_id in elements and elements[element_id] != element_name:
+            raise ValueError(f"conflicting element name: {element_id}")
+        occupations[onetsoc_code] = occupation_title
+        elements[element_id] = element_name
+        data_value = _decimal(row["Data Value"], "data value")
+        if not scale.minimum_value <= data_value <= scale.maximum_value:  # type: ignore[operator]
+            raise ValueError(f"data value outside scale {scale_id}")
+        sample_size = _integer(row["N"], "sample size", optional=True)
+        standard_error = _decimal(
+            row["Standard Error"], "standard error", optional=True
+        )
+        lower = _decimal(row["Lower CI Bound"], "lower CI bound", optional=True)
+        upper = _decimal(row["Upper CI Bound"], "upper CI bound", optional=True)
+        if sample_size is not None and sample_size <= 0:
+            raise ValueError("sample size must be positive")
+        if standard_error is not None and standard_error < 0:
+            raise ValueError("standard error must be non-negative")
+        if (lower is None) != (upper is None) or (lower is not None and lower > upper):
+            raise ValueError("invalid confidence interval")
+        category = _integer(row.get("Category", ""), "category", optional=True)
+        identity = (onetsoc_code, element_id, scale_id, category)
+        if identity in identities:
+            raise ValueError(f"duplicate rating identity: {identity}")
+        identities.add(identity)
+        domain_source = row["Domain Source"].strip()
+        if (
+            not onetsoc_code
+            or not occupation_title
+            or not element_id
+            or not element_name
+            or not domain_source
+        ):
+            raise ValueError("empty rating identity")
+        observations.append(
+            RatingObservation(
+                onetsoc_code=onetsoc_code,
+                occupation_title=occupation_title,
+                element_id=element_id,
+                element_name=element_name,
+                scale_id=scale_id,
+                category_value=category,
+                data_value=data_value,  # type: ignore[arg-type]
+                sample_size=sample_size,
+                standard_error=standard_error,
+                lower_ci_bound=lower,
+                upper_ci_bound=upper,
+                recommend_suppress=_flag(
+                    row["Recommend Suppress"], "recommend suppress"
+                ),
+                not_relevant=_flag(row.get("Not Relevant", ""), "not relevant"),
+                source_updated_date=_updated_month(row["Date"], observed_today),
+                domain_source_code=domain_source,
+            )
+        )
+    if not observations:
+        raise ValueError("rating file has no rows")
+    return observations
+
+
+def _parser() -> argparse.ArgumentParser:
+    """Build the explicit, provenance-bearing importer command contract."""
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument("--target-dsn", required=True)
+    parser.add_argument("--release-code", required=True)
+    parser.add_argument("--release-version", required=True)
+    parser.add_argument("--source-table-code", required=True)
+    parser.add_argument("--source-table-name", required=True)
+    parser.add_argument("--source-url", required=True)
+    parser.add_argument("--source-sha256", required=True)
+    parser.add_argument("--source-row-count", type=int, required=True)
+    parser.add_argument("--publisher", default="National Center for O*NET Development")
+    parser.add_argument(
+        "--license-url", default="https://creativecommons.org/licenses/by/4.0/"
+    )
+    parser.add_argument("--scales-file", type=Path, required=True)
+    parser.add_argument("--scales-url", required=True)
+    parser.add_argument("--scales-sha256", required=True)
+    parser.add_argument("--scales-row-count", type=int, required=True)
+    parser.add_argument("--ratings-file", type=Path, required=True)
+    return parser
+
+
+def _validate_args(args: argparse.Namespace) -> None:
+    """Reject ambiguous provenance and unsafe source identities before I/O."""
+    if (
+        not _SOURCE_CODE.fullmatch(args.source_table_code)
+        or args.source_table_code == _SCALES_SOURCE_CODE
+    ):
+        raise ValueError("source table code must be non-reserved lower snake case")
+    for field in ("release_code", "release_version", "source_table_name", "publisher"):
+        if not str(getattr(args, field)).strip():
+            raise ValueError(f"{field} must not be blank")
+    for field in ("source_sha256", "scales_sha256"):
+        if not _SHA256.fullmatch(str(getattr(args, field))):
+            raise ValueError(f"{field} must be one SHA-256 digest")
+    for field in ("source_row_count", "scales_row_count"):
+        if getattr(args, field) <= 0:
+            raise ValueError(f"{field} must be positive")
+    for field in ("source_url", "scales_url", "license_url"):
+        parsed = urlsplit(str(getattr(args, field)))
+        if parsed.scheme != "https" or not parsed.hostname or parsed.username:
+            raise ValueError(f"{field} must be an HTTPS URL without userinfo")
+    for field in ("ratings_file", "scales_file"):
+        if not getattr(args, field).is_file():
+            raise ValueError(f"{field} must be a regular file")
+
+
+async def _reject_reference_conflicts(
+    conn: asyncpg.Connection,
+    args: argparse.Namespace,
+    scales: dict[str, ScaleDefinition],
+    observations: list[RatingObservation],
+) -> None:
+    """Reject a reused source identity whose immutable source labels differ."""
+    release = await conn.fetchrow(
+        """select release_version, source_publisher_name, source_license_url
+             from occupational_data_release where data_release_code = $1""",
+        args.release_code,
+    )
+    if release is not None and tuple(release) != (
+        args.release_version,
+        args.publisher,
+        args.license_url,
+    ):
+        raise ValueError("conflicting release identity")
+    sources = (
+        (
+            args.source_table_code,
+            args.source_table_name,
+            args.source_url,
+            args.source_sha256.lower(),
+            args.source_row_count,
+        ),
+        (
+            _SCALES_SOURCE_CODE,
+            "Scales Reference",
+            args.scales_url,
+            args.scales_sha256.lower(),
+            args.scales_row_count,
+        ),
+    )
+    for source_code, source_name, source_url, source_digest, source_rows in sources:
+        source = await conn.fetchrow(
+            """select source_table_name, source_artifact_url,
+                      source_artifact_sha256, source_row_count
+                 from occupational_source_table
+                where data_release_code = $1 and source_table_code = $2""",
+            args.release_code,
+            source_code,
+        )
+        if source is not None and tuple(source) != (
+            source_name,
+            source_url,
+            source_digest,
+            source_rows,
+        ):
+            raise ValueError(f"conflicting source-table identity: {source_code}")
+    expected = {
+        "scale": {
+            item.scale_id: (
+                _SCALES_SOURCE_CODE,
+                item.scale_name,
+                item.minimum_value,
+                item.maximum_value,
+            )
+            for item in scales.values()
+        },
+        "occupation": {
+            item.onetsoc_code: (item.occupation_title,) for item in observations
+        },
+        "element": {item.element_id: (item.element_name,) for item in observations},
+    }
+    queries = {
+        "scale": "select scale_id, source_table_code, scale_name, minimum_value, maximum_value from occupational_scale_definition where data_release_code = $1",
+        "occupation": "select onetsoc_code, occupation_title from occupational_classification_entry where data_release_code = $1",
+        "element": "select element_id, element_name from occupational_element_definition where data_release_code = $1",
+    }
+    for kind, query in queries.items():
+        for row in await conn.fetch(query, args.release_code):
+            source_id, *values = tuple(row)
+            if (
+                source_id in expected[kind]
+                and tuple(values) != expected[kind][source_id]
+            ):
+                raise ValueError(f"conflicting {kind} identity: {source_id}")
+
+
+async def import_ratings(args: argparse.Namespace) -> dict[str, object]:
+    """Validate one pinned source artifact, then transactionally UPSERT its rows."""
+    _validate_args(args)
+    digest = hashlib.sha256(args.ratings_file.read_bytes()).hexdigest()
+    if digest != args.source_sha256.lower():
+        raise ValueError("rating artifact SHA-256 mismatch")
+    scales_digest = hashlib.sha256(args.scales_file.read_bytes()).hexdigest()
+    if scales_digest != args.scales_sha256.lower():
+        raise ValueError("scales artifact SHA-256 mismatch")
+    scales = read_scale_file(args.scales_file)
+    if len(scales) != args.scales_row_count:
+        raise ValueError("scales artifact row-count mismatch")
+    observations = read_rating_file(args.ratings_file, scales)
+    if len(observations) != args.source_row_count:
+        raise ValueError("rating artifact row-count mismatch")
+    conn = await asyncpg.connect(args.target_dsn)
+    release_partition = f"occupational_rating_release_{hashlib.sha256(args.release_code.encode()).hexdigest()[:16]}"
+    source_partition = f"occupational_rating_source_{hashlib.sha256(f'{args.release_code}\0{args.source_table_code}'.encode()).hexdigest()[:16]}"
+    try:
+        async with conn.transaction():
+            await conn.execute(
+                "select pg_advisory_xact_lock(hashtextextended($1, 0))",
+                args.release_code,
+            )
+            await _reject_reference_conflicts(conn, args, scales, observations)
+            release_literal = await conn.fetchval(
+                "select quote_literal($1)", args.release_code
+            )
+            source_literal = await conn.fetchval(
+                "select quote_literal($1)", args.source_table_code
+            )
+            await conn.execute(
+                """insert into occupational_data_release
+                       (data_release_code, release_version, source_publisher_name, source_license_url)
+                   values ($1, $2, $3, $4)
+                   on conflict (data_release_code) do nothing""",
+                args.release_code,
+                args.release_version,
+                args.publisher,
+                args.license_url,
+            )
+            await conn.execute(
+                f"create table if not exists {release_partition} partition of occupational_rating_observation for values in ({release_literal}) partition by list (source_table_code)"
+            )
+            await conn.execute(
+                f"create table if not exists {source_partition} partition of {release_partition} for values in ({source_literal})"
+            )
+            await conn.execute(
+                """insert into occupational_source_table
+                       (data_release_code, source_table_code, source_table_name,
+                        source_artifact_url, source_artifact_sha256, source_row_count)
+                   values ($1, $2, $3, $4, $5, $6)
+                   on conflict (data_release_code, source_table_code) do nothing""",
+                args.release_code,
+                args.source_table_code,
+                args.source_table_name,
+                args.source_url,
+                digest,
+                len(observations),
+            )
+            await conn.execute(
+                """insert into occupational_source_table
+                       (data_release_code, source_table_code, source_table_name,
+                        source_artifact_url, source_artifact_sha256, source_row_count)
+                   values ($1, $2, 'Scales Reference', $3, $4, $5)
+                   on conflict (data_release_code, source_table_code) do nothing""",
+                args.release_code,
+                _SCALES_SOURCE_CODE,
+                args.scales_url,
+                scales_digest,
+                len(scales),
+            )
+            await conn.executemany(
+                """insert into occupational_scale_definition
+                       (data_release_code, source_table_code, scale_id, scale_name,
+                        minimum_value, maximum_value)
+                   values ($1, $2, $3, $4, $5, $6)
+                   on conflict (data_release_code, scale_id) do nothing""",
+                [
+                    (
+                        args.release_code,
+                        _SCALES_SOURCE_CODE,
+                        item.scale_id,
+                        item.scale_name,
+                        item.minimum_value,
+                        item.maximum_value,
+                    )
+                    for item in scales.values()
+                ],
+            )
+            await conn.executemany(
+                """insert into occupational_classification_entry
+                       (data_release_code, onetsoc_code, occupation_title)
+                   values ($1, $2, $3) on conflict (data_release_code, onetsoc_code) do nothing""",
+                [
+                    (args.release_code, code, title)
+                    for code, title in {
+                        item.onetsoc_code: item.occupation_title
+                        for item in observations
+                    }.items()
+                ],
+            )
+            await conn.executemany(
+                """insert into occupational_element_definition
+                       (data_release_code, element_id, element_name)
+                   values ($1, $2, $3) on conflict (data_release_code, element_id) do nothing""",
+                [
+                    (args.release_code, code, name)
+                    for code, name in {
+                        item.element_id: item.element_name for item in observations
+                    }.items()
+                ],
+            )
+            await conn.executemany(
+                """insert into occupational_rating_observation
+                       (data_release_code, source_table_code, onetsoc_code, element_id,
+                        scale_id, category_value, data_value, sample_size, standard_error,
+                        lower_ci_bound, upper_ci_bound, recommend_suppress, not_relevant,
+                        source_updated_date, domain_source_code)
+                   values ($1,$2,$3,$4,$5,$6,$7,$8,$9,$10,$11,$12,$13,$14,$15)
+                   on conflict on constraint occupational_rating_identity_key do nothing""",
+                [
+                    (
+                        args.release_code,
+                        args.source_table_code,
+                        item.onetsoc_code,
+                        item.element_id,
+                        item.scale_id,
+                        item.category_value,
+                        item.data_value,
+                        item.sample_size,
+                        item.standard_error,
+                        item.lower_ci_bound,
+                        item.upper_ci_bound,
+                        item.recommend_suppress,
+                        item.not_relevant,
+                        item.source_updated_date,
+                        item.domain_source_code,
+                    )
+                    for item in observations
+                ],
+            )
+    finally:
+        await conn.close()
+    return {
+        "release_code": args.release_code,
+        "source_table_code": args.source_table_code,
+        "imported_rows": len(observations),
+        "source_sha256": digest,
+        "scales_sha256": scales_digest,
+    }
+
+
+def main() -> None:
+    """Run the command-line importer and print aggregate, non-identifying evidence."""
+    print(
+        json.dumps(asyncio.run(import_ratings(_parser().parse_args())), sort_keys=True)
+    )
+
+
+if __name__ == "__main__":
+    main()
diff --git a/tests/test_import_onet_ratings.py b/tests/test_import_onet_ratings.py
new file mode 100644
index 000000000..d2243cce9
--- /dev/null
+++ b/tests/test_import_onet_ratings.py
@@ -0,0 +1,181 @@
+"""Contracts for the official O*NET occupation-rating CSV importer."""
+
+import asyncio
+import hashlib
+from datetime import date
+from decimal import Decimal
+from pathlib import Path
+from types import SimpleNamespace
+
+import pytest
+
+from scripts.import_onet_ratings import (
+    import_ratings,
+    read_rating_file,
+    read_scale_file,
+)
+
+
+def _write(path: Path, text: str) -> Path:
+    path.write_text(text, encoding="utf-8", newline="")
+    return path
+
+
+def test_official_csv_preserves_decimal_missingness_and_uncertainty(
+    tmp_path: Path,
+) -> None:
+    scales = read_scale_file(
+        _write(
+            tmp_path / "scales.csv",
+            "Scale ID,Scale Name,Minimum,Maximum\nIM,Importance,1,5\n",
+        )
+    )
+    ratings = read_rating_file(
+        _write(
+            tmp_path / "abilities.csv",
+            "O*NET-SOC Code,Title,Element ID,Element Name,Scale ID,Scale Name,Data Value,N,Standard Error,Lower CI Bound,Upper CI Bound,Recommend Suppress,Not Relevant,Date,Domain Source\n"
+            "15-1252.00,Synthetic occupation,1.A.1.a.1,Oral Comprehension,IM,Importance,4.10,120,0.08,3.94,4.26,N,,08/2026,Analyst\n",
+        ),
+        scales,
+        today=date(2026, 8, 27),
+    )
+
+    assert len(ratings) == 1
+    assert ratings[0].data_value == Decimal("4.10")
+    assert ratings[0].not_relevant is None
+    assert ratings[0].source_updated_date == date(2026, 8, 1)
+
+
+@pytest.mark.parametrize(
+    ("field", "value", "message"),
+    [
+        ("Data Value", "5.01", "outside scale"),
+        ("Standard Error", "-0.01", "standard error"),
+        ("Date", "09/2026", "future"),
+        ("Recommend Suppress", "maybe", "flag"),
+    ],
+)
+def test_invalid_source_measurement_fails_before_persistence(
+    tmp_path: Path,
+    field: str,
+    value: str,
+    message: str,
+) -> None:
+    scales = read_scale_file(
+        _write(
+            tmp_path / "scales.csv",
+            "Scale ID,Scale Name,Minimum,Maximum\nIM,Importance,1,5\n",
+        )
+    )
+    row = {
+        "O*NET-SOC Code": "15-1252.00",
+        "Title": "Synthetic occupation",
+        "Element ID": "1.A.1.a.1",
+        "Element Name": "Oral Comprehension",
+        "Scale ID": "IM",
+        "Scale Name": "Importance",
+        "Data Value": "4.10",
+        "N": "120",
+        "Standard Error": "0.08",
+        "Lower CI Bound": "3.94",
+        "Upper CI Bound": "4.26",
+        "Recommend Suppress": "N",
+        "Not Relevant": "",
+        "Date": "08/2026",
+        "Domain Source": "Analyst",
+    }
+    row[field] = value
+    path = tmp_path / "invalid.csv"
+    import csv
+
+    with path.open("w", encoding="utf-8", newline="") as handle:
+        writer = csv.DictWriter(handle, fieldnames=row)
+        writer.writeheader()
+        writer.writerow(row)
+
+    with pytest.raises(ValueError, match=message):
+        read_rating_file(path, scales, today=date(2026, 8, 27))
+
+
+def test_conflicting_reference_name_fails_closed(tmp_path: Path) -> None:
+    scales = read_scale_file(
+        _write(
+            tmp_path / "scales.csv",
+            "Scale ID,Scale Name,Minimum,Maximum\nIM,Importance,1,5\n",
+        )
+    )
+    path = _write(
+        tmp_path / "conflict.csv",
+        "O*NET-SOC Code,Title,Element ID,Element Name,Scale ID,Scale Name,Data Value,N,Standard Error,Lower CI Bound,Upper CI Bound,Recommend Suppress,Not Relevant,Date,Domain Source\n"
+        "15-1252.00,Synthetic occupation,1.A.1.a.1,Oral Comprehension,IM,Importance,4.10,,,,,N,,08/2026,Analyst\n"
+        "15-1252.00,Conflicting title,1.A.1.a.1,Oral Comprehension,IM,Importance,4.20,,,,,N,,08/2026,Analyst\n",
+    )
+
+    with pytest.raises(ValueError, match="conflicting occupation title"):
+        read_rating_file(path, scales, today=date(2026, 8, 27))
+
+
+def test_category_table_may_omit_not_relevant_without_inventing_false(
+    tmp_path: Path,
+) -> None:
+    scales = read_scale_file(
+        _write(
+            tmp_path / "scales.csv",
+            "Scale ID,Scale Name,Minimum,Maximum\nPT,Percent,0,100\n",
+        )
+    )
+    rows = read_rating_file(
+        _write(
+            tmp_path / "education.csv",
+            "O*NET-SOC Code,Title,Element ID,Element Name,Scale ID,Scale Name,Category,Data Value,N,Standard Error,Lower CI Bound,Upper CI Bound,Recommend Suppress,Date,Domain Source\n"
+            "15-1252.00,Synthetic occupation,2.D.1,Education,PT,Percent,6,42.50,100,1.2,40.1,44.9,N,08/2026,Incumbent\n",
+        ),
+        scales,
+        today=date(2026, 8, 27),
+    )
+
+    assert rows[0].category_value == 6
+    assert rows[0].not_relevant is None
+
+
+def test_scales_digest_fails_before_database_connection(
+    tmp_path: Path,
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    scales = _write(
+        tmp_path / "scales.csv",
+        "Scale ID,Scale Name,Minimum,Maximum\nIM,Importance,1,5\n",
+    )
+    ratings = _write(
+        tmp_path / "abilities.csv",
+        "O*NET-SOC Code,Title,Element ID,Element Name,Scale ID,Scale Name,Data Value,N,Standard Error,Lower CI Bound,Upper CI Bound,Recommend Suppress,Not Relevant,Date,Domain Source\n"
+        "15-1252.00,Synthetic occupation,1.A.1.a.1,Oral Comprehension,IM,Importance,4.10,120,0.08,3.94,4.26,N,,08/2026,Analyst\n",
+    )
+    connected = False
+
+    async def fake_connect(_dsn: str) -> None:
+        nonlocal connected
+        connected = True
+
+    monkeypatch.setattr("scripts.import_onet_ratings.asyncpg.connect", fake_connect)
+    args = SimpleNamespace(
+        target_dsn="postgresql://unused",
+        release_code="onet-31.0-synthetic",
+        release_version="31.0-synthetic",
+        source_table_code="abilities",
+        source_table_name="Abilities",
+        source_url="https://example.test/abilities.csv",
+        source_sha256=hashlib.sha256(ratings.read_bytes()).hexdigest(),
+        source_row_count=1,
+        publisher="Synthetic publisher",
+        license_url="https://example.test/license",
+        scales_file=scales,
+        scales_url="https://example.test/scales.csv",
+        scales_sha256="0" * 64,
+        scales_row_count=1,
+        ratings_file=ratings,
+    )
+
+    with pytest.raises(ValueError, match="scales artifact SHA-256 mismatch"):
+        asyncio.run(import_ratings(args))
+    assert connected is False
diff --git a/tests/test_schema.py b/tests/test_schema.py
index 01f25b5b9..fe8c5d1ae 100644
--- a/tests/test_schema.py
+++ b/tests/test_schema.py
@@ -15,10 +15,12 @@
 from __future__ import annotations
 
 import asyncio
+import hashlib
 import os
 import uuid
 from decimal import Decimal
 from pathlib import Path
+from types import SimpleNamespace
 from urllib.parse import urlsplit, urlunsplit
 
 import asyncpg
@@ -27,6 +29,7 @@
 import pytest
 
 from backend.app.post_chat_ingestion import gather_global_chat_sources
+from scripts.import_onet_ratings import import_ratings
 
 _ADMIN_DSN = os.environ.get(
     "LINEAGEWEAVE_TEST_POSTGRES_ADMIN_DSN", "postgresql://localhost/postgres"
@@ -401,6 +404,51 @@ def test_onet_rating_store_partitions_upserts_and_rejects_invalid_error(schema_d
         cur.execute("rollback to savepoint invalid_standard_error")
 
 
+def test_onet_rating_importer_is_idempotent_against_postgresql(
+    schema_db,
+    tmp_path: Path,
+) -> None:
+    """A pinned synthetic artifact imports twice as one exact observation."""
+    scales = tmp_path / "scales.csv"
+    scales.write_text(
+        "Scale ID,Scale Name,Minimum,Maximum\nIM,Importance,1,5\n",
+        encoding="utf-8",
+    )
+    ratings = tmp_path / "abilities.csv"
+    ratings.write_text(
+        "O*NET-SOC Code,Title,Element ID,Element Name,Scale ID,Scale Name,Data Value,N,Standard Error,Lower CI Bound,Upper CI Bound,Recommend Suppress,Not Relevant,Date,Domain Source\n"
+        "15-1252.00,Synthetic occupation,1.A.1.a.1,Oral Comprehension,IM,Importance,4.10,120,0.08,3.94,4.26,N,,08/2026,Analyst\n",
+        encoding="utf-8",
+    )
+    args = SimpleNamespace(
+        target_dsn=schema_db.dsn,
+        release_code="onet-31.0-synthetic",
+        release_version="31.0-synthetic",
+        source_table_code="abilities",
+        source_table_name="Abilities",
+        source_url="https://example.test/abilities.csv",
+        source_sha256=hashlib.sha256(ratings.read_bytes()).hexdigest(),
+        source_row_count=1,
+        publisher="Synthetic publisher",
+        license_url="https://example.test/license",
+        scales_file=scales,
+        scales_url="https://example.test/scales.csv",
+        scales_sha256=hashlib.sha256(scales.read_bytes()).hexdigest(),
+        scales_row_count=1,
+        ratings_file=ratings,
+    )
+
+    assert asyncio.run(import_ratings(args))["imported_rows"] == 1
+    assert asyncio.run(import_ratings(args))["imported_rows"] == 1
+    with schema_db.cursor() as cur:
+        cur.execute(
+            """select count(*), min(data_value), bool_or(not_relevant is null)
+                 from occupational_rating_observation
+                where data_release_code = 'onet-31.0-synthetic'"""
+        )
+        assert cur.fetchone() == (1, Decimal("4.10"), True)
+
+
 def test_global_ask_evidence_search_indexes_exist_on_normalized_tables(schema_db) -> None:
     """The real PostgreSQL schema owns all nine evidence-search indexes."""
     with schema_db.cursor() as cur:

From 76b3ef2a221d1b75280ca8ccdd8d3e854b2a362c Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Thu, 27 Aug 2026 04:13:23 +0900
Subject: [PATCH 02/17] fix(ontology): accept sparse O*NET model profiles

---
 scripts/import_onet_ratings.py    | 15 +++++----------
 tests/test_import_onet_ratings.py | 24 ++++++++++++++++++++++++
 2 files changed, 29 insertions(+), 10 deletions(-)

diff --git a/scripts/import_onet_ratings.py b/scripts/import_onet_ratings.py
index a09cbe4bc..ed645bf5f 100644
--- a/scripts/import_onet_ratings.py
+++ b/scripts/import_onet_ratings.py
@@ -24,11 +24,6 @@
     "Scale ID",
     "Scale Name",
     "Data Value",
-    "N",
-    "Standard Error",
-    "Lower CI Bound",
-    "Upper CI Bound",
-    "Recommend Suppress",
     "Date",
     "Domain Source",
 }
@@ -185,12 +180,12 @@ def read_rating_file(
         data_value = _decimal(row["Data Value"], "data value")
         if not scale.minimum_value <= data_value <= scale.maximum_value:  # type: ignore[operator]
             raise ValueError(f"data value outside scale {scale_id}")
-        sample_size = _integer(row["N"], "sample size", optional=True)
+        sample_size = _integer(row.get("N", ""), "sample size", optional=True)
         standard_error = _decimal(
-            row["Standard Error"], "standard error", optional=True
+            row.get("Standard Error", ""), "standard error", optional=True
         )
-        lower = _decimal(row["Lower CI Bound"], "lower CI bound", optional=True)
-        upper = _decimal(row["Upper CI Bound"], "upper CI bound", optional=True)
+        lower = _decimal(row.get("Lower CI Bound", ""), "lower CI bound", optional=True)
+        upper = _decimal(row.get("Upper CI Bound", ""), "upper CI bound", optional=True)
         if sample_size is not None and sample_size <= 0:
             raise ValueError("sample size must be positive")
         if standard_error is not None and standard_error < 0:
@@ -225,7 +220,7 @@ def read_rating_file(
                 lower_ci_bound=lower,
                 upper_ci_bound=upper,
                 recommend_suppress=_flag(
-                    row["Recommend Suppress"], "recommend suppress"
+                    row.get("Recommend Suppress", ""), "recommend suppress"
                 ),
                 not_relevant=_flag(row.get("Not Relevant", ""), "not relevant"),
                 source_updated_date=_updated_month(row["Date"], observed_today),
diff --git a/tests/test_import_onet_ratings.py b/tests/test_import_onet_ratings.py
index d2243cce9..d09852f76 100644
--- a/tests/test_import_onet_ratings.py
+++ b/tests/test_import_onet_ratings.py
@@ -138,6 +138,30 @@ def test_category_table_may_omit_not_relevant_without_inventing_false(
     assert rows[0].not_relevant is None
 
 
+def test_machine_generated_profile_keeps_unpublished_uncertainty_missing(
+    tmp_path: Path,
+) -> None:
+    scales = read_scale_file(
+        _write(
+            tmp_path / "scales.csv",
+            "Scale ID,Scale Name,Minimum,Maximum\nDR,Distinctiveness Rank,0,7\n",
+        )
+    )
+    rows = read_rating_file(
+        _write(
+            tmp_path / "work_styles.csv",
+            "O*NET-SOC Code,Title,Element ID,Element Name,Scale ID,Scale Name,Data Value,Date,Domain Source\n"
+            "15-1252.00,Synthetic occupation,1.D.1.a,Innovation,DR,Distinctiveness Rank,7.00,08/2026,AI/Expert\n",
+        ),
+        scales,
+        today=date(2026, 8, 27),
+    )
+
+    assert rows[0].sample_size is None
+    assert rows[0].standard_error is None
+    assert rows[0].recommend_suppress is None
+
+
 def test_scales_digest_fails_before_database_connection(
     tmp_path: Path,
     monkeypatch: pytest.MonkeyPatch,

From 3ee8b90d4bf7c5d381faed5145455af97bfa6a6f Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Thu, 27 Aug 2026 04:41:07 +0900
Subject: [PATCH 03/17] docs(gaps): refresh O*NET delivery stack

---
 docs/product-technical-gap-baseline.md | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/docs/product-technical-gap-baseline.md b/docs/product-technical-gap-baseline.md
index 1735a65ff..51638152f 100644
--- a/docs/product-technical-gap-baseline.md
+++ b/docs/product-technical-gap-baseline.md
@@ -1,7 +1,7 @@
 # Product & Technical Gap Baseline
 
 > Current queue overlay: 2026-08-27 KST. Protected `main` was
-> `ff7431bd1851c03e737808d22c6a2d43968582f9`; 35 PRs and 10 issues were
+> `ff7431bd1851c03e737808d22c6a2d43968582f9`; 38 PRs and 10 issues were
 > open. This overlay supersedes the older queue count and exact-head table
 > below, which remain historical evidence. Re-fetch the head, checks, reviews,
 > threads, applicable rulesets, and merge SHA immediately before any lifecycle
@@ -35,9 +35,11 @@ explicit unavailable state, not a reason to infer mappings from labels.
 
 | PR | Exact observed head | Base | Observed gate state |
 |---:|---|---|---|
-| #732 | `1c622545` | `feat/onet-31-content-model-ontology` | unstable; full tests in progress; frontend, CodeRabbit, and Devin Review passed; one informational review thread open |
-| #731 | `2c848373` | `feat/soc-2018-full-hierarchy` | unstable; full tests, frontend, and Devin Review queued |
-| #724 | `f5ee37b0` | `feat/io-occupational-taxonomy` | clean; full tests, frontend, CodeRabbit, and Devin Review passed; parent-first stack gate remains |
+| #735 | `56d8cbfa` | `feat/onet-occupation-ratings-contract` | clean immediately after exact parent merge; hosted checks had not yet registered on this head, so this is not merge-readiness evidence |
+| #734 | `0dedf33b` | `feat/onet-content-model-linkages` | unstable; two hosted gates pending after exact parent reconciliation |
+| #732 | `3e30402d` | `feat/onet-31-content-model-ontology` | unstable; one hosted gate pending after exact parent reconciliation |
+| #731 | `9654bb8b` | `feat/soc-2018-full-hierarchy` | unstable; one hosted gate pending after exact parent reconciliation |
+| #724 | `fdec8f65` | `feat/io-occupational-taxonomy` | clean; hosted gates passed; parent-first stack gate remains |
 | #719 | `6ee2278a` | `feat/io-psych-construct-ontology` | unstable; full tests, frontend, CodeRabbit, and Devin Review passed |
 | #718 | `2723fea3` | `feat/fja-worker-function-ontology` | unstable; full tests, frontend, CodeRabbit, and Devin Review passed |
 | #717 | `771a8edf` | `feat/voice-of-x-complete-taxonomy` | unstable; 1 pending check(s) |

From 1b229c646a1dc10046e127dd7a7fa0a76da08e5a Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Thu, 27 Aug 2026 04:42:26 +0900
Subject: [PATCH 04/17] fix(test): preserve credentialed importer DSN

---
 tests/test_schema.py | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/tests/test_schema.py b/tests/test_schema.py
index 7770690c7..d6cf67712 100644
--- a/tests/test_schema.py
+++ b/tests/test_schema.py
@@ -434,7 +434,9 @@ def test_onet_rating_importer_is_idempotent_against_postgresql(
         encoding="utf-8",
     )
     args = SimpleNamespace(
-        target_dsn=schema_db.dsn,
+        target_dsn=urlunsplit(
+            urlsplit(_ADMIN_DSN)._replace(path=f"/{schema_db.info.dbname}")
+        ),
         release_code="onet-31.0-synthetic",
         release_version="31.0-synthetic",
         source_table_code="abilities",

From c0f31f115301fc17f75410a28a9a8bb702d77ffa Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Thu, 27 Aug 2026 04:51:21 +0900
Subject: [PATCH 05/17] fix(import): reject password-only source URL userinfo

---
 scripts/import_onet_ratings.py    | 7 ++++++-
 tests/test_import_onet_ratings.py | 6 ++++++
 2 files changed, 12 insertions(+), 1 deletion(-)

diff --git a/scripts/import_onet_ratings.py b/scripts/import_onet_ratings.py
index bcb518dd4..85f01af49 100644
--- a/scripts/import_onet_ratings.py
+++ b/scripts/import_onet_ratings.py
@@ -276,7 +276,12 @@ def _validate_args(args: argparse.Namespace) -> None:
             raise ValueError(f"{field} must be positive")
     for field in ("source_url", "scales_url", "license_url"):
         parsed = urlsplit(str(getattr(args, field)))
-        if parsed.scheme != "https" or not parsed.hostname or parsed.username:
+        if (
+            parsed.scheme != "https"
+            or not parsed.hostname
+            or parsed.username is not None
+            or parsed.password is not None
+        ):
             raise ValueError(f"{field} must be an HTTPS URL without userinfo")
     for field in ("ratings_file", "scales_file"):
         if not getattr(args, field).is_file():
diff --git a/tests/test_import_onet_ratings.py b/tests/test_import_onet_ratings.py
index 3619744f8..4007d4e98 100644
--- a/tests/test_import_onet_ratings.py
+++ b/tests/test_import_onet_ratings.py
@@ -205,6 +205,12 @@ async def fake_connect(_dsn: str) -> None:
         ratings_file=ratings,
     )
 
+    args.source_url = "https://:secret@example.test/abilities.csv"
+    with pytest.raises(ValueError, match="without userinfo"):
+        asyncio.run(import_ratings(args))
+    assert connected is False
+
+    args.source_url = "https://example.test/abilities.csv"
     with pytest.raises(ValueError, match="scales artifact SHA-256 mismatch"):
         asyncio.run(import_ratings(args))
     assert connected is False

From 9bc9e93d20fc10191705613356b7c8ca101c5db7 Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Thu, 27 Aug 2026 04:54:57 +0900
Subject: [PATCH 06/17] feat(ontology): expose occupation rating evidence

---
 ARCHITECTURE.md                             |   1 +
 CHANGELOG.md                                |   5 +
 backend/app/main.py                         |  29 +++-
 backend/app/occupation_rating_ingestion.py  | 127 +++++++++++++++
 docs/adr/0258-occupation-rating-read-api.md |  49 ++++++
 docs/adr/README.md                          |   1 +
 docs/product-requirements.md                |  13 ++
 docs/product-technical-gap-baseline.md      |   2 +-
 tests/test_occupation_rating_ingestion.py   | 162 ++++++++++++++++++++
 tests/test_schema.py                        |  20 +++
 10 files changed, 407 insertions(+), 2 deletions(-)
 create mode 100644 backend/app/occupation_rating_ingestion.py
 create mode 100644 docs/adr/0258-occupation-rating-read-api.md
 create mode 100644 tests/test_occupation_rating_ingestion.py

diff --git a/ARCHITECTURE.md b/ARCHITECTURE.md
index 55e00b78d..6a24bce13 100644
--- a/ARCHITECTURE.md
+++ b/ARCHITECTURE.md
@@ -80,6 +80,7 @@ flowchart LR
 | `temporal_expressions.py` | Pure Korean relative-time resolver for Global Ask (ADR 0150) |
 | `ask_time_axis.py` | Event-time vs ingestion-time clock choice for that window (ADR 0202) |
 | `ontology.py` | Loads the governed Turtle source tree (`lineageweave-kg.ttl` plus generated fragments), the formal OWL 2/RDFS/SKOS vocabulary for the Knowledge Graph's node/edge types, source taxonomies, and published O*NET linkages (ADR 0004, ADR 0252, ADR 0255, ADR 0256) |
+| `backend/app/occupation_rating_ingestion.py` | Projects authenticated, bounded occupation-rating source evidence with exact decimal and artifact provenance semantics (ADR 0258) |
 | `ontology_neighborhood.py` | Bounded typed ontology/provenance neighborhood (ADR 0184); PostgreSQL stays authoritative, OWL subclass is not an instance edge |
 | `ontology_source_cursor.py` | Opaque HMAC source-window continuation (ADR 0124); keyset pagination, never OFFSET |
 | `period_report.py` | Fit GRM/GPCM on persisted IRT rows, FIPC-select, EAP-score a period (ADR 0003 slice 3; Bock & Mislevy, 1982) |
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 79869185c..15193b4e0 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -8,6 +8,11 @@ All notable changes to this project are documented here. Format follows
 
 ### Added
 
+- Authenticated occupation profiles can now read one exact imported
+  release/source with rating and scale artifact provenance, exact decimal
+  strings, uncertainty, suppression/relevance flags, and explicit unavailable
+  versus empty outcomes. The endpoint derives no ranking or recommendation
+  (ADR 0258).
 - O*NET occupation-rating source evidence now has a replay-safe PostgreSQL
   contract with normalized release, source-table, scale, occupation, element,
   and observation tables. Exact release/source LIST partitions fail closed;
diff --git a/backend/app/main.py b/backend/app/main.py
index 6457bbde1..60dd9e1a5 100644
--- a/backend/app/main.py
+++ b/backend/app/main.py
@@ -30,7 +30,7 @@
 
 import asyncpg
 import redis.asyncio as redis
-from fastapi import Depends, FastAPI, HTTPException, Query, status
+from fastapi import Depends, FastAPI, HTTPException, Path, Query, status
 from fastapi.middleware.cors import CORSMiddleware
 from pydantic import BaseModel
 
@@ -98,6 +98,7 @@
     upsert_commitment_ticket,
 )
 from backend.app.operations_dashboard import fetch_operations_dashboard
+from backend.app.occupation_rating_ingestion import fetch_occupation_ratings
 from backend.app.keyman_ingestion import ingest_post_keymen
 from backend.app.knowledge_graph import (
     corporate_entity_exists,
@@ -2275,6 +2276,32 @@ async def read_ontology_neighborhood(
     return payload
 
 
+@app.get("/api/occupations/{onetsoc_code}/ratings")
+async def read_occupation_ratings(
+    onetsoc_code: str = Path(..., pattern=r"^[0-9]{2}-[0-9]{4}\.[0-9]{2}$"),
+    data_release_code: str = Query(
+        ..., min_length=1, max_length=63, pattern=r"^[a-z0-9][a-z0-9.-]*$"
+    ),
+    source_table_code: str = Query(
+        ..., min_length=1, max_length=63, pattern=r"^[a-z][a-z0-9_]*$"
+    ),
+    limit: int = Query(100, ge=1, le=500),
+    offset: int = Query(0, ge=0, le=10000),
+    _account: CurrentAccount = Depends(get_current_account),
+    pool: asyncpg.Pool = Depends(get_pool),
+) -> dict[str, object]:
+    """Return one authenticated, provenance-bearing occupation source profile."""
+    async with pool.acquire() as conn:
+        return await fetch_occupation_ratings(
+            conn,
+            data_release_code=data_release_code,
+            source_table_code=source_table_code,
+            onetsoc_code=onetsoc_code,
+            limit=limit,
+            offset=offset,
+        )
+
+
 @app.get("/api/posts/{post_id}/counterparties")
 async def read_post_counterparties(
     post_id: str,
diff --git a/backend/app/occupation_rating_ingestion.py b/backend/app/occupation_rating_ingestion.py
new file mode 100644
index 000000000..320fe248d
--- /dev/null
+++ b/backend/app/occupation_rating_ingestion.py
@@ -0,0 +1,127 @@
+"""Read exact imported occupation ratings without deriving a score or weight."""
+
+from __future__ import annotations
+
+from decimal import Decimal
+from typing import Any, Protocol
+
+
+class RatingReadConnection(Protocol):
+    """Small asyncpg-compatible surface used by the rating read projection."""
+
+    async def fetchrow(self, query: str, *args: object) -> Any:
+        """Return one row or ``None``."""
+
+    async def fetch(self, query: str, *args: object) -> list[Any]:
+        """Return ordered rows."""
+
+
+def _decimal_text(value: Decimal | None) -> str | None:
+    """Return the exact database decimal representation or honest absence."""
+    return str(value) if value is not None else None
+
+
+async def fetch_occupation_ratings(
+    conn: RatingReadConnection,
+    *,
+    data_release_code: str,
+    source_table_code: str,
+    onetsoc_code: str,
+    limit: int,
+    offset: int,
+) -> dict[str, object]:
+    """Return one bounded source profile and explicit artifact availability."""
+    source = await conn.fetchrow(
+        """select rating_source.source_table_name,
+                  rating_source.source_artifact_url,
+                  rating_source.source_artifact_sha256,
+                  rating_source.source_row_count,
+                  scale_source.source_artifact_url as scale_artifact_url,
+                  scale_source.source_artifact_sha256 as scale_artifact_sha256,
+                  scale_source.source_row_count as scale_source_row_count
+             from occupational_source_table rating_source
+             left join occupational_source_table scale_source
+               on scale_source.data_release_code = rating_source.data_release_code
+              and scale_source.source_table_code = 'scales_reference'
+            where rating_source.data_release_code = $1
+              and rating_source.source_table_code = $2""",
+        data_release_code,
+        source_table_code,
+    )
+    if source is None:
+        return {
+            "data_release_code": data_release_code,
+            "source_table_code": source_table_code,
+            "onetsoc_code": onetsoc_code,
+            "source_available": False,
+            "source": None,
+            "items": [],
+            "next_offset": None,
+        }
+    rows = await conn.fetch(
+        """select observation.element_id, element.element_name,
+                  observation.scale_id, scale.scale_name,
+                  scale.minimum_value, scale.maximum_value,
+                  observation.category_value, observation.data_value,
+                  observation.sample_size, observation.standard_error,
+                  observation.lower_ci_bound, observation.upper_ci_bound,
+                  observation.recommend_suppress, observation.not_relevant,
+                  observation.source_updated_month, observation.domain_source_code
+             from occupational_rating_observation observation
+             join occupational_element_definition element
+               on element.data_release_code = observation.data_release_code
+              and element.element_id = observation.element_id
+             join occupational_scale_definition scale
+               on scale.data_release_code = observation.data_release_code
+              and scale.scale_id = observation.scale_id
+            where observation.data_release_code = $1
+              and observation.source_table_code = $2
+              and observation.onetsoc_code = $3
+            order by observation.element_id, observation.scale_id,
+                     observation.category_value nulls first
+            limit $4 offset $5""",
+        data_release_code,
+        source_table_code,
+        onetsoc_code,
+        limit + 1,
+        offset,
+    )
+    page = rows[:limit]
+    items = [
+        {
+            "element_id": row["element_id"],
+            "element_name": row["element_name"],
+            "scale_id": row["scale_id"],
+            "scale_name": row["scale_name"],
+            "minimum_value": _decimal_text(row["minimum_value"]),
+            "maximum_value": _decimal_text(row["maximum_value"]),
+            "category_value": row["category_value"],
+            "data_value": _decimal_text(row["data_value"]),
+            "sample_size": row["sample_size"],
+            "standard_error": _decimal_text(row["standard_error"]),
+            "lower_ci_bound": _decimal_text(row["lower_ci_bound"]),
+            "upper_ci_bound": _decimal_text(row["upper_ci_bound"]),
+            "recommend_suppress": row["recommend_suppress"],
+            "not_relevant": row["not_relevant"],
+            "source_updated_month": row["source_updated_month"],
+            "domain_source_code": row["domain_source_code"],
+        }
+        for row in page
+    ]
+    return {
+        "data_release_code": data_release_code,
+        "source_table_code": source_table_code,
+        "onetsoc_code": onetsoc_code,
+        "source_available": True,
+        "source": {
+            "source_table_name": source["source_table_name"],
+            "source_artifact_url": source["source_artifact_url"],
+            "source_artifact_sha256": source["source_artifact_sha256"],
+            "source_row_count": source["source_row_count"],
+            "scale_artifact_url": source["scale_artifact_url"],
+            "scale_artifact_sha256": source["scale_artifact_sha256"],
+            "scale_source_row_count": source["scale_source_row_count"],
+        },
+        "items": items,
+        "next_offset": offset + limit if len(rows) > limit else None,
+    }
diff --git a/docs/adr/0258-occupation-rating-read-api.md b/docs/adr/0258-occupation-rating-read-api.md
new file mode 100644
index 000000000..a22f5bab1
--- /dev/null
+++ b/docs/adr/0258-occupation-rating-read-api.md
@@ -0,0 +1,49 @@
+# ADR 0258: Authenticated occupation-rating source read API
+
+**Status:** Accepted  
+**Date:** 2026-08-27  
+**Extends:** ADR 0120, ADR 0184, ADR 0257
+
+## Context
+
+ADR 0257 preserves released occupation-to-element observations, but a database
+import alone does not let a product user inspect what a job profile says. A
+read contract must distinguish an unimported source from an imported source
+with no row for one occupation, preserve low-precision and not-relevant flags,
+and avoid presenting a published rating as a local weight or recommendation.
+
+## Decision
+
+1. Add an authenticated, read-only occupation-rating endpoint. O*NET source
+   observations are licensed public reference data and are not tenant records;
+   any authenticated LineageWeave account may read an imported artifact.
+2. Require exact release, source-table, and O*NET-SOC codes. Return
+   `source_available=false` when that pinned artifact is not imported; return
+   `source_available=true` with an empty item list when it is imported but has
+   no observation for the requested occupation.
+3. Return the rating and Scales Reference artifact URLs, SHA-256 values, and
+   row counts. Every observation retains element/scale identity, declared
+   bounds, optional category, exact decimal strings, sample/error/interval,
+   suppression, relevance, source month, and domain source.
+4. Order by element, scale, and category and use bounded offset pagination.
+   Per-occupation source partitions bound this projection; a cursor needs a
+   later decision only if measured production latency requires it.
+5. Do not aggregate, rank, normalize, infer person traits, or recommend an
+   occupation. Suppressed values remain visible with the suppression flag so a
+   user can audit the source without mistaking low precision for absence.
+6. API and frontend copy describe the evidence and the user's next action,
+   never importer, partition, model-provider, or orchestration internals.
+
+## Consequences
+
+The semantic layer gains an honest product read boundary without duplicating
+psychometric arithmetic. An accessible UI and its Storybook states remain a
+separate delivery step after this API has authenticated runtime evidence.
+
+## References
+
+National Center for O*NET Development. (2026). *O*NET 31.0 database* [Data
+set]. https://www.onetcenter.org/database.html
+
+PostgreSQL Global Development Group. (2026). *PostgreSQL 18 documentation:
+Queries—limit and offset*. https://www.postgresql.org/docs/current/queries-limit.html
diff --git a/docs/adr/README.md b/docs/adr/README.md
index 74e74592c..1a36fa1a8 100644
--- a/docs/adr/README.md
+++ b/docs/adr/README.md
@@ -34,6 +34,7 @@ decision from them.
 | [`SOC_2018_HIERARCHY_REFERENCES.md`](../doctoring/SOC_2018_HIERARCHY_REFERENCES.md) | [0252](0252-complete-2018-soc-hierarchy.md) |
 | [`ONET_31_LINKAGE_REFERENCES.md`](../doctoring/ONET_31_LINKAGE_REFERENCES.md) | [0256](0256-onet-content-model-published-linkages.md) |
 | [`ONET_RATING_STORE_REFERENCES.md`](../doctoring/ONET_RATING_STORE_REFERENCES.md) | [0257](0257-onet-occupation-rating-observation-store.md) |
+| Occupation-rating authenticated read projection | [0258](0258-occupation-rating-read-api.md) |
 
 [0011](0011-prov-o-standard-relations.md) and [0065](0065-prov-o-provenance-boundary.md) cite the dated W3C PROV-O and PROV-DM Recommendations (https://www.w3.org/TR/2013/REC-prov-o-20130430/ and https://www.w3.org/TR/2013/REC-prov-dm-20130430/).
 
diff --git a/docs/product-requirements.md b/docs/product-requirements.md
index 7e3e116e5..d84b284b0 100644
--- a/docs/product-requirements.md
+++ b/docs/product-requirements.md
@@ -148,6 +148,19 @@ repeated null-category UPSERT is idempotent.
 API, UI, and derived modeling remain unavailable until separate accepted
 delivery records.
 
+### PRD-FR-2E — Occupation-rating evidence read
+
+- Let an authenticated user open one exact release/source/occupation profile
+  with both rating and scale artifact provenance (ADR 0258).
+- Distinguish an unavailable imported source from an available source with no
+  observation for the occupation.
+- Preserve exact decimal text, uncertainty, suppression, relevance, source
+  month, domain source, and declared bounds; derive no ranking or recommendation.
+
+Acceptance: invalid identifiers and unbounded pages are rejected; an unavailable
+source never appears as a negative profile; pagination is deterministic; and a
+suppressed observation retains its value and warning flag together.
+
 ### PRD-FR-3 — Bounded ontology exploration
 
 - Apply RBAC/ABAC, source eligibility, and knowledge cutoff before graph
diff --git a/docs/product-technical-gap-baseline.md b/docs/product-technical-gap-baseline.md
index 51638152f..bd7e5a884 100644
--- a/docs/product-technical-gap-baseline.md
+++ b/docs/product-technical-gap-baseline.md
@@ -29,7 +29,7 @@ explicit unavailable state, not a reason to infer mappings from labels.
 | Occupation-to-construct relations | ADR 0257 defines a candidate 3NF, release/source-partitioned immutable observation store and deterministic pinned-CSV importer preserving value, optional category, sample/error/CI, suppression, relevance, exact `MM/YYYY` source update month, source digest, and domain provenance. The official O*NET 31.0 Abilities file (94,640 rows, 910 occupations, 52 elements; SHA-256 `7e9cd79791ce6014e1d26d0a449ae5b1e7aa7ef52d39b3934c3bb8d438104b88`) and all 33 Scales Reference rows (SHA-256 `bcba23858ce21ecaacbde303a8993e35d46724b4afb8c9ec2b10e04f42adcfc9`) imported into a throwaway local PostgreSQL database with all 94,640 observations, 55 suppression flags, 7,572 not-relevant flags, and source months from `12/2004` through `08/2026`; every scale retained `scales_reference` artifact provenance, the database was dropped afterward, and no corpus is committed or claimed deployed | Pass exact-head review/checks and protected merge; validate and import every selected official rating artifact through an authorized runtime, returning only aggregate evidence; never invent or locally normalize a weight |
 | Job-family and job-series semantics | No authoritative employer-specific job architecture is present | Define an organization-neutral import contract that preserves the authorized source hierarchy and distinguishes standard occupation codes from employer job families/series; no label-based binding |
 | Temporal and multilevel interpretation | Static vocabulary only; no person-level inference is asserted | Version valid and transaction time, preserve occupation/organization/unit nesting and multiple membership, and require TEPP or the owning Rust psychometric service before any calibrated temporal or multilevel result |
-| Product consumption | The read model has no persisted semantic-layer consumer or authenticated UI evidence | Add a provenance-bearing API and accessible ontology exploration flow, then verify synthetic Storybook edge states plus authenticated aggregate runtime evidence without exposing identifying records |
+| Product consumption | ADR 0258 defines a candidate authenticated occupation-rating API that distinguishes unavailable source from empty profile and returns exact decimal/uncertainty/warning/artifact provenance without ranking; accessible UI and authenticated runtime evidence remain absent | Pass exact-head review/checks and protected merge; add the accessible exploration UI, synthetic Storybook unavailable/empty/suppressed/paginated states, screenshots, and authenticated aggregate runtime evidence without exposing identifying records |
 
 ### Current exact-head PR queue
 
diff --git a/tests/test_occupation_rating_ingestion.py b/tests/test_occupation_rating_ingestion.py
new file mode 100644
index 000000000..3f2244d7f
--- /dev/null
+++ b/tests/test_occupation_rating_ingestion.py
@@ -0,0 +1,162 @@
+"""Tests for the provenance-bearing occupation-rating read projection."""
+
+import asyncio
+from decimal import Decimal
+
+from backend.app.main import read_occupation_ratings
+from backend.app.occupation_rating_ingestion import fetch_occupation_ratings
+
+
+class FakeConnection:
+    """Minimal ordered asyncpg stand-in for one projection query."""
+
+    def __init__(self, source, rows=()):
+        self.source = source
+        self.rows = list(rows)
+        self.fetch_called = False
+
+    async def fetchrow(self, _query: str, *_args: object):
+        """Return configured source metadata."""
+        return self.source
+
+    async def fetch(self, _query: str, *_args: object):
+        """Return configured observation rows."""
+        self.fetch_called = True
+        return self.rows
+
+
+class FakeAcquire:
+    """Async pool-acquire context for route wiring."""
+
+    def __init__(self, conn: FakeConnection):
+        self.conn = conn
+
+    async def __aenter__(self) -> FakeConnection:
+        """Return the configured connection."""
+        return self.conn
+
+    async def __aexit__(self, *_args: object) -> None:
+        """Release without external state."""
+
+
+class FakePool:
+    """Minimal pool exposing one acquisition context."""
+
+    def __init__(self, conn: FakeConnection):
+        self.conn = conn
+
+    def acquire(self) -> FakeAcquire:
+        """Return one deterministic acquisition context."""
+        return FakeAcquire(self.conn)
+
+
+def test_unimported_source_is_not_an_empty_observed_profile() -> None:
+    conn = FakeConnection(None)
+
+    result = asyncio.run(
+        fetch_occupation_ratings(
+            conn,
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+            onetsoc_code="15-1252.00",
+            limit=100,
+            offset=0,
+        )
+    )
+
+    assert result["source_available"] is False
+    assert result["items"] == []
+    assert conn.fetch_called is False
+
+
+def test_rating_projection_preserves_exact_decimal_and_warning_flags() -> None:
+    source = {
+        "source_table_name": "Abilities",
+        "source_artifact_url": "https://example.test/abilities.csv",
+        "source_artifact_sha256": "a" * 64,
+        "source_row_count": 2,
+        "scale_artifact_url": "https://example.test/scales.csv",
+        "scale_artifact_sha256": "b" * 64,
+        "scale_source_row_count": 33,
+    }
+    row = {
+        "element_id": "1.A.1.a.1",
+        "element_name": "Oral Comprehension",
+        "scale_id": "IM",
+        "scale_name": "Importance",
+        "minimum_value": Decimal("1.00"),
+        "maximum_value": Decimal("5.00"),
+        "category_value": None,
+        "data_value": Decimal("4.10"),
+        "sample_size": 8,
+        "standard_error": Decimal("0.1830"),
+        "lower_ci_bound": Decimal("3.7414"),
+        "upper_ci_bound": Decimal("4.4586"),
+        "recommend_suppress": True,
+        "not_relevant": None,
+        "source_updated_month": "08/2026",
+        "domain_source_code": "Analyst",
+    }
+    conn = FakeConnection(source, (row, row))
+
+    result = asyncio.run(
+        fetch_occupation_ratings(
+            conn,
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+            onetsoc_code="15-1252.00",
+            limit=1,
+            offset=0,
+        )
+    )
+
+    item = result["items"][0]
+    assert item["data_value"] == "4.10"
+    assert item["standard_error"] == "0.1830"
+    assert item["recommend_suppress"] is True
+    assert item["not_relevant"] is None
+    assert result["source"]["scale_artifact_sha256"] == "b" * 64
+    assert result["next_offset"] == 1
+
+
+def test_empty_profile_keeps_imported_scale_provenance() -> None:
+    source = {
+        "source_table_name": "Abilities",
+        "source_artifact_url": "https://example.test/abilities.csv",
+        "source_artifact_sha256": "a" * 64,
+        "source_row_count": 2,
+        "scale_artifact_url": "https://example.test/scales.csv",
+        "scale_artifact_sha256": "b" * 64,
+        "scale_source_row_count": 33,
+    }
+
+    result = asyncio.run(
+        fetch_occupation_ratings(
+            FakeConnection(source),
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+            onetsoc_code="15-9999.99",
+            limit=100,
+            offset=0,
+        )
+    )
+
+    assert result["source_available"] is True
+    assert result["items"] == []
+    assert result["source"]["scale_artifact_sha256"] == "b" * 64
+
+
+def test_authenticated_route_delegates_to_bounded_projection() -> None:
+    result = asyncio.run(
+        read_occupation_ratings(
+            onetsoc_code="15-1252.00",
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+            limit=100,
+            offset=0,
+            _account=object(),
+            pool=FakePool(FakeConnection(None)),
+        )
+    )
+
+    assert result["source_available"] is False
diff --git a/tests/test_schema.py b/tests/test_schema.py
index d6cf67712..49dacb2a2 100644
--- a/tests/test_schema.py
+++ b/tests/test_schema.py
@@ -28,6 +28,7 @@
 import psycopg2.errors
 import pytest
 
+from backend.app.occupation_rating_ingestion import fetch_occupation_ratings
 from backend.app.post_chat_ingestion import gather_global_chat_sources
 from scripts.import_onet_ratings import import_ratings
 
@@ -463,6 +464,25 @@ def test_onet_rating_importer_is_idempotent_against_postgresql(
         )
         assert cur.fetchone() == (1, Decimal("4.10"), True)
 
+    async def read_imported_profile() -> dict[str, object]:
+        conn = await asyncpg.connect(args.target_dsn)
+        try:
+            return await fetch_occupation_ratings(
+                conn,
+                data_release_code=args.release_code,
+                source_table_code=args.source_table_code,
+                onetsoc_code="15-1252.00",
+                limit=100,
+                offset=0,
+            )
+        finally:
+            await conn.close()
+
+    profile = asyncio.run(read_imported_profile())
+    assert profile["source_available"] is True
+    assert profile["items"][0]["data_value"] == "4.10"
+    assert profile["source"]["scale_artifact_sha256"] == args.scales_sha256
+
 
 def test_global_ask_evidence_search_indexes_exist_on_normalized_tables(schema_db) -> None:
     """The real PostgreSQL schema owns all nine evidence-search indexes."""

From b03559993117cbdb1523109da302b0799236ae25 Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Thu, 27 Aug 2026 05:28:30 +0900
Subject: [PATCH 07/17] fix: reject malformed O*NET CSV rows

---
 scripts/import_onet_ratings.py    |  7 ++++++-
 tests/test_import_onet_ratings.py | 17 +++++++++++++++++
 2 files changed, 23 insertions(+), 1 deletion(-)

diff --git a/scripts/import_onet_ratings.py b/scripts/import_onet_ratings.py
index 85f01af49..7887f740b 100644
--- a/scripts/import_onet_ratings.py
+++ b/scripts/import_onet_ratings.py
@@ -124,7 +124,12 @@ def _rows(path: Path, required: set[str]) -> list[dict[str, str]]:
         missing = sorted(required - fields)
         if missing:
             raise ValueError(f"missing CSV columns: {', '.join(missing)}")
-        return [dict(row) for row in reader]
+        rows = []
+        for line_number, row in enumerate(reader, start=2):
+            if None in row or any(value is None for value in row.values()):
+                raise ValueError(f"malformed CSV row: {line_number}")
+            rows.append(dict(row))
+        return rows
 
 
 def read_scale_file(path: Path) -> dict[str, ScaleDefinition]:
diff --git a/tests/test_import_onet_ratings.py b/tests/test_import_onet_ratings.py
index 4007d4e98..74fdc509e 100644
--- a/tests/test_import_onet_ratings.py
+++ b/tests/test_import_onet_ratings.py
@@ -120,6 +120,23 @@ def test_conflicting_reference_name_fails_closed(tmp_path: Path) -> None:
         read_rating_file(path, scales, today=date(2026, 8, 27))
 
 
+def test_short_csv_row_fails_with_import_error(tmp_path: Path) -> None:
+    scales = read_scale_file(
+        _write(
+            tmp_path / "scales.csv",
+            "Scale ID,Scale Name,Minimum,Maximum\nIM,Importance,1,5\n",
+        )
+    )
+    path = _write(
+        tmp_path / "short.csv",
+        "O*NET-SOC Code,Title,Element ID,Element Name,Scale ID,Scale Name,Data Value,Date,Domain Source\n"
+        "15-1252.00,Synthetic occupation,1.A.1.a.1\n",
+    )
+
+    with pytest.raises(ValueError, match="malformed CSV row: 2"):
+        read_rating_file(path, scales, today=date(2026, 8, 27))
+
+
 def test_category_table_may_omit_not_relevant_without_inventing_false(
     tmp_path: Path,
 ) -> None:

From b74a841a8af97a69920974b3027ee947b748fc37 Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Thu, 27 Aug 2026 05:33:54 +0900
Subject: [PATCH 08/17] feat(ontology): add occupation evidence view

---
 ARCHITECTURE.md                               |   1 +
 CHANGELOG.md                                  |   4 +
 .../adr/0259-occupation-rating-evidence-ui.md |  52 ++++++
 docs/adr/README.md                            |   1 +
 docs/product-requirements.md                  |  14 ++
 docs/product-technical-gap-baseline.md        |   2 +-
 docs/storybook-inventory.md                   |   1 +
 frontend/src/App.css                          |  34 ++++
 frontend/src/App.tsx                          |  18 +-
 frontend/src/api.test.ts                      |  27 ++-
 frontend/src/api.ts                           |  59 +++++++
 .../OccupationRatingProfile.stories.tsx       |  50 ++++++
 .../OccupationRatingProfile.test.tsx          |  57 +++++++
 .../components/OccupationRatingProfile.tsx    | 154 ++++++++++++++++++
 14 files changed, 465 insertions(+), 9 deletions(-)
 create mode 100644 docs/adr/0259-occupation-rating-evidence-ui.md
 create mode 100644 frontend/src/components/OccupationRatingProfile.stories.tsx
 create mode 100644 frontend/src/components/OccupationRatingProfile.test.tsx
 create mode 100644 frontend/src/components/OccupationRatingProfile.tsx

diff --git a/ARCHITECTURE.md b/ARCHITECTURE.md
index 6a24bce13..1a112853d 100644
--- a/ARCHITECTURE.md
+++ b/ARCHITECTURE.md
@@ -81,6 +81,7 @@ flowchart LR
 | `ask_time_axis.py` | Event-time vs ingestion-time clock choice for that window (ADR 0202) |
 | `ontology.py` | Loads the governed Turtle source tree (`lineageweave-kg.ttl` plus generated fragments), the formal OWL 2/RDFS/SKOS vocabulary for the Knowledge Graph's node/edge types, source taxonomies, and published O*NET linkages (ADR 0004, ADR 0252, ADR 0255, ADR 0256) |
 | `backend/app/occupation_rating_ingestion.py` | Projects authenticated, bounded occupation-rating source evidence with exact decimal and artifact provenance semantics (ADR 0258) |
+| `frontend/src/components/OccupationRatingProfile.tsx` | Reads exact occupation evidence in the existing Dashboard and preserves absence, uncertainty, and warning semantics (ADR 0259) |
 | `ontology_neighborhood.py` | Bounded typed ontology/provenance neighborhood (ADR 0184); PostgreSQL stays authoritative, OWL subclass is not an instance edge |
 | `ontology_source_cursor.py` | Opaque HMAC source-window continuation (ADR 0124); keyset pagination, never OFFSET |
 | `period_report.py` | Fit GRM/GPCM on persisted IRT rows, FIPC-select, EAP-score a period (ADR 0003 slice 3; Bock & Mislevy, 1982) |
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 15193b4e0..01055a341 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -8,6 +8,10 @@ All notable changes to this project are documented here. Format follows
 
 ### Added
 
+- The existing Dashboard now includes an authenticated occupation-evidence
+  view with exact source selection, sample/error context, textual suppression
+  warnings, artifact links, responsive table access, and distinct unavailable
+  versus empty next actions (ADR 0259).
 - Authenticated occupation profiles can now read one exact imported
   release/source with rating and scale artifact provenance, exact decimal
   strings, uncertainty, suppression/relevance flags, and explicit unavailable
diff --git a/docs/adr/0259-occupation-rating-evidence-ui.md b/docs/adr/0259-occupation-rating-evidence-ui.md
new file mode 100644
index 000000000..efcf08726
--- /dev/null
+++ b/docs/adr/0259-occupation-rating-evidence-ui.md
@@ -0,0 +1,52 @@
+# ADR 0259: Occupation-rating evidence in the existing Dashboard
+
+- Status: Accepted
+- Date: 2026-08-27
+- Extends: ADR 0183, ADR 0206, ADR 0258
+- Figma file ID: `1Su3lDRmiZdcUs47t1QwIX`
+
+## Context
+
+ADR 0258 makes an exact imported occupation profile readable, but an API does
+not let an authenticated user find a published work characteristic or notice
+that a value has low precision. ADR 0183 fixes the analyst GNB and prohibits a
+new destination for every evidence type. The existing Dashboard is the place
+for evidence-oriented next actions and already owns responsive table and form
+tokens under ADR 0206.
+
+## Decision
+
+1. Add the occupation profile below the existing operations evidence on the
+   Dashboard. Do not add or rename a GNB destination.
+2. Require the user to submit an exact O*NET-SOC code, data release, and source
+   table. Native form validation rejects malformed occupation codes before a
+   request; the API remains the trust-boundary validator.
+3. Show each exact published value beside its declared scale bounds, optional
+   category, sample size, standard error, confidence interval, source month,
+   domain source, suppression warning, and not-relevant flag. Do not calculate
+   a score, rank, weight, trait estimate, or recommendation.
+4. Keep `source unavailable` distinct from `occupation has no observations`.
+   Both states give a next action instead of displaying zero or a blank table.
+5. Link the rating artifact and scale definition. The API carries their
+   digests and row counts for provenance; a later disclosure control may show
+   those identifiers when user research demonstrates that it aids the task.
+6. Reuse the existing Dashboard Figma file, design tokens, native controls,
+   responsive overflow, focus behavior, and reduced-motion baseline. The table
+   has a named keyboard-focusable region and every warning is text, not color.
+7. Storybook records populated, narrow, source-unavailable, and empty-profile
+   scenes using synthetic records only. Runtime screenshot review covers the
+   populated desktop and narrow scenes.
+
+## Consequences
+
+Users can inspect source evidence without confusing absence, low precision, or
+not-relevant responses with a negative occupational conclusion. The interface
+does not introduce a local psychometric or inference implementation.
+
+## References
+
+National Center for O*NET Development. (2026). *O*NET 31.0 database* [Data
+set]. https://www.onetcenter.org/database.html
+
+World Wide Web Consortium. (2024). *Web Content Accessibility Guidelines
+(WCAG) 2.2* (W3C Recommendation). https://www.w3.org/TR/WCAG22/
diff --git a/docs/adr/README.md b/docs/adr/README.md
index 1a36fa1a8..27911acd1 100644
--- a/docs/adr/README.md
+++ b/docs/adr/README.md
@@ -35,6 +35,7 @@ decision from them.
 | [`ONET_31_LINKAGE_REFERENCES.md`](../doctoring/ONET_31_LINKAGE_REFERENCES.md) | [0256](0256-onet-content-model-published-linkages.md) |
 | [`ONET_RATING_STORE_REFERENCES.md`](../doctoring/ONET_RATING_STORE_REFERENCES.md) | [0257](0257-onet-occupation-rating-observation-store.md) |
 | Occupation-rating authenticated read projection | [0258](0258-occupation-rating-read-api.md) |
+| Occupation-rating Dashboard evidence view | [0259](0259-occupation-rating-evidence-ui.md) |
 
 [0011](0011-prov-o-standard-relations.md) and [0065](0065-prov-o-provenance-boundary.md) cite the dated W3C PROV-O and PROV-DM Recommendations (https://www.w3.org/TR/2013/REC-prov-o-20130430/ and https://www.w3.org/TR/2013/REC-prov-dm-20130430/).
 
diff --git a/docs/product-requirements.md b/docs/product-requirements.md
index d84b284b0..181f7c755 100644
--- a/docs/product-requirements.md
+++ b/docs/product-requirements.md
@@ -161,6 +161,20 @@ Acceptance: invalid identifiers and unbounded pages are rejected; an unavailable
 source never appears as a negative profile; pagination is deterministic; and a
 suppressed observation retains its value and warning flag together.
 
+### PRD-FR-2F — Occupation-rating evidence view
+
+- Let an authenticated user submit an exact O*NET-SOC code, release, and source
+  from the existing Dashboard without changing the governed GNB (ADR 0259).
+- Display published values beside bounds, sample/error/interval evidence,
+  source time, and text warnings; link both source artifacts.
+- Give different next actions for unavailable source, empty occupation,
+  transport failure, and additional pages.
+
+Acceptance: keyboard users can operate the form and named horizontally
+scrollable table; narrow layouts retain complete values; suppression remains
+visible beside its value; and Storybook covers populated, narrow, unavailable,
+and empty states using synthetic data.
+
 ### PRD-FR-3 — Bounded ontology exploration
 
 - Apply RBAC/ABAC, source eligibility, and knowledge cutoff before graph
diff --git a/docs/product-technical-gap-baseline.md b/docs/product-technical-gap-baseline.md
index bd7e5a884..1541e95c2 100644
--- a/docs/product-technical-gap-baseline.md
+++ b/docs/product-technical-gap-baseline.md
@@ -29,7 +29,7 @@ explicit unavailable state, not a reason to infer mappings from labels.
 | Occupation-to-construct relations | ADR 0257 defines a candidate 3NF, release/source-partitioned immutable observation store and deterministic pinned-CSV importer preserving value, optional category, sample/error/CI, suppression, relevance, exact `MM/YYYY` source update month, source digest, and domain provenance. The official O*NET 31.0 Abilities file (94,640 rows, 910 occupations, 52 elements; SHA-256 `7e9cd79791ce6014e1d26d0a449ae5b1e7aa7ef52d39b3934c3bb8d438104b88`) and all 33 Scales Reference rows (SHA-256 `bcba23858ce21ecaacbde303a8993e35d46724b4afb8c9ec2b10e04f42adcfc9`) imported into a throwaway local PostgreSQL database with all 94,640 observations, 55 suppression flags, 7,572 not-relevant flags, and source months from `12/2004` through `08/2026`; every scale retained `scales_reference` artifact provenance, the database was dropped afterward, and no corpus is committed or claimed deployed | Pass exact-head review/checks and protected merge; validate and import every selected official rating artifact through an authorized runtime, returning only aggregate evidence; never invent or locally normalize a weight |
 | Job-family and job-series semantics | No authoritative employer-specific job architecture is present | Define an organization-neutral import contract that preserves the authorized source hierarchy and distinguishes standard occupation codes from employer job families/series; no label-based binding |
 | Temporal and multilevel interpretation | Static vocabulary only; no person-level inference is asserted | Version valid and transaction time, preserve occupation/organization/unit nesting and multiple membership, and require TEPP or the owning Rust psychometric service before any calibrated temporal or multilevel result |
-| Product consumption | ADR 0258 defines a candidate authenticated occupation-rating API that distinguishes unavailable source from empty profile and returns exact decimal/uncertainty/warning/artifact provenance without ranking; accessible UI and authenticated runtime evidence remain absent | Pass exact-head review/checks and protected merge; add the accessible exploration UI, synthetic Storybook unavailable/empty/suppressed/paginated states, screenshots, and authenticated aggregate runtime evidence without exposing identifying records |
+| Product consumption | ADR 0258 defines a candidate authenticated occupation-rating API; ADR 0259 adds a candidate Dashboard evidence view with exact source selection, sample/error context, text warnings, artifact links, responsive overflow, and distinct unavailable/empty states. Component/API tests and Storybook build pass; synthetic populated scenes were visually audited at 1440×900 and 390×844, including the mobile horizontal-scroll next action. Protected delivery and authenticated runtime evidence remain absent | Pass exact-head review/checks and protected merge; verify the authenticated API and rendered Dashboard against an authorized imported source using only aggregate/non-identifying evidence, and add a source catalog selector after an authoritative catalog read contract exists |
 
 ### Current exact-head PR queue
 
diff --git a/docs/storybook-inventory.md b/docs/storybook-inventory.md
index 81ab3a8af..c9e1570bf 100644
--- a/docs/storybook-inventory.md
+++ b/docs/storybook-inventory.md
@@ -6,6 +6,7 @@ operator-facing control you can click before changing product CSS.
 | Story | Operator next action | Token / module |
 |---|---|---|
 | `Workspace/OperationsDashboard` | Compare Event and post counts, inspect external-information coverage, then open the cited source behind a claim, handover, or repeat-issue fact. `EvidenceReady`, `NarrowViewport`, `AnalysisPendingAndMissingEvidence`, `AnalysisFailed`, and `LoadError` cover populated, mobile, unavailable-evidence, analysis-pending, retryable failure, and transport-error states. | `--color-dashboard-*`, `OperationsDashboard` |
+| `Ontology/OccupationRatingProfile` | Enter an exact O*NET-SOC code, inspect the published value beside its sample/error and warning, then open the rating or scale artifact. `InteractiveEvidenceReady`, `EvidenceReady`, `NarrowViewport`, `SourceUnavailable`, and `EmptyOccupation` cover the full form, populated table, horizontal mobile access, and honest absence states. | `OccupationRatingProfile`, native table, `--color-border`, `--size-control-min` |
 | `Post/SimilarVocPanel` | Compare ontology/semantic similar VOC and prior action evidence, then open the source; unavailable states show no fabricated TEPP theta or weight. | `SimilarVocPanel.css`, `SimilarVocPanel` |
 | `Evidence/CitationChip` | Click a cited title to open that source post. | `--color-chip-border`, `--radius-chip`, `CitationChip` |
 | `Evidence/OrganizationAliasChip` | Click a cataloged org; the parenthetical is the unique corroborated SKOS companion. | `--color-chip-border`, `--radius-chip`, `OrganizationAliasChip` |
diff --git a/frontend/src/App.css b/frontend/src/App.css
index 901b17b1a..33adf4221 100644
--- a/frontend/src/App.css
+++ b/frontend/src/App.css
@@ -1439,8 +1439,42 @@
 .dashboard-case-card dd { margin: 0; font-weight: 600; }
 .dashboard-case-card button { margin-top: auto; }
 
+.occupation-rating-profile {
+  max-width: 1440px;
+  margin: 0 auto 2rem;
+  padding: 2rem;
+  color: var(--color-text-heading);
+}
+
+.occupation-rating-form {
+  display: grid;
+  grid-template-columns: repeat(3, minmax(10rem, 1fr)) auto;
+  align-items: end;
+  gap: var(--space-control-gap);
+  margin: 1rem 0;
+}
+
+.occupation-rating-form label,
+.occupation-rating-source {
+  display: grid;
+  gap: var(--space-control-gap);
+}
+
+.occupation-rating-form input { min-height: var(--size-control-min); }
+.occupation-rating-source { grid-template-columns: repeat(auto-fit, minmax(12rem, 1fr)); margin: 1rem 0; }
+.occupation-rating-scroll-hint { display: none; }
+.occupation-rating-table { overflow-x: auto; border: 1px solid var(--color-border); }
+.occupation-rating-table table { width: 100%; min-width: 64rem; border-collapse: collapse; }
+.occupation-rating-table caption { padding: 0.75rem; text-align: left; }
+.occupation-rating-table th,
+.occupation-rating-table td { padding: 0.75rem; border-bottom: 1px solid var(--color-border); text-align: left; vertical-align: top; }
+.occupation-rating-table small { display: block; color: var(--color-text); }
+
 @media (max-width: 900px) {
   .operations-dashboard { padding: 1rem; }
+  .occupation-rating-profile { padding: 1rem; }
+  .occupation-rating-form { grid-template-columns: 1fr; }
+  .occupation-rating-scroll-hint { display: block; }
   .operations-dashboard-heading { align-items: start; flex-direction: column; }
   .dashboard-metrics { grid-template-columns: repeat(2, minmax(0, 1fr)); }
   .dashboard-case-grid { grid-template-columns: 1fr; }
diff --git a/frontend/src/App.tsx b/frontend/src/App.tsx
index fbba1d9f2..671e12aab 100644
--- a/frontend/src/App.tsx
+++ b/frontend/src/App.tsx
@@ -100,6 +100,7 @@ import { SimilarVocPanel } from "./components/SimilarVocPanel";
 import { chatEvidenceKindLabel } from "./evidenceKindLabels";
 import { WorkspaceNav, type WorkspaceDestination } from "./components/WorkspaceNav";
 import { OperationsDashboard } from "./components/OperationsDashboard";
+import { OccupationRatingProfile } from "./components/OccupationRatingProfile";
 import { initialWorkspaceDestination } from "./gnbChrome";
 import { LineageDag } from "./LineageDag";
 import { PostBody } from "./PostBody";
@@ -5119,13 +5120,16 @@ export default function App({ showLabPanels = false }: { showLabPanels?: boolean
       />
       <main>
         {destination === "dashboard" ? (
-          <OperationsDashboard
-            accessToken={accessToken}
-            onOpenPost={(postId) => {
-              setPostToOpen(postId);
-              setDestination("board");
-            }}
-          />
+          <>
+            <OperationsDashboard
+              accessToken={accessToken}
+              onOpenPost={(postId) => {
+                setPostToOpen(postId);
+                setDestination("board");
+              }}
+            />
+            <OccupationRatingProfile accessToken={accessToken} />
+          </>
         ) : null}
         {destination === "board" ? (
           <PostList
diff --git a/frontend/src/api.test.ts b/frontend/src/api.test.ts
index 3182afb04..cf0b30d40 100644
--- a/frontend/src/api.test.ts
+++ b/frontend/src/api.test.ts
@@ -1,5 +1,11 @@
 import { afterEach, describe, expect, it, vi } from "vitest";
-import { BackendError, fetchMe, fetchOperationsDashboard, updateTenantConfig } from "./api";
+import {
+  BackendError,
+  fetchMe,
+  fetchOccupationRatings,
+  fetchOperationsDashboard,
+  updateTenantConfig,
+} from "./api";
 
 afterEach(() => {
   vi.unstubAllGlobals();
@@ -19,6 +25,25 @@ describe("backendFetch provider-error boundary", () => {
     );
   });
 
+  it("encodes an exact occupation rating source request", async () => {
+    const fetchMock = vi.fn().mockResolvedValue(
+      new Response(JSON.stringify({ source_available: false, items: [] }), {
+        headers: { "Content-Type": "application/json" },
+      }),
+    );
+    vi.stubGlobal("fetch", fetchMock);
+
+    await fetchOccupationRatings("access-token", {
+      onetsocCode: "15-1252.00",
+      dataReleaseCode: "onet-31.0",
+      sourceTableCode: "abilities",
+    });
+
+    expect(fetchMock.mock.calls[0][0]).toContain(
+      "/api/occupations/15-1252.00/ratings?data_release_code=onet-31.0&source_table_code=abilities&limit=100&offset=0",
+    );
+  });
+
   it("does not expose provider details from server failures", async () => {
     vi.stubGlobal(
       "fetch",
diff --git a/frontend/src/api.ts b/frontend/src/api.ts
index fca5882d0..1e904bbc5 100644
--- a/frontend/src/api.ts
+++ b/frontend/src/api.ts
@@ -960,6 +960,65 @@ export function fetchOntologyNeighborhood(
   return backendFetch(`/api/ontology/neighborhood?${params.toString()}`, accessToken);
 }
 
+export interface OccupationRatingItem {
+  element_id: string;
+  element_name: string;
+  scale_id: string;
+  scale_name: string;
+  minimum_value: string;
+  maximum_value: string;
+  category_value: number | null;
+  data_value: string;
+  sample_size: number | null;
+  standard_error: string | null;
+  lower_ci_bound: string | null;
+  upper_ci_bound: string | null;
+  recommend_suppress: boolean | null;
+  not_relevant: boolean | null;
+  source_updated_month: string | null;
+  domain_source_code: string | null;
+}
+
+export interface OccupationRatingProfile {
+  data_release_code: string;
+  source_table_code: string;
+  onetsoc_code: string;
+  source_available: boolean;
+  source: {
+    source_table_name: string;
+    source_artifact_url: string;
+    source_artifact_sha256: string;
+    source_row_count: number;
+    scale_artifact_url: string | null;
+    scale_artifact_sha256: string | null;
+    scale_source_row_count: number | null;
+  } | null;
+  items: OccupationRatingItem[];
+  next_offset: number | null;
+}
+
+export function fetchOccupationRatings(
+  accessToken: string,
+  query: {
+    onetsocCode: string;
+    dataReleaseCode: string;
+    sourceTableCode: string;
+    limit?: number;
+    offset?: number;
+  },
+): Promise<OccupationRatingProfile> {
+  const params = new URLSearchParams({
+    data_release_code: query.dataReleaseCode,
+    source_table_code: query.sourceTableCode,
+    limit: String(query.limit ?? 100),
+    offset: String(query.offset ?? 0),
+  });
+  return backendFetch(
+    `/api/occupations/${encodeURIComponent(query.onetsocCode)}/ratings?${params.toString()}`,
+    accessToken,
+  );
+}
+
 export function extractPostKeymen(
   accessToken: string,
   postId: string,
diff --git a/frontend/src/components/OccupationRatingProfile.stories.tsx b/frontend/src/components/OccupationRatingProfile.stories.tsx
new file mode 100644
index 000000000..d12114844
--- /dev/null
+++ b/frontend/src/components/OccupationRatingProfile.stories.tsx
@@ -0,0 +1,50 @@
+import type { Meta, StoryObj } from "@storybook/react-vite";
+import { expect, userEvent, within } from "storybook/test";
+import { OccupationRatingProfile, OccupationRatingProfileView } from "./OccupationRatingProfile";
+import "../App.css";
+
+const ready = {
+  data_release_code: "onet-31.0", source_table_code: "abilities", onetsoc_code: "15-1252.00", source_available: true,
+  source: { source_table_name: "Abilities", source_artifact_url: "https://example.test/abilities.csv", source_artifact_sha256: "a".repeat(64), source_row_count: 94640, scale_artifact_url: "https://example.test/scales.csv", scale_artifact_sha256: "b".repeat(64), scale_source_row_count: 33 },
+  items: [
+    { element_id: "1.A.1.a.1", element_name: "Oral Comprehension", scale_id: "IM", scale_name: "Importance", minimum_value: "1.00", maximum_value: "5.00", category_value: null, data_value: "4.10", sample_size: 120, standard_error: "0.0800", lower_ci_bound: "3.9432", upper_ci_bound: "4.2568", recommend_suppress: true, not_relevant: null, source_updated_month: "08/2026", domain_source_code: "Analyst" },
+    { element_id: "1.A.1.a.2", element_name: "Written Comprehension", scale_id: "LV", scale_name: "Level", minimum_value: "0.00", maximum_value: "7.00", category_value: null, data_value: "5.25", sample_size: 118, standard_error: "0.1100", lower_ci_bound: "5.0344", upper_ci_bound: "5.4656", recommend_suppress: false, not_relevant: false, source_updated_month: "08/2026", domain_source_code: "Analyst" },
+  ],
+  next_offset: null,
+};
+
+const meta = { title: "Ontology/OccupationRatingProfile", component: OccupationRatingProfileView, parameters: { layout: "fullscreen" }, args: { profile: ready } } satisfies Meta<typeof OccupationRatingProfileView>;
+export default meta;
+type Story = StoryObj<typeof meta>;
+
+export const EvidenceReady: Story = {
+  play: async ({ canvasElement }) => {
+    const canvas = within(canvasElement);
+    await expect(canvas.getByText("4.10")).toBeVisible();
+    await expect(canvas.getByText(/정밀도가 낮아/)).toBeVisible();
+  },
+};
+
+export const InteractiveEvidenceReady: Story = {
+  render: () => <OccupationRatingProfile accessToken="synthetic-token" />,
+  beforeEach: () => {
+    const previousFetch = globalThis.fetch;
+    globalThis.fetch = async () => new Response(JSON.stringify(ready), {
+      headers: { "Content-Type": "application/json" },
+    });
+    return () => { globalThis.fetch = previousFetch; };
+  },
+  play: async ({ canvasElement }) => {
+    const canvas = within(canvasElement);
+    await userEvent.type(canvas.getByLabelText("O*NET-SOC 직업 코드"), "15-1252.00");
+    await userEvent.click(canvas.getByRole("button", { name: "직업 근거 열기" }));
+    await expect(canvas.findByText("4.10")).resolves.toBeVisible();
+  },
+};
+
+export const NarrowViewport: Story = {
+  ...InteractiveEvidenceReady,
+  parameters: { viewport: { defaultViewport: "mobile1" } },
+};
+export const SourceUnavailable: Story = { args: { profile: { ...ready, source_available: false, source: null, items: [] } } };
+export const EmptyOccupation: Story = { args: { profile: { ...ready, items: [] } } };
diff --git a/frontend/src/components/OccupationRatingProfile.test.tsx b/frontend/src/components/OccupationRatingProfile.test.tsx
new file mode 100644
index 000000000..2f7cd9125
--- /dev/null
+++ b/frontend/src/components/OccupationRatingProfile.test.tsx
@@ -0,0 +1,57 @@
+import { render, screen } from "@testing-library/react";
+import userEvent from "@testing-library/user-event";
+import { describe, expect, it, vi } from "vitest";
+import { fetchOccupationRatings, type OccupationRatingProfile as Payload } from "../api";
+import { OccupationRatingProfile, OccupationRatingProfileView } from "./OccupationRatingProfile";
+
+vi.mock("../api", async (importOriginal) => ({
+  ...(await importOriginal<typeof import("../api")>()),
+  fetchOccupationRatings: vi.fn(),
+}));
+
+const ready: Payload = {
+  data_release_code: "onet-31.0",
+  source_table_code: "abilities",
+  onetsoc_code: "15-1252.00",
+  source_available: true,
+  source: {
+    source_table_name: "Abilities",
+    source_artifact_url: "https://example.test/abilities.csv",
+    source_artifact_sha256: "a".repeat(64),
+    source_row_count: 2,
+    scale_artifact_url: "https://example.test/scales.csv",
+    scale_artifact_sha256: "b".repeat(64),
+    scale_source_row_count: 33,
+  },
+  items: [{
+    element_id: "1.A.1.a.1", element_name: "Oral Comprehension",
+    scale_id: "IM", scale_name: "Importance", minimum_value: "1.00", maximum_value: "5.00",
+    category_value: null, data_value: "4.10", sample_size: 120, standard_error: "0.0800",
+    lower_ci_bound: "3.9432", upper_ci_bound: "4.2568", recommend_suppress: true,
+    not_relevant: true, source_updated_month: "08/2026", domain_source_code: "Analyst",
+  }],
+  next_offset: null,
+};
+
+describe("OccupationRatingProfile", () => {
+  it("submits exact identifiers and renders warnings beside the retained value", async () => {
+    vi.mocked(fetchOccupationRatings).mockResolvedValue(ready);
+    render(<OccupationRatingProfile accessToken="synthetic-token" />);
+    await userEvent.type(screen.getByLabelText("O*NET-SOC 직업 코드"), "15-1252.00");
+    await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
+    expect(fetchOccupationRatings).toHaveBeenCalledWith("synthetic-token", {
+      onetsocCode: "15-1252.00", dataReleaseCode: "onet-31.0", sourceTableCode: "abilities", offset: 0,
+    });
+    expect(await screen.findByText("4.10")).toBeInTheDocument();
+    expect(screen.getByText(/정밀도가 낮아/)).toBeInTheDocument();
+    expect(screen.getByText(/해당 없음 응답이 포함됩니다/)).toBeInTheDocument();
+    expect(screen.getByText(/표를 가로로 밀어/)).toBeInTheDocument();
+  });
+
+  it("distinguishes an unavailable artifact from an empty occupation profile", () => {
+    const { rerender } = render(<OccupationRatingProfileView profile={{ ...ready, source_available: false, source: null, items: [] }} />);
+    expect(screen.getByRole("status")).toHaveTextContent("아직 준비되지 않았습니다");
+    rerender(<OccupationRatingProfileView profile={{ ...ready, items: [] }} />);
+    expect(screen.getByRole("status")).toHaveTextContent("관측값이 없습니다");
+  });
+});
diff --git a/frontend/src/components/OccupationRatingProfile.tsx b/frontend/src/components/OccupationRatingProfile.tsx
new file mode 100644
index 000000000..7107c4155
--- /dev/null
+++ b/frontend/src/components/OccupationRatingProfile.tsx
@@ -0,0 +1,154 @@
+import { useState } from "react";
+import {
+  fetchOccupationRatings,
+  type OccupationRatingProfile as OccupationRatingProfilePayload,
+} from "../api";
+
+type Props = { accessToken: string };
+
+/** Lets an authenticated user inspect one exact imported occupation profile. */
+export function OccupationRatingProfile({ accessToken }: Props) {
+  const [onetsocCode, setOnetsocCode] = useState("");
+  const [releaseCode, setReleaseCode] = useState("onet-31.0");
+  const [sourceCode, setSourceCode] = useState("abilities");
+  const [profile, setProfile] = useState<OccupationRatingProfilePayload | null>(null);
+  const [status, setStatus] = useState<"idle" | "loading" | "error">("idle");
+
+  function load(offset = 0) {
+    setStatus("loading");
+    fetchOccupationRatings(accessToken, {
+      onetsocCode,
+      dataReleaseCode: releaseCode,
+      sourceTableCode: sourceCode,
+      offset,
+    })
+      .then((payload) => {
+        setProfile((current) =>
+          offset && current
+            ? { ...payload, items: [...current.items, ...payload.items] }
+            : payload,
+        );
+        setStatus("idle");
+      })
+      .catch(() => setStatus("error"));
+  }
+
+  return (
+    <section className="occupation-rating-profile" aria-labelledby="occupation-rating-heading">
+      <header>
+        <p className="dashboard-eyebrow">공개 직업 근거</p>
+        <h2 id="occupation-rating-heading">직업별 업무 특성 확인</h2>
+        <p>직업 코드와 근거 표를 선택해 관측값, 오차, 사용 주의사항을 함께 확인하세요.</p>
+      </header>
+      <form
+        className="occupation-rating-form"
+        onSubmit={(event) => {
+          event.preventDefault();
+          load();
+        }}
+      >
+        <label>
+          O*NET-SOC 직업 코드
+          <input
+            required
+            pattern="[0-9]{2}-[0-9]{4}\.[0-9]{2}"
+            placeholder="15-1252.00"
+            value={onetsocCode}
+            onChange={(event) => setOnetsocCode(event.target.value)}
+          />
+        </label>
+        <label>
+          데이터 릴리스
+          <input required value={releaseCode} onChange={(event) => setReleaseCode(event.target.value)} />
+        </label>
+        <label>
+          근거 표
+          <input required value={sourceCode} onChange={(event) => setSourceCode(event.target.value)} />
+        </label>
+        <button className="btn-secondary" type="submit" disabled={status === "loading"}>
+          {status === "loading" ? "근거를 불러오는 중" : "직업 근거 열기"}
+        </button>
+      </form>
+      {status === "error" ? (
+        <p role="alert">직업 근거를 불러오지 못했습니다. 코드와 접근 권한을 확인한 뒤 다시 시도하세요.</p>
+      ) : null}
+      {profile ? <OccupationRatingProfileView profile={profile} /> : null}
+      {profile?.next_offset != null ? (
+        <button
+          className="btn-secondary"
+          type="button"
+          disabled={status === "loading"}
+          onClick={() => load(profile.next_offset ?? 0)}
+        >
+          다음 관측값 불러오기
+        </button>
+      ) : null}
+    </section>
+  );
+}
+
+/** Renders an exact occupation profile for runtime and Storybook scenes. */
+export function OccupationRatingProfileView({
+  profile,
+}: {
+  profile: OccupationRatingProfilePayload;
+}) {
+  if (!profile.source_available) {
+    return (
+      <p role="status">
+        선택한 릴리스와 근거 표가 아직 준비되지 않았습니다. 다른 근거 표를 선택하거나 데이터 담당자에게 가져오기를 요청하세요.
+      </p>
+    );
+  }
+  if (profile.items.length === 0) {
+    return (
+      <p role="status">
+        이 근거 표에는 선택한 직업의 관측값이 없습니다. 직업 코드나 근거 표를 바꿔 확인하세요.
+      </p>
+    );
+  }
+  return (
+    <>
+      <div className="occupation-rating-source">
+        <strong>{profile.source?.source_table_name}</strong>
+        <span>{profile.data_release_code} · {profile.onetsoc_code}</span>
+        <a href={profile.source?.source_artifact_url} target="_blank" rel="noreferrer">평정 원문 열기</a>
+        {profile.source?.scale_artifact_url ? (
+          <a href={profile.source.scale_artifact_url} target="_blank" rel="noreferrer">척도 정의 열기</a>
+        ) : null}
+      </div>
+      <p className="occupation-rating-scroll-hint">표를 가로로 밀어 오차와 사용 주의를 확인하세요.</p>
+      <div className="occupation-rating-table" role="region" tabIndex={0} aria-label="직업 업무 특성 관측값">
+        <table>
+          <caption>값과 오차 및 사용 주의사항</caption>
+          <thead>
+            <tr>
+              <th>업무 특성</th><th>척도</th><th>값</th><th>표본·오차</th><th>출처 시점</th><th>사용 주의</th>
+            </tr>
+          </thead>
+          <tbody>
+            {profile.items.map((item) => (
+              <tr key={`${item.element_id}-${item.scale_id}-${item.category_value ?? "none"}`}>
+                <td>{item.element_name}<small>{item.element_id}</small></td>
+                <td>{item.scale_name} ({item.minimum_value}–{item.maximum_value})</td>
+                <td>{item.data_value}{item.category_value == null ? null : ` · 범주 ${item.category_value}`}</td>
+                <td>
+                  {item.sample_size == null ? "표본 수 없음" : `N ${item.sample_size}`}
+                  {item.standard_error == null ? null : ` · SE ${item.standard_error}`}
+                  {item.lower_ci_bound == null || item.upper_ci_bound == null ? null : ` · CI ${item.lower_ci_bound}–${item.upper_ci_bound}`}
+                </td>
+                <td>{item.source_updated_month ?? "시점 없음"}{item.domain_source_code ? ` · ${item.domain_source_code}` : ""}</td>
+                <td>
+                  {[
+                    item.recommend_suppress ? "정밀도가 낮아 해석 전 원문을 확인하세요." : null,
+                    item.not_relevant ? "해당 없음 응답이 포함됩니다." : null,
+                  ].filter(Boolean).join(" ") || "공개 근거와 함께 해석하세요."}
+                </td>
+              </tr>
+            ))}
+          </tbody>
+        </table>
+      </div>
+    </>
+  );
+}

From 3d1d3171f9c4bd4efcedacc7af893dcaf84d6778 Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Thu, 27 Aug 2026 05:40:20 +0900
Subject: [PATCH 09/17] docs(gaps): refresh exact PR queue

---
 docs/product-technical-gap-baseline.md | 79 +++++++++++++++-----------
 1 file changed, 45 insertions(+), 34 deletions(-)

diff --git a/docs/product-technical-gap-baseline.md b/docs/product-technical-gap-baseline.md
index 1541e95c2..1632ad818 100644
--- a/docs/product-technical-gap-baseline.md
+++ b/docs/product-technical-gap-baseline.md
@@ -35,40 +35,51 @@ explicit unavailable state, not a reason to infer mappings from labels.
 
 | PR | Exact observed head | Base | Observed gate state |
 |---:|---|---|---|
-| #735 | `56d8cbfa` | `feat/onet-occupation-ratings-contract` | clean immediately after exact parent merge; hosted checks had not yet registered on this head, so this is not merge-readiness evidence |
-| #734 | `0dedf33b` | `feat/onet-content-model-linkages` | unstable; two hosted gates pending after exact parent reconciliation |
-| #732 | `3e30402d` | `feat/onet-31-content-model-ontology` | unstable; one hosted gate pending after exact parent reconciliation |
-| #731 | `9654bb8b` | `feat/soc-2018-full-hierarchy` | unstable; one hosted gate pending after exact parent reconciliation |
-| #724 | `fdec8f65` | `feat/io-occupational-taxonomy` | clean; hosted gates passed; parent-first stack gate remains |
-| #719 | `6ee2278a` | `feat/io-psych-construct-ontology` | unstable; full tests, frontend, CodeRabbit, and Devin Review passed |
-| #718 | `2723fea3` | `feat/fja-worker-function-ontology` | unstable; full tests, frontend, CodeRabbit, and Devin Review passed |
-| #717 | `771a8edf` | `feat/voice-of-x-complete-taxonomy` | unstable; 1 pending check(s) |
-| #716 | `8b54b2f7` | `fix/structured-workflow-exact-pin` | clean; no non-passing check observed |
-| #714 | `aa93318f` | `main` | blocked; no non-passing check observed |
-| #713 | `cc3dfc14` | `main` | blocked; review required; 13 pending check(s) |
-| #711 | `8902e37f` | `feat/dashboard-case-metrics` | clean; no non-passing check observed |
-| #710 | `8df04b68` | `main` | blocked; review required; no non-passing check observed |
-| #709 | `8ef4090c` | `main` | blocked; review required; 11 pending check(s) |
-| #704 | `027323cf` | `main` | blocked; review required; 2 failed check(s) |
-| #702 | `5de66ab9` | `main` | blocked; review required; 2 pending check(s) |
-| #701 | `cc3351a9` | `main` | blocked; review required; 1 failed check(s) |
-| #700 | `1bc99eca` | `main` | blocked; review required; 1 failed check(s) |
-| #680 | `efe864e5` | `main` | blocked; 1 failed check(s) |
-| #679 | `13ecf41d` | `main` | blocked; no non-passing check observed |
-| #672 | `a3e87a89` | `main` | blocked; review required; 1 failed check(s) |
-| #668 | `1194f44d` | `main` | blocked; review required; 1 failed check(s) |
-| #667 | `c2d11a8a` | `main` | blocked; review required; 2 pending check(s) |
-| #658 | `15d670f0` | `main` | blocked; review required; 1 failed check(s) |
-| #657 | `9f71681c` | `main` | blocked; review required; 1 failed check(s) |
-| #644 | `f53dd28e` | `main` | blocked; review required; 1 failed check(s) |
-| #643 | `8767de1b` | `main` | blocked; review required; 1 failed check(s); 1 pending check(s) |
-| #640 | `5594029c` | `main` | blocked; no non-passing check observed |
-| #639 | `2f4b1bff` | `main` | blocked; review required; 1 failed check(s) |
-| #632 | `24262a99` | `main` | blocked; review required; 1 failed check(s) |
-| #629 | `b721b0f2` | `main` | blocked; review required; 1 failed check(s) |
-
-> Dashboard delivery snapshot: 2026-08-26 07:15 KST. Protected `main` was
-> `494b54e2245040bcf02b45376f221c37cd437e76`. This local branch is not
+| #740 | `52814353` | `feat/onet-rating-read-api` | unstable; three non-passing hosted contexts after opening the occupation evidence UI PR |
+| #739 | `9a3f380f` | `feat/operations-candidate-priority` | unstable; two non-passing hosted contexts |
+| #738 | `026ba803` | `feat/onet-rating-importer` | unstable; one hosted context still running after exact parent reconciliation; unresolved review threads 0 |
+| #735 | `b0355999` | `feat/onet-occupation-ratings-contract` | unstable; one non-passing hosted context |
+| #734 | `4c3677af` | `feat/onet-content-model-linkages` | clean snapshot; parent-first stack gate remains |
+| #733 | `b53c1edd` | `feat/io-psych-construct-extraction` | unstable; two non-passing hosted contexts |
+| #732 | `7f60aa8e` | `feat/onet-31-content-model-ontology` | clean snapshot; parent-first stack gate remains |
+| #731 | `b3b9b360` | `feat/soc-2018-full-hierarchy` | clean snapshot; parent-first stack gate remains |
+| #728 | `e52a8272` | `feat/dashboard-case-metrics` | clean snapshot; parent-first stack gate remains |
+| #726 | `d6a12fbb` | `feat/io-psych-construct-catalog` | unstable; one non-passing hosted context |
+| #724 | `1d2f8052` | `feat/io-occupational-taxonomy` | clean snapshot; parent-first stack gate remains |
+| #723 | `316fc190` | `feat/io-psych-construct-persistence` | clean snapshot; parent-first stack gate remains |
+| #721 | `9214c50f` | `feat/io-psych-construct-ontology` | clean snapshot; parent-first stack gate remains |
+| #720 | `dda0531d` | `main` | blocked; review required; one non-passing hosted context |
+| #719 | `6ee2278a` | `feat/io-psych-construct-ontology` | clean snapshot; parent-first stack gate remains |
+| #718 | `2723fea3` | `feat/fja-worker-function-ontology` | clean snapshot; parent-first stack gate remains |
+| #717 | `bf355876` | `feat/voice-of-x-complete-taxonomy` | unstable; one non-passing hosted context |
+| #716 | `0a8bd0b6` | `fix/structured-workflow-exact-pin` | unstable; two non-passing hosted contexts |
+| #714 | `76a602c8` | `main` | blocked; one non-passing hosted context |
+| #713 | `850494c3` | `main` | blocked; review required; one non-passing hosted context |
+| #711 | `05e5f520` | `feat/dashboard-case-metrics` | clean snapshot; parent-first stack gate remains |
+| #710 | `27a917ee` | `main` | blocked; review required |
+| #709 | `8ef4090c` | `main` | blocked; review required; all hosted checks passed; auto-merge awaits the independent approval gate |
+| #704 | `7b9a70ee` | `main` | blocked; review required; one non-passing hosted context |
+| #702 | `05bdd5b7` | `main` | blocked; two non-passing hosted contexts |
+| #701 | `cc3351a9` | `main` | blocked; review required; one non-passing hosted context |
+| #700 | `1bc99eca` | `main` | blocked; review required; one non-passing hosted context |
+| #680 | `b05e3100` | `main` | blocked; one non-passing hosted context |
+| #679 | `135dfe7c` | `main` | blocked; review required; two non-passing hosted contexts |
+| #672 | `a3e87a89` | `main` | blocked; review required; one non-passing hosted context |
+| #668 | `234f975b` | `main` | blocked; review required; three non-passing hosted contexts |
+| #667 | `92a64c40` | `main` | blocked; one non-passing hosted context |
+| #658 | `15d670f0` | `main` | blocked; review required; one non-passing hosted context |
+| #657 | `9f71681c` | `main` | blocked; review required; one non-passing hosted context |
+| #644 | `f53dd28e` | `main` | blocked; review required; one non-passing hosted context |
+| #643 | `8767de1b` | `main` | blocked; review required; one non-passing hosted context |
+| #640 | `ebfe60af` | `main` | blocked; one non-passing hosted context |
+| #639 | `2f4b1bff` | `main` | blocked; review required; one non-passing hosted context |
+| #632 | `24262a99` | `main` | blocked; review required; one non-passing hosted context |
+| #629 | `b721b0f2` | `main` | blocked; review required; one non-passing hosted context |
+
+> Dashboard delivery snapshot: 2026-08-27 05:40 KST. Protected `main` was
+> `ff7431bd1851c03e737808d22c6a2d43968582f9`. `mergeStateStatus` and hosted
+> context counts are observations, not protected merge-readiness evidence.
+> This local branch is not
 > protected-main release evidence.
 
 ## Operations Dashboard PRD/TRD traceability

From 5697189c25c05639c6adcdec7126437bb0feb408 Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Thu, 27 Aug 2026 05:59:23 +0900
Subject: [PATCH 10/17] fix(ui): bind occupation evidence pages

---
 .../OccupationRatingProfile.test.tsx          | 38 +++++++++++++++++++
 .../components/OccupationRatingProfile.tsx    | 20 ++++++----
 2 files changed, 51 insertions(+), 7 deletions(-)

diff --git a/frontend/src/components/OccupationRatingProfile.test.tsx b/frontend/src/components/OccupationRatingProfile.test.tsx
index 2f7cd9125..cddaec717 100644
--- a/frontend/src/components/OccupationRatingProfile.test.tsx
+++ b/frontend/src/components/OccupationRatingProfile.test.tsx
@@ -48,6 +48,44 @@ describe("OccupationRatingProfile", () => {
     expect(screen.getByText(/표를 가로로 밀어/)).toBeInTheDocument();
   });
 
+  it("keeps pagination bound to the loaded profile after form edits", async () => {
+    vi.mocked(fetchOccupationRatings)
+      .mockResolvedValueOnce({ ...ready, next_offset: 100 })
+      .mockResolvedValueOnce({ ...ready, items: [{ ...ready.items[0], scale_id: "LV" }] });
+    render(<OccupationRatingProfile accessToken="synthetic-token" />);
+    const occupation = screen.getByLabelText("O*NET-SOC 직업 코드");
+    await userEvent.type(occupation, "15-1252.00");
+    await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
+    await screen.findByText("4.10");
+
+    await userEvent.clear(occupation);
+    await userEvent.type(occupation, "11-1011.00");
+    await userEvent.click(screen.getByRole("button", { name: "다음 관측값 불러오기" }));
+
+    expect(fetchOccupationRatings).toHaveBeenLastCalledWith("synthetic-token", {
+      onetsocCode: "15-1252.00", dataReleaseCode: "onet-31.0", sourceTableCode: "abilities", offset: 100,
+    });
+    expect(await screen.findAllByText("4.10")).toHaveLength(2);
+  });
+
+  it("removes stale evidence while a fresh occupation loads", async () => {
+    vi.mocked(fetchOccupationRatings)
+      .mockResolvedValueOnce(ready)
+      .mockImplementationOnce(() => new Promise(() => undefined));
+    render(<OccupationRatingProfile accessToken="synthetic-token" />);
+    const occupation = screen.getByLabelText("O*NET-SOC 직업 코드");
+    await userEvent.type(occupation, "15-1252.00");
+    await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
+    await screen.findByText("4.10");
+
+    await userEvent.clear(occupation);
+    await userEvent.type(occupation, "11-1011.00");
+    await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
+
+    expect(screen.queryByText("4.10")).not.toBeInTheDocument();
+    expect(screen.getByRole("button", { name: "근거를 불러오는 중" })).toBeDisabled();
+  });
+
   it("distinguishes an unavailable artifact from an empty occupation profile", () => {
     const { rerender } = render(<OccupationRatingProfileView profile={{ ...ready, source_available: false, source: null, items: [] }} />);
     expect(screen.getByRole("status")).toHaveTextContent("아직 준비되지 않았습니다");
diff --git a/frontend/src/components/OccupationRatingProfile.tsx b/frontend/src/components/OccupationRatingProfile.tsx
index 7107c4155..20833f3ee 100644
--- a/frontend/src/components/OccupationRatingProfile.tsx
+++ b/frontend/src/components/OccupationRatingProfile.tsx
@@ -14,17 +14,23 @@ export function OccupationRatingProfile({ accessToken }: Props) {
   const [profile, setProfile] = useState<OccupationRatingProfilePayload | null>(null);
   const [status, setStatus] = useState<"idle" | "loading" | "error">("idle");
 
-  function load(offset = 0) {
+  function load(offset: number | null = null) {
+    const request = offset == null
+      ? { onetsocCode, dataReleaseCode: releaseCode, sourceTableCode: sourceCode }
+      : {
+          onetsocCode: profile?.onetsoc_code ?? onetsocCode,
+          dataReleaseCode: profile?.data_release_code ?? releaseCode,
+          sourceTableCode: profile?.source_table_code ?? sourceCode,
+        };
+    if (offset == null) setProfile(null);
     setStatus("loading");
     fetchOccupationRatings(accessToken, {
-      onetsocCode,
-      dataReleaseCode: releaseCode,
-      sourceTableCode: sourceCode,
-      offset,
+      ...request,
+      offset: offset ?? 0,
     })
       .then((payload) => {
         setProfile((current) =>
-          offset && current
+          offset != null && current
             ? { ...payload, items: [...current.items, ...payload.items] }
             : payload,
         );
@@ -78,7 +84,7 @@ export function OccupationRatingProfile({ accessToken }: Props) {
           className="btn-secondary"
           type="button"
           disabled={status === "loading"}
-          onClick={() => load(profile.next_offset ?? 0)}
+          onClick={() => load(profile.next_offset)}
         >
           다음 관측값 불러오기
         </button>

From e6a1d6217fb9491c6c15882573e94090c2ede071 Mon Sep 17 00:00:00 2001
From: Codex <codex@localhost>
Date: Thu, 27 Aug 2026 06:00:13 +0900
Subject: [PATCH 11/17] docs: keep occupation rating ADR lint-clean

---
 docs/adr/0258-occupation-rating-read-api.md | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/docs/adr/0258-occupation-rating-read-api.md b/docs/adr/0258-occupation-rating-read-api.md
index a22f5bab1..7af31c817 100644
--- a/docs/adr/0258-occupation-rating-read-api.md
+++ b/docs/adr/0258-occupation-rating-read-api.md
@@ -1,7 +1,8 @@
 # ADR 0258: Authenticated occupation-rating source read API
 
-**Status:** Accepted  
-**Date:** 2026-08-27  
+**Status:** Accepted
+
+**Date:** 2026-08-27
 **Extends:** ADR 0120, ADR 0184, ADR 0257
 
 ## Context

From e2042093b572ed74120ea090828e484c7e11f1be Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Wed, 26 Aug 2026 14:00:28 -0700
Subject: [PATCH 12/17] feat(ontology): expose occupation rating evidence
 (#738)

* feat(ontology): expose occupation rating evidence

* docs: keep occupation rating ADR lint-clean

---------

Co-authored-by: Codex <codex@localhost>
---
 ARCHITECTURE.md                             |   1 +
 CHANGELOG.md                                |   5 +
 backend/app/main.py                         |  29 +++-
 backend/app/occupation_rating_ingestion.py  | 127 +++++++++++++++
 docs/adr/0258-occupation-rating-read-api.md |  50 ++++++
 docs/adr/README.md                          |   1 +
 docs/product-requirements.md                |  13 ++
 docs/product-technical-gap-baseline.md      |   2 +-
 tests/test_occupation_rating_ingestion.py   | 162 ++++++++++++++++++++
 tests/test_schema.py                        |  20 +++
 10 files changed, 408 insertions(+), 2 deletions(-)
 create mode 100644 backend/app/occupation_rating_ingestion.py
 create mode 100644 docs/adr/0258-occupation-rating-read-api.md
 create mode 100644 tests/test_occupation_rating_ingestion.py

diff --git a/ARCHITECTURE.md b/ARCHITECTURE.md
index 55e00b78d..6a24bce13 100644
--- a/ARCHITECTURE.md
+++ b/ARCHITECTURE.md
@@ -80,6 +80,7 @@ flowchart LR
 | `temporal_expressions.py` | Pure Korean relative-time resolver for Global Ask (ADR 0150) |
 | `ask_time_axis.py` | Event-time vs ingestion-time clock choice for that window (ADR 0202) |
 | `ontology.py` | Loads the governed Turtle source tree (`lineageweave-kg.ttl` plus generated fragments), the formal OWL 2/RDFS/SKOS vocabulary for the Knowledge Graph's node/edge types, source taxonomies, and published O*NET linkages (ADR 0004, ADR 0252, ADR 0255, ADR 0256) |
+| `backend/app/occupation_rating_ingestion.py` | Projects authenticated, bounded occupation-rating source evidence with exact decimal and artifact provenance semantics (ADR 0258) |
 | `ontology_neighborhood.py` | Bounded typed ontology/provenance neighborhood (ADR 0184); PostgreSQL stays authoritative, OWL subclass is not an instance edge |
 | `ontology_source_cursor.py` | Opaque HMAC source-window continuation (ADR 0124); keyset pagination, never OFFSET |
 | `period_report.py` | Fit GRM/GPCM on persisted IRT rows, FIPC-select, EAP-score a period (ADR 0003 slice 3; Bock & Mislevy, 1982) |
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 79869185c..15193b4e0 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -8,6 +8,11 @@ All notable changes to this project are documented here. Format follows
 
 ### Added
 
+- Authenticated occupation profiles can now read one exact imported
+  release/source with rating and scale artifact provenance, exact decimal
+  strings, uncertainty, suppression/relevance flags, and explicit unavailable
+  versus empty outcomes. The endpoint derives no ranking or recommendation
+  (ADR 0258).
 - O*NET occupation-rating source evidence now has a replay-safe PostgreSQL
   contract with normalized release, source-table, scale, occupation, element,
   and observation tables. Exact release/source LIST partitions fail closed;
diff --git a/backend/app/main.py b/backend/app/main.py
index 6457bbde1..60dd9e1a5 100644
--- a/backend/app/main.py
+++ b/backend/app/main.py
@@ -30,7 +30,7 @@
 
 import asyncpg
 import redis.asyncio as redis
-from fastapi import Depends, FastAPI, HTTPException, Query, status
+from fastapi import Depends, FastAPI, HTTPException, Path, Query, status
 from fastapi.middleware.cors import CORSMiddleware
 from pydantic import BaseModel
 
@@ -98,6 +98,7 @@
     upsert_commitment_ticket,
 )
 from backend.app.operations_dashboard import fetch_operations_dashboard
+from backend.app.occupation_rating_ingestion import fetch_occupation_ratings
 from backend.app.keyman_ingestion import ingest_post_keymen
 from backend.app.knowledge_graph import (
     corporate_entity_exists,
@@ -2275,6 +2276,32 @@ async def read_ontology_neighborhood(
     return payload
 
 
+@app.get("/api/occupations/{onetsoc_code}/ratings")
+async def read_occupation_ratings(
+    onetsoc_code: str = Path(..., pattern=r"^[0-9]{2}-[0-9]{4}\.[0-9]{2}$"),
+    data_release_code: str = Query(
+        ..., min_length=1, max_length=63, pattern=r"^[a-z0-9][a-z0-9.-]*$"
+    ),
+    source_table_code: str = Query(
+        ..., min_length=1, max_length=63, pattern=r"^[a-z][a-z0-9_]*$"
+    ),
+    limit: int = Query(100, ge=1, le=500),
+    offset: int = Query(0, ge=0, le=10000),
+    _account: CurrentAccount = Depends(get_current_account),
+    pool: asyncpg.Pool = Depends(get_pool),
+) -> dict[str, object]:
+    """Return one authenticated, provenance-bearing occupation source profile."""
+    async with pool.acquire() as conn:
+        return await fetch_occupation_ratings(
+            conn,
+            data_release_code=data_release_code,
+            source_table_code=source_table_code,
+            onetsoc_code=onetsoc_code,
+            limit=limit,
+            offset=offset,
+        )
+
+
 @app.get("/api/posts/{post_id}/counterparties")
 async def read_post_counterparties(
     post_id: str,
diff --git a/backend/app/occupation_rating_ingestion.py b/backend/app/occupation_rating_ingestion.py
new file mode 100644
index 000000000..320fe248d
--- /dev/null
+++ b/backend/app/occupation_rating_ingestion.py
@@ -0,0 +1,127 @@
+"""Read exact imported occupation ratings without deriving a score or weight."""
+
+from __future__ import annotations
+
+from decimal import Decimal
+from typing import Any, Protocol
+
+
+class RatingReadConnection(Protocol):
+    """Small asyncpg-compatible surface used by the rating read projection."""
+
+    async def fetchrow(self, query: str, *args: object) -> Any:
+        """Return one row or ``None``."""
+
+    async def fetch(self, query: str, *args: object) -> list[Any]:
+        """Return ordered rows."""
+
+
+def _decimal_text(value: Decimal | None) -> str | None:
+    """Return the exact database decimal representation or honest absence."""
+    return str(value) if value is not None else None
+
+
+async def fetch_occupation_ratings(
+    conn: RatingReadConnection,
+    *,
+    data_release_code: str,
+    source_table_code: str,
+    onetsoc_code: str,
+    limit: int,
+    offset: int,
+) -> dict[str, object]:
+    """Return one bounded source profile and explicit artifact availability."""
+    source = await conn.fetchrow(
+        """select rating_source.source_table_name,
+                  rating_source.source_artifact_url,
+                  rating_source.source_artifact_sha256,
+                  rating_source.source_row_count,
+                  scale_source.source_artifact_url as scale_artifact_url,
+                  scale_source.source_artifact_sha256 as scale_artifact_sha256,
+                  scale_source.source_row_count as scale_source_row_count
+             from occupational_source_table rating_source
+             left join occupational_source_table scale_source
+               on scale_source.data_release_code = rating_source.data_release_code
+              and scale_source.source_table_code = 'scales_reference'
+            where rating_source.data_release_code = $1
+              and rating_source.source_table_code = $2""",
+        data_release_code,
+        source_table_code,
+    )
+    if source is None:
+        return {
+            "data_release_code": data_release_code,
+            "source_table_code": source_table_code,
+            "onetsoc_code": onetsoc_code,
+            "source_available": False,
+            "source": None,
+            "items": [],
+            "next_offset": None,
+        }
+    rows = await conn.fetch(
+        """select observation.element_id, element.element_name,
+                  observation.scale_id, scale.scale_name,
+                  scale.minimum_value, scale.maximum_value,
+                  observation.category_value, observation.data_value,
+                  observation.sample_size, observation.standard_error,
+                  observation.lower_ci_bound, observation.upper_ci_bound,
+                  observation.recommend_suppress, observation.not_relevant,
+                  observation.source_updated_month, observation.domain_source_code
+             from occupational_rating_observation observation
+             join occupational_element_definition element
+               on element.data_release_code = observation.data_release_code
+              and element.element_id = observation.element_id
+             join occupational_scale_definition scale
+               on scale.data_release_code = observation.data_release_code
+              and scale.scale_id = observation.scale_id
+            where observation.data_release_code = $1
+              and observation.source_table_code = $2
+              and observation.onetsoc_code = $3
+            order by observation.element_id, observation.scale_id,
+                     observation.category_value nulls first
+            limit $4 offset $5""",
+        data_release_code,
+        source_table_code,
+        onetsoc_code,
+        limit + 1,
+        offset,
+    )
+    page = rows[:limit]
+    items = [
+        {
+            "element_id": row["element_id"],
+            "element_name": row["element_name"],
+            "scale_id": row["scale_id"],
+            "scale_name": row["scale_name"],
+            "minimum_value": _decimal_text(row["minimum_value"]),
+            "maximum_value": _decimal_text(row["maximum_value"]),
+            "category_value": row["category_value"],
+            "data_value": _decimal_text(row["data_value"]),
+            "sample_size": row["sample_size"],
+            "standard_error": _decimal_text(row["standard_error"]),
+            "lower_ci_bound": _decimal_text(row["lower_ci_bound"]),
+            "upper_ci_bound": _decimal_text(row["upper_ci_bound"]),
+            "recommend_suppress": row["recommend_suppress"],
+            "not_relevant": row["not_relevant"],
+            "source_updated_month": row["source_updated_month"],
+            "domain_source_code": row["domain_source_code"],
+        }
+        for row in page
+    ]
+    return {
+        "data_release_code": data_release_code,
+        "source_table_code": source_table_code,
+        "onetsoc_code": onetsoc_code,
+        "source_available": True,
+        "source": {
+            "source_table_name": source["source_table_name"],
+            "source_artifact_url": source["source_artifact_url"],
+            "source_artifact_sha256": source["source_artifact_sha256"],
+            "source_row_count": source["source_row_count"],
+            "scale_artifact_url": source["scale_artifact_url"],
+            "scale_artifact_sha256": source["scale_artifact_sha256"],
+            "scale_source_row_count": source["scale_source_row_count"],
+        },
+        "items": items,
+        "next_offset": offset + limit if len(rows) > limit else None,
+    }
diff --git a/docs/adr/0258-occupation-rating-read-api.md b/docs/adr/0258-occupation-rating-read-api.md
new file mode 100644
index 000000000..7af31c817
--- /dev/null
+++ b/docs/adr/0258-occupation-rating-read-api.md
@@ -0,0 +1,50 @@
+# ADR 0258: Authenticated occupation-rating source read API
+
+**Status:** Accepted
+
+**Date:** 2026-08-27
+**Extends:** ADR 0120, ADR 0184, ADR 0257
+
+## Context
+
+ADR 0257 preserves released occupation-to-element observations, but a database
+import alone does not let a product user inspect what a job profile says. A
+read contract must distinguish an unimported source from an imported source
+with no row for one occupation, preserve low-precision and not-relevant flags,
+and avoid presenting a published rating as a local weight or recommendation.
+
+## Decision
+
+1. Add an authenticated, read-only occupation-rating endpoint. O*NET source
+   observations are licensed public reference data and are not tenant records;
+   any authenticated LineageWeave account may read an imported artifact.
+2. Require exact release, source-table, and O*NET-SOC codes. Return
+   `source_available=false` when that pinned artifact is not imported; return
+   `source_available=true` with an empty item list when it is imported but has
+   no observation for the requested occupation.
+3. Return the rating and Scales Reference artifact URLs, SHA-256 values, and
+   row counts. Every observation retains element/scale identity, declared
+   bounds, optional category, exact decimal strings, sample/error/interval,
+   suppression, relevance, source month, and domain source.
+4. Order by element, scale, and category and use bounded offset pagination.
+   Per-occupation source partitions bound this projection; a cursor needs a
+   later decision only if measured production latency requires it.
+5. Do not aggregate, rank, normalize, infer person traits, or recommend an
+   occupation. Suppressed values remain visible with the suppression flag so a
+   user can audit the source without mistaking low precision for absence.
+6. API and frontend copy describe the evidence and the user's next action,
+   never importer, partition, model-provider, or orchestration internals.
+
+## Consequences
+
+The semantic layer gains an honest product read boundary without duplicating
+psychometric arithmetic. An accessible UI and its Storybook states remain a
+separate delivery step after this API has authenticated runtime evidence.
+
+## References
+
+National Center for O*NET Development. (2026). *O*NET 31.0 database* [Data
+set]. https://www.onetcenter.org/database.html
+
+PostgreSQL Global Development Group. (2026). *PostgreSQL 18 documentation:
+Queries—limit and offset*. https://www.postgresql.org/docs/current/queries-limit.html
diff --git a/docs/adr/README.md b/docs/adr/README.md
index 74e74592c..1a36fa1a8 100644
--- a/docs/adr/README.md
+++ b/docs/adr/README.md
@@ -34,6 +34,7 @@ decision from them.
 | [`SOC_2018_HIERARCHY_REFERENCES.md`](../doctoring/SOC_2018_HIERARCHY_REFERENCES.md) | [0252](0252-complete-2018-soc-hierarchy.md) |
 | [`ONET_31_LINKAGE_REFERENCES.md`](../doctoring/ONET_31_LINKAGE_REFERENCES.md) | [0256](0256-onet-content-model-published-linkages.md) |
 | [`ONET_RATING_STORE_REFERENCES.md`](../doctoring/ONET_RATING_STORE_REFERENCES.md) | [0257](0257-onet-occupation-rating-observation-store.md) |
+| Occupation-rating authenticated read projection | [0258](0258-occupation-rating-read-api.md) |
 
 [0011](0011-prov-o-standard-relations.md) and [0065](0065-prov-o-provenance-boundary.md) cite the dated W3C PROV-O and PROV-DM Recommendations (https://www.w3.org/TR/2013/REC-prov-o-20130430/ and https://www.w3.org/TR/2013/REC-prov-dm-20130430/).
 
diff --git a/docs/product-requirements.md b/docs/product-requirements.md
index 7e3e116e5..d84b284b0 100644
--- a/docs/product-requirements.md
+++ b/docs/product-requirements.md
@@ -148,6 +148,19 @@ repeated null-category UPSERT is idempotent.
 API, UI, and derived modeling remain unavailable until separate accepted
 delivery records.
 
+### PRD-FR-2E — Occupation-rating evidence read
+
+- Let an authenticated user open one exact release/source/occupation profile
+  with both rating and scale artifact provenance (ADR 0258).
+- Distinguish an unavailable imported source from an available source with no
+  observation for the occupation.
+- Preserve exact decimal text, uncertainty, suppression, relevance, source
+  month, domain source, and declared bounds; derive no ranking or recommendation.
+
+Acceptance: invalid identifiers and unbounded pages are rejected; an unavailable
+source never appears as a negative profile; pagination is deterministic; and a
+suppressed observation retains its value and warning flag together.
+
 ### PRD-FR-3 — Bounded ontology exploration
 
 - Apply RBAC/ABAC, source eligibility, and knowledge cutoff before graph
diff --git a/docs/product-technical-gap-baseline.md b/docs/product-technical-gap-baseline.md
index 51638152f..bd7e5a884 100644
--- a/docs/product-technical-gap-baseline.md
+++ b/docs/product-technical-gap-baseline.md
@@ -29,7 +29,7 @@ explicit unavailable state, not a reason to infer mappings from labels.
 | Occupation-to-construct relations | ADR 0257 defines a candidate 3NF, release/source-partitioned immutable observation store and deterministic pinned-CSV importer preserving value, optional category, sample/error/CI, suppression, relevance, exact `MM/YYYY` source update month, source digest, and domain provenance. The official O*NET 31.0 Abilities file (94,640 rows, 910 occupations, 52 elements; SHA-256 `7e9cd79791ce6014e1d26d0a449ae5b1e7aa7ef52d39b3934c3bb8d438104b88`) and all 33 Scales Reference rows (SHA-256 `bcba23858ce21ecaacbde303a8993e35d46724b4afb8c9ec2b10e04f42adcfc9`) imported into a throwaway local PostgreSQL database with all 94,640 observations, 55 suppression flags, 7,572 not-relevant flags, and source months from `12/2004` through `08/2026`; every scale retained `scales_reference` artifact provenance, the database was dropped afterward, and no corpus is committed or claimed deployed | Pass exact-head review/checks and protected merge; validate and import every selected official rating artifact through an authorized runtime, returning only aggregate evidence; never invent or locally normalize a weight |
 | Job-family and job-series semantics | No authoritative employer-specific job architecture is present | Define an organization-neutral import contract that preserves the authorized source hierarchy and distinguishes standard occupation codes from employer job families/series; no label-based binding |
 | Temporal and multilevel interpretation | Static vocabulary only; no person-level inference is asserted | Version valid and transaction time, preserve occupation/organization/unit nesting and multiple membership, and require TEPP or the owning Rust psychometric service before any calibrated temporal or multilevel result |
-| Product consumption | The read model has no persisted semantic-layer consumer or authenticated UI evidence | Add a provenance-bearing API and accessible ontology exploration flow, then verify synthetic Storybook edge states plus authenticated aggregate runtime evidence without exposing identifying records |
+| Product consumption | ADR 0258 defines a candidate authenticated occupation-rating API that distinguishes unavailable source from empty profile and returns exact decimal/uncertainty/warning/artifact provenance without ranking; accessible UI and authenticated runtime evidence remain absent | Pass exact-head review/checks and protected merge; add the accessible exploration UI, synthetic Storybook unavailable/empty/suppressed/paginated states, screenshots, and authenticated aggregate runtime evidence without exposing identifying records |
 
 ### Current exact-head PR queue
 
diff --git a/tests/test_occupation_rating_ingestion.py b/tests/test_occupation_rating_ingestion.py
new file mode 100644
index 000000000..3f2244d7f
--- /dev/null
+++ b/tests/test_occupation_rating_ingestion.py
@@ -0,0 +1,162 @@
+"""Tests for the provenance-bearing occupation-rating read projection."""
+
+import asyncio
+from decimal import Decimal
+
+from backend.app.main import read_occupation_ratings
+from backend.app.occupation_rating_ingestion import fetch_occupation_ratings
+
+
+class FakeConnection:
+    """Minimal ordered asyncpg stand-in for one projection query."""
+
+    def __init__(self, source, rows=()):
+        self.source = source
+        self.rows = list(rows)
+        self.fetch_called = False
+
+    async def fetchrow(self, _query: str, *_args: object):
+        """Return configured source metadata."""
+        return self.source
+
+    async def fetch(self, _query: str, *_args: object):
+        """Return configured observation rows."""
+        self.fetch_called = True
+        return self.rows
+
+
+class FakeAcquire:
+    """Async pool-acquire context for route wiring."""
+
+    def __init__(self, conn: FakeConnection):
+        self.conn = conn
+
+    async def __aenter__(self) -> FakeConnection:
+        """Return the configured connection."""
+        return self.conn
+
+    async def __aexit__(self, *_args: object) -> None:
+        """Release without external state."""
+
+
+class FakePool:
+    """Minimal pool exposing one acquisition context."""
+
+    def __init__(self, conn: FakeConnection):
+        self.conn = conn
+
+    def acquire(self) -> FakeAcquire:
+        """Return one deterministic acquisition context."""
+        return FakeAcquire(self.conn)
+
+
+def test_unimported_source_is_not_an_empty_observed_profile() -> None:
+    conn = FakeConnection(None)
+
+    result = asyncio.run(
+        fetch_occupation_ratings(
+            conn,
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+            onetsoc_code="15-1252.00",
+            limit=100,
+            offset=0,
+        )
+    )
+
+    assert result["source_available"] is False
+    assert result["items"] == []
+    assert conn.fetch_called is False
+
+
+def test_rating_projection_preserves_exact_decimal_and_warning_flags() -> None:
+    source = {
+        "source_table_name": "Abilities",
+        "source_artifact_url": "https://example.test/abilities.csv",
+        "source_artifact_sha256": "a" * 64,
+        "source_row_count": 2,
+        "scale_artifact_url": "https://example.test/scales.csv",
+        "scale_artifact_sha256": "b" * 64,
+        "scale_source_row_count": 33,
+    }
+    row = {
+        "element_id": "1.A.1.a.1",
+        "element_name": "Oral Comprehension",
+        "scale_id": "IM",
+        "scale_name": "Importance",
+        "minimum_value": Decimal("1.00"),
+        "maximum_value": Decimal("5.00"),
+        "category_value": None,
+        "data_value": Decimal("4.10"),
+        "sample_size": 8,
+        "standard_error": Decimal("0.1830"),
+        "lower_ci_bound": Decimal("3.7414"),
+        "upper_ci_bound": Decimal("4.4586"),
+        "recommend_suppress": True,
+        "not_relevant": None,
+        "source_updated_month": "08/2026",
+        "domain_source_code": "Analyst",
+    }
+    conn = FakeConnection(source, (row, row))
+
+    result = asyncio.run(
+        fetch_occupation_ratings(
+            conn,
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+            onetsoc_code="15-1252.00",
+            limit=1,
+            offset=0,
+        )
+    )
+
+    item = result["items"][0]
+    assert item["data_value"] == "4.10"
+    assert item["standard_error"] == "0.1830"
+    assert item["recommend_suppress"] is True
+    assert item["not_relevant"] is None
+    assert result["source"]["scale_artifact_sha256"] == "b" * 64
+    assert result["next_offset"] == 1
+
+
+def test_empty_profile_keeps_imported_scale_provenance() -> None:
+    source = {
+        "source_table_name": "Abilities",
+        "source_artifact_url": "https://example.test/abilities.csv",
+        "source_artifact_sha256": "a" * 64,
+        "source_row_count": 2,
+        "scale_artifact_url": "https://example.test/scales.csv",
+        "scale_artifact_sha256": "b" * 64,
+        "scale_source_row_count": 33,
+    }
+
+    result = asyncio.run(
+        fetch_occupation_ratings(
+            FakeConnection(source),
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+            onetsoc_code="15-9999.99",
+            limit=100,
+            offset=0,
+        )
+    )
+
+    assert result["source_available"] is True
+    assert result["items"] == []
+    assert result["source"]["scale_artifact_sha256"] == "b" * 64
+
+
+def test_authenticated_route_delegates_to_bounded_projection() -> None:
+    result = asyncio.run(
+        read_occupation_ratings(
+            onetsoc_code="15-1252.00",
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+            limit=100,
+            offset=0,
+            _account=object(),
+            pool=FakePool(FakeConnection(None)),
+        )
+    )
+
+    assert result["source_available"] is False
diff --git a/tests/test_schema.py b/tests/test_schema.py
index 15e94591b..d12fa59de 100644
--- a/tests/test_schema.py
+++ b/tests/test_schema.py
@@ -28,6 +28,7 @@
 import psycopg2.errors
 import pytest
 
+from backend.app.occupation_rating_ingestion import fetch_occupation_ratings
 from backend.app.post_chat_ingestion import gather_global_chat_sources
 from scripts.import_onet_ratings import import_ratings
 
@@ -527,6 +528,25 @@ def test_onet_rating_importer_is_idempotent_against_postgresql(
         )
         assert cur.fetchone() == (1, Decimal("4.10"), True)
 
+    async def read_imported_profile() -> dict[str, object]:
+        conn = await asyncpg.connect(args.target_dsn)
+        try:
+            return await fetch_occupation_ratings(
+                conn,
+                data_release_code=args.release_code,
+                source_table_code=args.source_table_code,
+                onetsoc_code="15-1252.00",
+                limit=100,
+                offset=0,
+            )
+        finally:
+            await conn.close()
+
+    profile = asyncio.run(read_imported_profile())
+    assert profile["source_available"] is True
+    assert profile["items"][0]["data_value"] == "4.10"
+    assert profile["source"]["scale_artifact_sha256"] == args.scales_sha256
+
 
 def test_global_ask_evidence_search_indexes_exist_on_normalized_tables(schema_db) -> None:
     """The real PostgreSQL schema owns all nine evidence-search indexes."""

From 4f7b928cb14276353e84e21d6fe92793dac264ba Mon Sep 17 00:00:00 2001
From: Codex <codex@localhost>
Date: Thu, 27 Aug 2026 06:06:03 +0900
Subject: [PATCH 13/17] fix(ui): validate occupation evidence links

---
 .../OccupationRatingProfile.test.tsx          | 10 +++++++++
 .../components/OccupationRatingProfile.tsx    | 22 ++++++++++++++++---
 2 files changed, 29 insertions(+), 3 deletions(-)

diff --git a/frontend/src/components/OccupationRatingProfile.test.tsx b/frontend/src/components/OccupationRatingProfile.test.tsx
index cddaec717..3b4913992 100644
--- a/frontend/src/components/OccupationRatingProfile.test.tsx
+++ b/frontend/src/components/OccupationRatingProfile.test.tsx
@@ -92,4 +92,14 @@ describe("OccupationRatingProfile", () => {
     rerender(<OccupationRatingProfileView profile={{ ...ready, items: [] }} />);
     expect(screen.getByRole("status")).toHaveTextContent("관측값이 없습니다");
   });
+
+  it("does not turn a non-http artifact value into a customer link", () => {
+    render(<OccupationRatingProfileView profile={{
+      ...ready,
+      source: { ...ready.source!, source_artifact_url: "javascript:alert(1)" },
+    }} />);
+
+    expect(screen.queryByRole("link", { name: "평정 원문 열기" })).not.toBeInTheDocument();
+    expect(screen.getByText(/데이터 담당자에게 출처 확인을 요청하세요/)).toBeInTheDocument();
+  });
 });
diff --git a/frontend/src/components/OccupationRatingProfile.tsx b/frontend/src/components/OccupationRatingProfile.tsx
index 20833f3ee..c40d94886 100644
--- a/frontend/src/components/OccupationRatingProfile.tsx
+++ b/frontend/src/components/OccupationRatingProfile.tsx
@@ -6,6 +6,16 @@ import {
 
 type Props = { accessToken: string };
 
+function safeHttpUrl(value: string | null | undefined): string | null {
+  if (!value) return null;
+  try {
+    const parsed = new URL(value);
+    return parsed.protocol === "http:" || parsed.protocol === "https:" ? parsed.toString() : null;
+  } catch {
+    return null;
+  }
+}
+
 /** Lets an authenticated user inspect one exact imported occupation profile. */
 export function OccupationRatingProfile({ accessToken }: Props) {
   const [onetsocCode, setOnetsocCode] = useState("");
@@ -113,14 +123,20 @@ export function OccupationRatingProfileView({
       </p>
     );
   }
+  const sourceArtifactUrl = safeHttpUrl(profile.source?.source_artifact_url);
+  const scaleArtifactUrl = safeHttpUrl(profile.source?.scale_artifact_url);
   return (
     <>
       <div className="occupation-rating-source">
         <strong>{profile.source?.source_table_name}</strong>
         <span>{profile.data_release_code} · {profile.onetsoc_code}</span>
-        <a href={profile.source?.source_artifact_url} target="_blank" rel="noreferrer">평정 원문 열기</a>
-        {profile.source?.scale_artifact_url ? (
-          <a href={profile.source.scale_artifact_url} target="_blank" rel="noreferrer">척도 정의 열기</a>
+        {sourceArtifactUrl ? (
+          <a href={sourceArtifactUrl} target="_blank" rel="noreferrer">평정 원문 열기</a>
+        ) : (
+          <span>원문 링크를 사용할 수 없습니다. 데이터 담당자에게 출처 확인을 요청하세요.</span>
+        )}
+        {scaleArtifactUrl ? (
+          <a href={scaleArtifactUrl} target="_blank" rel="noreferrer">척도 정의 열기</a>
         ) : null}
       </div>
       <p className="occupation-rating-scroll-hint">표를 가로로 밀어 오차와 사용 주의를 확인하세요.</p>

From b24e8fa36e60f0029bfea7cc161b24c886da27da Mon Sep 17 00:00:00 2001
From: Codex <codex@localhost>
Date: Thu, 27 Aug 2026 06:10:28 +0900
Subject: [PATCH 14/17] fix(ui): fence superseded occupation evidence

---
 .../OccupationRatingProfile.test.tsx          | 22 ++++++++++++++++++-
 .../components/OccupationRatingProfile.tsx    | 10 +++++++--
 2 files changed, 29 insertions(+), 3 deletions(-)

diff --git a/frontend/src/components/OccupationRatingProfile.test.tsx b/frontend/src/components/OccupationRatingProfile.test.tsx
index 3b4913992..93166047d 100644
--- a/frontend/src/components/OccupationRatingProfile.test.tsx
+++ b/frontend/src/components/OccupationRatingProfile.test.tsx
@@ -1,4 +1,4 @@
-import { render, screen } from "@testing-library/react";
+import { fireEvent, render, screen } from "@testing-library/react";
 import userEvent from "@testing-library/user-event";
 import { describe, expect, it, vi } from "vitest";
 import { fetchOccupationRatings, type OccupationRatingProfile as Payload } from "../api";
@@ -86,6 +86,26 @@ describe("OccupationRatingProfile", () => {
     expect(screen.getByRole("button", { name: "근거를 불러오는 중" })).toBeDisabled();
   });
 
+  it("ignores a superseded occupation response that finishes last", async () => {
+    let finishFirst: ((profile: Payload) => void) | undefined;
+    vi.mocked(fetchOccupationRatings)
+      .mockImplementationOnce(() => new Promise((resolve) => { finishFirst = resolve; }))
+      .mockResolvedValueOnce({ ...ready, onetsoc_code: "11-1011.00", items: [{ ...ready.items[0], data_value: "3.20" }] });
+    render(<OccupationRatingProfile accessToken="synthetic-token" />);
+    const occupation = screen.getByLabelText("O*NET-SOC 직업 코드");
+    await userEvent.type(occupation, "15-1252.00");
+    await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
+
+    await userEvent.clear(occupation);
+    await userEvent.type(occupation, "11-1011.00");
+    fireEvent.submit(occupation.closest("form")!);
+    expect(await screen.findByText("3.20")).toBeInTheDocument();
+
+    finishFirst?.(ready);
+    expect(screen.queryByText("4.10")).not.toBeInTheDocument();
+    expect(screen.getByText("3.20")).toBeInTheDocument();
+  });
+
   it("distinguishes an unavailable artifact from an empty occupation profile", () => {
     const { rerender } = render(<OccupationRatingProfileView profile={{ ...ready, source_available: false, source: null, items: [] }} />);
     expect(screen.getByRole("status")).toHaveTextContent("아직 준비되지 않았습니다");
diff --git a/frontend/src/components/OccupationRatingProfile.tsx b/frontend/src/components/OccupationRatingProfile.tsx
index c40d94886..8ce8a7fc8 100644
--- a/frontend/src/components/OccupationRatingProfile.tsx
+++ b/frontend/src/components/OccupationRatingProfile.tsx
@@ -1,4 +1,4 @@
-import { useState } from "react";
+import { useRef, useState } from "react";
 import {
   fetchOccupationRatings,
   type OccupationRatingProfile as OccupationRatingProfilePayload,
@@ -23,8 +23,11 @@ export function OccupationRatingProfile({ accessToken }: Props) {
   const [sourceCode, setSourceCode] = useState("abilities");
   const [profile, setProfile] = useState<OccupationRatingProfilePayload | null>(null);
   const [status, setStatus] = useState<"idle" | "loading" | "error">("idle");
+  const requestSequence = useRef(0);
 
   function load(offset: number | null = null) {
+    const requestId = requestSequence.current + 1;
+    requestSequence.current = requestId;
     const request = offset == null
       ? { onetsocCode, dataReleaseCode: releaseCode, sourceTableCode: sourceCode }
       : {
@@ -39,6 +42,7 @@ export function OccupationRatingProfile({ accessToken }: Props) {
       offset: offset ?? 0,
     })
       .then((payload) => {
+        if (requestSequence.current !== requestId) return;
         setProfile((current) =>
           offset != null && current
             ? { ...payload, items: [...current.items, ...payload.items] }
@@ -46,7 +50,9 @@ export function OccupationRatingProfile({ accessToken }: Props) {
         );
         setStatus("idle");
       })
-      .catch(() => setStatus("error"));
+      .catch(() => {
+        if (requestSequence.current === requestId) setStatus("error");
+      });
   }
 
   return (

From e6efa6fe0d622eca55adee1dcdefc71a0eb430ed Mon Sep 17 00:00:00 2001
From: Codex <codex@localhost>
Date: Thu, 27 Aug 2026 06:34:21 +0900
Subject: [PATCH 15/17] fix(ui): hide stale occupation pagination

---
 frontend/src/components/OccupationRatingProfile.test.tsx | 9 +++++----
 frontend/src/components/OccupationRatingProfile.tsx      | 6 +++++-
 2 files changed, 10 insertions(+), 5 deletions(-)

diff --git a/frontend/src/components/OccupationRatingProfile.test.tsx b/frontend/src/components/OccupationRatingProfile.test.tsx
index 93166047d..11334c948 100644
--- a/frontend/src/components/OccupationRatingProfile.test.tsx
+++ b/frontend/src/components/OccupationRatingProfile.test.tsx
@@ -48,7 +48,7 @@ describe("OccupationRatingProfile", () => {
     expect(screen.getByText(/표를 가로로 밀어/)).toBeInTheDocument();
   });
 
-  it("keeps pagination bound to the loaded profile after form edits", async () => {
+  it("hides stale pagination after form edits and loads the new profile", async () => {
     vi.mocked(fetchOccupationRatings)
       .mockResolvedValueOnce({ ...ready, next_offset: 100 })
       .mockResolvedValueOnce({ ...ready, items: [{ ...ready.items[0], scale_id: "LV" }] });
@@ -60,12 +60,13 @@ describe("OccupationRatingProfile", () => {
 
     await userEvent.clear(occupation);
     await userEvent.type(occupation, "11-1011.00");
-    await userEvent.click(screen.getByRole("button", { name: "다음 관측값 불러오기" }));
+    expect(screen.queryByRole("button", { name: "다음 관측값 불러오기" })).not.toBeInTheDocument();
+    await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
 
     expect(fetchOccupationRatings).toHaveBeenLastCalledWith("synthetic-token", {
-      onetsocCode: "15-1252.00", dataReleaseCode: "onet-31.0", sourceTableCode: "abilities", offset: 100,
+      onetsocCode: "11-1011.00", dataReleaseCode: "onet-31.0", sourceTableCode: "abilities", offset: 0,
     });
-    expect(await screen.findAllByText("4.10")).toHaveLength(2);
+    expect(await screen.findByText("4.10")).toBeInTheDocument();
   });
 
   it("removes stale evidence while a fresh occupation loads", async () => {
diff --git a/frontend/src/components/OccupationRatingProfile.tsx b/frontend/src/components/OccupationRatingProfile.tsx
index 8ce8a7fc8..2d1445187 100644
--- a/frontend/src/components/OccupationRatingProfile.tsx
+++ b/frontend/src/components/OccupationRatingProfile.tsx
@@ -24,6 +24,10 @@ export function OccupationRatingProfile({ accessToken }: Props) {
   const [profile, setProfile] = useState<OccupationRatingProfilePayload | null>(null);
   const [status, setStatus] = useState<"idle" | "loading" | "error">("idle");
   const requestSequence = useRef(0);
+  const profileMatchesForm = profile != null
+    && profile.onetsoc_code === onetsocCode
+    && profile.data_release_code === releaseCode
+    && profile.source_table_code === sourceCode;
 
   function load(offset: number | null = null) {
     const requestId = requestSequence.current + 1;
@@ -95,7 +99,7 @@ export function OccupationRatingProfile({ accessToken }: Props) {
         <p role="alert">직업 근거를 불러오지 못했습니다. 코드와 접근 권한을 확인한 뒤 다시 시도하세요.</p>
       ) : null}
       {profile ? <OccupationRatingProfileView profile={profile} /> : null}
-      {profile?.next_offset != null ? (
+      {profileMatchesForm && profile.next_offset != null ? (
         <button
           className="btn-secondary"
           type="button"

From 8dd5adb2295e8ae0a0254069d2e4610edac8dba0 Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Wed, 26 Aug 2026 14:56:19 -0700
Subject: [PATCH 16/17] feat(ontology): define O*NET occupation rating store
 (#734)

* feat(ontology): define O*NET occupation rating store

* fix(ontology): preserve O*NET scale artifact provenance

* fix(schema): preserve O*NET update month

* fix(schema): reject rating evidence mutation

* test: cover O*NET category percentages

---------

Co-authored-by: Codex <codex@localhost>
---
 CHANGELOG.md                                  |   7 +
 ...net-occupation-rating-observation-store.md |  80 ++++++
 docs/adr/README.md                            |   1 +
 .../doctoring/ONET_RATING_STORE_REFERENCES.md |  26 ++
 docs/product-requirements.md                  |  21 ++
 docs/product-technical-gap-baseline.md        |   4 +-
 .../0222_onet_rating_observation_store.sql    | 226 +++++++++++++++++
 tests/test_onet_rating_schema.py              |  34 +++
 tests/test_schema.py                          | 239 ++++++++++++++++++
 9 files changed, 636 insertions(+), 2 deletions(-)
 create mode 100644 docs/adr/0257-onet-occupation-rating-observation-store.md
 create mode 100644 docs/doctoring/ONET_RATING_STORE_REFERENCES.md
 create mode 100644 migrations/0222_onet_rating_observation_store.sql
 create mode 100644 tests/test_onet_rating_schema.py

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 552afeebe..d8a2f0f2b 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -8,6 +8,13 @@ All notable changes to this project are documented here. Format follows
 
 ### Added
 
+- O*NET occupation-rating source evidence now has a replay-safe PostgreSQL
+  contract with normalized release, source-table, scale, occupation, element,
+  and observation tables. Exact release/source LIST partitions fail closed;
+  nullable categories remain idempotent identities, divergent duplicates and
+  truncation fail closed, and no source value is promoted to a local weight or
+  person score. Task Ratings remain outside this content-model-element store
+  pending their own normalized identity contract (ADR 0257).
 - All eight O*NET 31.0 published linkage tables now contribute 1,417 directed
   Ability/Essential Skill/Transferable Skill/Work Style relations to Work
   Activities and Work Context. Every direct relation has an exact reified
diff --git a/docs/adr/0257-onet-occupation-rating-observation-store.md b/docs/adr/0257-onet-occupation-rating-observation-store.md
new file mode 100644
index 000000000..aae24d860
--- /dev/null
+++ b/docs/adr/0257-onet-occupation-rating-observation-store.md
@@ -0,0 +1,80 @@
+# ADR 0257: O*NET occupation-rating observation store
+
+**Status:** Accepted  
+**Date:** 2026-08-27  
+**Extends:** ADR 0166, ADR 0255, ADR 0256
+
+## Context
+
+O*NET 31.0 publishes occupation-specific ratings for content-model abilities, essential and
+transferable skills, knowledge, education, training and experience, interests,
+work styles, work activities, work context, and adjacent content-model
+domains. A rating is not merely an edge: its meaning depends on the release,
+source table, occupation, element, scale, optional response category, sample
+size, standard error, 95% confidence interval, precision-suppression flag,
+relevance flag, source update month, and domain source. O*NET publishes that
+field as a seven-character `MM/YYYY` value; coercing it to a database date
+would invent a day (National Center for O*NET Development, 2026b).
+
+Flattening those attributes into ontology predicates would erase measurement
+and provenance boundaries. Loading them into Python mathematical code would
+also violate LineageWeave's externalized-compute boundary.
+
+## Decision
+
+1. Store source releases, source tables, scales, occupations, content-model
+   elements, and rating observations in separate third-normal-form tables.
+2. Preserve published numeric values exactly as decimal observations. They are
+   source ratings, never locally estimated weights, person scores, causal
+   effects, or calibrated psychometric parameters.
+3. An observation key is release + source table + occupation + element + scale
+   + optional category. PostgreSQL `UNIQUE NULLS NOT DISTINCT` keeps an absent
+   category an honest single absence rather than replacing it with a sentinel.
+4. Partition observations first by release and then by source-table code.
+   These are authoritative lifecycle/query boundaries and require no invented
+   hash modulus. An importer must create both exact LIST partitions before
+   inserting; without them PostgreSQL rejects the row.
+5. Every insert uses the owning release/table artifact digest and idempotent
+   `ON CONFLICT DO NOTHING`. An exact duplicate is idempotent, while a row with
+   the same identity and different source values fails closed. Endpoint names
+   and scale names must match their normalized reference
+   rows; each scale definition names its owning source-table artifact, and the
+   importer rejects disagreement rather than overwriting identity.
+6. Preserve `recommend_suppress` and `not_relevant` independently. A suppressed
+   value remains stored with its warning; a not-relevant value is not converted
+   to zero. Missing `n`, error, or interval values remain null.
+7. Range and uncertainty constraints reject negative sample/error values,
+   inverted confidence intervals, malformed or future source update months, malformed source
+   digests, and values outside their declared scale bounds before persistence. Scale bounds
+   govern the published `Data Value`, not the optional response-category code. In particular,
+   O*NET's `CXP` rows store a category in `Category` while `Data Value` is the percentage that
+   endorsed it, so the authoritative `CXP` bounds are 0 through 100 (National Center for
+   O*NET Development, 2026b).
+8. This content-model-element store excludes Task Ratings, whose integer Task
+   IDs and task-statement identity require a separate normalized target table;
+   it does not reinterpret a Task ID as a content-model element.
+9. This store is immutable source evidence. Row mutation and whole-store
+   truncation fail closed. Any later aggregation, comparison,
+   temporal model, multilevel model, or occupational recommendation belongs to
+   TEPP/fast-mlsirm or another owning Rust service and must cite these rows.
+
+## Consequences
+
+LineageWeave can import the governed public O*NET content-model rating corpus without
+manufacturing semantics or embedding large production datasets in git.
+Release/source partitions localize hot imports and permit exact detach/archive
+operations. A separate API/UI ADR is still required before exposing ratings.
+
+## References
+
+National Center for O*NET Development. (2026a). *O*NET 31.0 database* [Data
+set]. https://www.onetcenter.org/database.html
+
+National Center for O*NET Development. (2026b). *Work context: O*NET 31.0 data
+dictionary*. https://www.onetcenter.org/dictionary/31.0/excel/work_context.html
+
+PostgreSQL Global Development Group. (2026). *PostgreSQL 18 documentation:
+Table partitioning*. https://www.postgresql.org/docs/current/ddl-partitioning.html
+
+PostgreSQL Global Development Group. (2026). *PostgreSQL 18 documentation:
+CREATE TABLE*. https://www.postgresql.org/docs/current/sql-createtable.html
diff --git a/docs/adr/README.md b/docs/adr/README.md
index 441f62517..74e74592c 100644
--- a/docs/adr/README.md
+++ b/docs/adr/README.md
@@ -33,6 +33,7 @@ decision from them.
 | [`ONET_31_CONTENT_MODEL_REFERENCES.md`](../doctoring/ONET_31_CONTENT_MODEL_REFERENCES.md) | [0255](0255-onet-31-content-model-ontology.md) |
 | [`SOC_2018_HIERARCHY_REFERENCES.md`](../doctoring/SOC_2018_HIERARCHY_REFERENCES.md) | [0252](0252-complete-2018-soc-hierarchy.md) |
 | [`ONET_31_LINKAGE_REFERENCES.md`](../doctoring/ONET_31_LINKAGE_REFERENCES.md) | [0256](0256-onet-content-model-published-linkages.md) |
+| [`ONET_RATING_STORE_REFERENCES.md`](../doctoring/ONET_RATING_STORE_REFERENCES.md) | [0257](0257-onet-occupation-rating-observation-store.md) |
 
 [0011](0011-prov-o-standard-relations.md) and [0065](0065-prov-o-provenance-boundary.md) cite the dated W3C PROV-O and PROV-DM Recommendations (https://www.w3.org/TR/2013/REC-prov-o-20130430/ and https://www.w3.org/TR/2013/REC-prov-dm-20130430/).
 
diff --git a/docs/doctoring/ONET_RATING_STORE_REFERENCES.md b/docs/doctoring/ONET_RATING_STORE_REFERENCES.md
new file mode 100644
index 000000000..6258ed2b2
--- /dev/null
+++ b/docs/doctoring/ONET_RATING_STORE_REFERENCES.md
@@ -0,0 +1,26 @@
+# O*NET occupation-rating store evidence
+
+This supporting note records the source and database capabilities governed by
+ADR 0257. It introduces no independent architecture decision.
+
+O*NET 31.0 occupation data tables publish occupation and element identifiers,
+scale identifiers and names, decimal values, optional category values, sample
+sizes, standard errors, confidence bounds, suppression/relevance flags, update
+dates, and domain sources. These fields remain source observations; they are
+not locally estimated psychometric weights.
+
+PostgreSQL declarative partitioning provides exact LIST boundaries and
+partition pruning. A unique constraint on a partitioned table includes its
+partition key; `UNIQUE NULLS NOT DISTINCT` makes a missing category one stable
+identity without a sentinel value.
+
+## APA 7 references
+
+National Center for O*NET Development. (2026). *O*NET 31.0 database* [Data
+set]. https://www.onetcenter.org/database.html
+
+PostgreSQL Global Development Group. (2026). *PostgreSQL 18 documentation:
+Table partitioning*. https://www.postgresql.org/docs/current/ddl-partitioning.html
+
+PostgreSQL Global Development Group. (2026). *PostgreSQL 18 documentation:
+CREATE TABLE*. https://www.postgresql.org/docs/current/sql-createtable.html
diff --git a/docs/product-requirements.md b/docs/product-requirements.md
index 82b60df78..dae982b4f 100644
--- a/docs/product-requirements.md
+++ b/docs/product-requirements.md
@@ -124,6 +124,27 @@ prohibit FJA equivalence, require exact Post/evidence/PROV statement structure,
 and reproduce every pinned O*NET linkage with its exact source table. Runtime
 persistence and UI remain unavailable until their separate ADR acceptance.
 
+### PRD-FR-2D — Occupation-rating source observations
+
+- Persist released occupation-to-element ratings as source observations, not
+  ontology weights: release, source table, occupation, element, scale,
+  optional category, value, sample/error/interval, suppression, relevance,
+  exact source update month, and domain source remain independently auditable
+  (ADR 0257); the product must not invent a day for O*NET's `MM/YYYY` field.
+- Keep normalized reference identities in third normal form and partition the
+  observation store by exact release then source table. An unknown partition
+  fails closed instead of entering a catch-all table.
+- Preserve decimals and missingness exactly. No local aggregation,
+  normalization, person inference, or psychometric estimation is permitted.
+- Reject divergent duplicate identities and owner-level truncation. Task
+  Ratings remain unavailable until their integer Task IDs and statements have
+  a separate normalized source-target contract.
+
+Acceptance: the replay-safe migration creates the normalized store; PostgreSQL
+integration proves missing partitions fail closed, null-category UPSERT is
+idempotent, and invalid uncertainty is rejected. Corpus import, API, UI, and
+derived modeling remain unavailable until separate accepted delivery records.
+
 ### PRD-FR-3 — Bounded ontology exploration
 
 - Apply RBAC/ABAC, source eligibility, and knowledge cutoff before graph
diff --git a/docs/product-technical-gap-baseline.md b/docs/product-technical-gap-baseline.md
index 30521e065..c99d17e75 100644
--- a/docs/product-technical-gap-baseline.md
+++ b/docs/product-technical-gap-baseline.md
@@ -26,7 +26,7 @@ explicit unavailable state, not a reason to infer mappings from labels.
 | Classification depth | ADR 0252's candidate branch imports all 1,447 official 2018 SOC nodes with source-declared hierarchy, pinned XLSX/CSV digests, a deterministic Turtle renderer, and fail-closed read model | Pass exact-head review/checks and protected merge; add ISCO/ESCO crosswalks only where the publishing authority supplies them |
 | Construct granularity | ADR 0255's candidate branch publishes all 3,006 O*NET 31.0 Content Model Reference concepts with exact IDs, names, descriptions, and source-defined parents; it imports no ratings or person assertions | Pass exact-head review/checks and protected merge; import separately released occupation-element observations only through the provenance contract below |
 | Construct-to-work relations | ADR 0256's candidate branch imports all 1,417 official Ability/Essential Skill/Transferable Skill/Work Style linkages to Work Activities and Work Context with assertion-level source provenance and no invented weights | Pass exact-head review/checks and protected merge; keep these published relevance links distinct from occupation ratings and causal claims |
-| Occupation-to-construct relations | No occupation rating is imported; ADR 0256 explicitly separates those observations from vocabulary linkages | Persist released source observations with source version, occupation code, element identifier, scale identifier, value, sample/error/CI/suppression metadata when supplied, and provenance; never invent or locally normalize a weight |
+| Occupation-to-construct relations | ADR 0257 defines a candidate 3NF, release/source-partitioned immutable observation store preserving value, optional category, sample/error/CI, suppression, relevance, exact `MM/YYYY` source update month, source digest, and domain provenance; no corpus is imported | Pass exact-head review/checks and protected merge; add a deterministic source importer for pinned public artifacts; never invent a date, weight, or local normalization |
 | Job-family and job-series semantics | No authoritative employer-specific job architecture is present | Define an organization-neutral import contract that preserves the authorized source hierarchy and distinguishes standard occupation codes from employer job families/series; no label-based binding |
 | Temporal and multilevel interpretation | Static vocabulary only; no person-level inference is asserted | Version valid and transaction time, preserve occupation/organization/unit nesting and multiple membership, and require TEPP or the owning Rust psychometric service before any calibrated temporal or multilevel result |
 | Product consumption | The read model has no persisted semantic-layer consumer or authenticated UI evidence | Add a provenance-bearing API and accessible ontology exploration flow, then verify synthetic Storybook edge states plus authenticated aggregate runtime evidence without exposing identifying records |
@@ -35,7 +35,7 @@ explicit unavailable state, not a reason to infer mappings from labels.
 
 | PR | Exact observed head | Base | Observed gate state |
 |---:|---|---|---|
-| #732 | `b2dd06f4` | `feat/onet-31-content-model-ontology` | unstable; full tests, frontend, and Devin Review queued; CodeRabbit passed |
+| #732 | `1c622545` | `feat/onet-31-content-model-ontology` | unstable; full tests in progress; frontend, CodeRabbit, and Devin Review passed; one informational review thread open |
 | #731 | `2c848373` | `feat/soc-2018-full-hierarchy` | unstable; full tests, frontend, and Devin Review queued |
 | #724 | `f5ee37b0` | `feat/io-occupational-taxonomy` | clean; full tests, frontend, CodeRabbit, and Devin Review passed; parent-first stack gate remains |
 | #719 | `6ee2278a` | `feat/io-psych-construct-ontology` | unstable; full tests, frontend, CodeRabbit, and Devin Review passed |
diff --git a/migrations/0222_onet_rating_observation_store.sql b/migrations/0222_onet_rating_observation_store.sql
new file mode 100644
index 000000000..b07a0461d
--- /dev/null
+++ b/migrations/0222_onet_rating_observation_store.sql
@@ -0,0 +1,226 @@
+-- ADR 0257: normalized, immutable O*NET occupation-rating source evidence.
+-- Release and source-table LIST partitions are created by the importer before
+-- data insertion; no default partition may silently absorb an unknown source.
+
+begin;
+
+create table if not exists occupational_data_release (
+    data_release_code text primary key,
+    release_version text not null,
+    source_publisher_name text not null,
+    source_license_url text not null,
+    imported_at timestamptz not null default now(),
+    constraint occupational_data_release_code_check
+        check (btrim(data_release_code) <> ''),
+    constraint occupational_release_version_check
+        check (btrim(release_version) <> '')
+);
+
+create table if not exists occupational_source_table (
+    data_release_code text not null,
+    source_table_code text not null,
+    source_table_name text not null,
+    source_artifact_url text not null,
+    source_artifact_sha256 text not null,
+    source_row_count bigint not null,
+    primary key (data_release_code, source_table_code),
+    constraint occupational_source_table_release_fkey
+        foreign key (data_release_code)
+        references occupational_data_release (data_release_code),
+    constraint occupational_source_table_code_check
+        check (btrim(source_table_code) <> ''),
+    constraint occupational_source_artifact_check
+        check (source_artifact_sha256 ~ '^[0-9a-f]{64}$'),
+    constraint occupational_source_row_count_check
+        check (source_row_count > 0)
+);
+
+create table if not exists occupational_scale_definition (
+    data_release_code text not null,
+    source_table_code text not null,
+    scale_id text not null,
+    scale_name text not null,
+    minimum_value numeric not null,
+    maximum_value numeric not null,
+    primary key (data_release_code, scale_id),
+    constraint occupational_scale_release_fkey
+        foreign key (data_release_code)
+        references occupational_data_release (data_release_code),
+    constraint occupational_scale_source_table_fkey
+        foreign key (data_release_code, source_table_code)
+        references occupational_source_table (data_release_code, source_table_code),
+    constraint occupational_scale_id_check check (btrim(scale_id) <> ''),
+    constraint occupational_scale_bounds_check
+        check (minimum_value <= maximum_value)
+);
+
+create table if not exists occupational_classification_entry (
+    data_release_code text not null,
+    onetsoc_code text not null,
+    occupation_title text not null,
+    primary key (data_release_code, onetsoc_code),
+    constraint occupational_classification_release_fkey
+        foreign key (data_release_code)
+        references occupational_data_release (data_release_code),
+    constraint occupational_onetsoc_code_check
+        check (onetsoc_code ~ '^[0-9]{2}-[0-9]{4}\.[0-9]{2}$'),
+    constraint occupational_title_check check (btrim(occupation_title) <> '')
+);
+
+create table if not exists occupational_element_definition (
+    data_release_code text not null,
+    element_id text not null,
+    element_name text not null,
+    primary key (data_release_code, element_id),
+    constraint occupational_element_release_fkey
+        foreign key (data_release_code)
+        references occupational_data_release (data_release_code),
+    constraint occupational_element_id_check
+        check (element_id ~ '^[1-6](\.[A-Za-z0-9]+)*$'),
+    constraint occupational_element_name_check check (btrim(element_name) <> '')
+);
+
+create table if not exists occupational_rating_observation (
+    data_release_code text not null,
+    source_table_code text not null,
+    onetsoc_code text not null,
+    element_id text not null,
+    scale_id text not null,
+    category_value integer,
+    data_value numeric not null,
+    sample_size integer,
+    standard_error numeric,
+    lower_ci_bound numeric,
+    upper_ci_bound numeric,
+    recommend_suppress boolean,
+    not_relevant boolean,
+    source_updated_month text not null,
+    domain_source_code text not null,
+    constraint occupational_rating_source_table_fkey
+        foreign key (data_release_code, source_table_code)
+        references occupational_source_table (data_release_code, source_table_code),
+    constraint occupational_rating_classification_fkey
+        foreign key (data_release_code, onetsoc_code)
+        references occupational_classification_entry (data_release_code, onetsoc_code),
+    constraint occupational_rating_element_fkey
+        foreign key (data_release_code, element_id)
+        references occupational_element_definition (data_release_code, element_id),
+    constraint occupational_rating_scale_fkey
+        foreign key (data_release_code, scale_id)
+        references occupational_scale_definition (data_release_code, scale_id),
+    constraint occupational_rating_identity_key
+        unique nulls not distinct
+        (data_release_code, source_table_code, onetsoc_code, element_id, scale_id, category_value),
+    constraint occupational_rating_sample_size_check
+        check (sample_size is null or sample_size > 0),
+    constraint occupational_rating_standard_error_check
+        check (standard_error is null or standard_error >= 0),
+    constraint occupational_rating_interval_presence_check
+        check ((lower_ci_bound is null) = (upper_ci_bound is null)),
+    constraint occupational_rating_interval_order_check
+        check (lower_ci_bound is null or lower_ci_bound <= upper_ci_bound),
+    constraint occupational_rating_data_value_check
+        check (data_value::text not in ('NaN', 'Infinity', '-Infinity')),
+    constraint occupational_rating_domain_source_check
+        check (btrim(domain_source_code) <> ''),
+    constraint occupational_rating_source_updated_month_check
+        check (source_updated_month ~ '^(0[1-9]|1[0-2])/[0-9]{4}$')
+) partition by list (data_release_code);
+
+create index if not exists occupational_rating_occupation_element_idx
+    on occupational_rating_observation
+    (data_release_code, onetsoc_code, element_id, scale_id);
+
+create index if not exists occupational_rating_element_occupation_idx
+    on occupational_rating_observation
+    (data_release_code, element_id, scale_id, onetsoc_code);
+
+create or replace function validate_occupational_rating_insert()
+returns trigger
+language plpgsql
+as $$
+declare
+    declared_minimum numeric;
+    declared_maximum numeric;
+    existing_observation occupational_rating_observation%rowtype;
+begin
+    if new.source_updated_month !~ '^(0[1-9]|1[0-2])/[0-9]{4}$' then
+        raise check_violation using message = 'source_updated_month must be MM/YYYY';
+    end if;
+    if to_date('01/' || new.source_updated_month, 'DD/MM/YYYY')
+       > date_trunc('month', current_date)::date then
+        raise check_violation using message = 'source_updated_month must not be in the future';
+    end if;
+    select minimum_value, maximum_value
+      into declared_minimum, declared_maximum
+      from occupational_scale_definition
+     where data_release_code = new.data_release_code
+       and scale_id = new.scale_id;
+    if declared_minimum is not null
+       and new.data_value not between declared_minimum and declared_maximum then
+        raise check_violation using message = 'data_value is outside the declared scale bounds';
+    end if;
+    select observation.*
+      into existing_observation
+      from occupational_rating_observation observation
+     where observation.data_release_code = new.data_release_code
+       and observation.source_table_code = new.source_table_code
+       and observation.onetsoc_code = new.onetsoc_code
+       and observation.element_id = new.element_id
+       and observation.scale_id = new.scale_id
+       and observation.category_value is not distinct from new.category_value;
+    if found and row(
+        existing_observation.data_value,
+        existing_observation.sample_size,
+        existing_observation.standard_error,
+        existing_observation.lower_ci_bound,
+        existing_observation.upper_ci_bound,
+        existing_observation.recommend_suppress,
+        existing_observation.not_relevant,
+        existing_observation.source_updated_month,
+        existing_observation.domain_source_code
+    ) is distinct from row(
+        new.data_value,
+        new.sample_size,
+        new.standard_error,
+        new.lower_ci_bound,
+        new.upper_ci_bound,
+        new.recommend_suppress,
+        new.not_relevant,
+        new.source_updated_month,
+        new.domain_source_code
+    ) then
+        raise check_violation using message = 'occupational rating identity conflicts with immutable evidence';
+    end if;
+    return new;
+end;
+$$;
+
+drop trigger if exists occupational_rating_validate_insert
+    on occupational_rating_observation;
+create trigger occupational_rating_validate_insert
+before insert on occupational_rating_observation
+for each row execute function validate_occupational_rating_insert();
+
+create or replace function reject_occupational_rating_mutation()
+returns trigger
+language plpgsql
+as $$
+begin
+    raise check_violation using message = 'occupational rating evidence is immutable';
+end;
+$$;
+
+drop trigger if exists occupational_rating_reject_mutation
+    on occupational_rating_observation;
+create trigger occupational_rating_reject_mutation
+before update or delete on occupational_rating_observation
+for each row execute function reject_occupational_rating_mutation();
+
+drop trigger if exists occupational_rating_reject_truncate
+    on occupational_rating_observation;
+create trigger occupational_rating_reject_truncate
+before truncate on occupational_rating_observation
+for each statement execute function reject_occupational_rating_mutation();
+
+commit;
diff --git a/tests/test_onet_rating_schema.py b/tests/test_onet_rating_schema.py
new file mode 100644
index 000000000..095dbdddf
--- /dev/null
+++ b/tests/test_onet_rating_schema.py
@@ -0,0 +1,34 @@
+"""Static contracts for the normalized O*NET rating observation store."""
+
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parents[1]
+MIGRATION = ROOT / "migrations" / "0222_onet_rating_observation_store.sql"
+
+
+def test_migration_declares_normalized_partitioned_observation_contract() -> None:
+    assert MIGRATION.is_file()
+    sql = MIGRATION.read_text(encoding="utf-8").casefold()
+    for table_name in (
+        "occupational_data_release",
+        "occupational_source_table",
+        "occupational_scale_definition",
+        "occupational_classification_entry",
+        "occupational_element_definition",
+        "occupational_rating_observation",
+    ):
+        assert f"create table if not exists {table_name}" in sql
+    assert "partition by list (data_release_code)" in sql
+    assert "unique nulls not distinct" in sql
+    assert "occupational_scale_source_table_fkey" in sql
+    assert "validate_occupational_rating_insert" in sql
+    assert "reject_occupational_rating_mutation" in sql
+    assert "before truncate on occupational_rating_observation" in sql
+    assert "identity conflicts with immutable evidence" in sql
+    assert "recommend_suppress" in sql
+    assert "not_relevant" in sql
+    assert "standard_error" in sql
+    assert "lower_ci_bound" in sql
+    assert "upper_ci_bound" in sql
+    assert "source_updated_month text not null" in sql
+    assert "source_updated_date" not in sql
diff --git a/tests/test_schema.py b/tests/test_schema.py
index 49caf6f8c..4a689dd76 100644
--- a/tests/test_schema.py
+++ b/tests/test_schema.py
@@ -17,6 +17,7 @@
 import asyncio
 import os
 import uuid
+from decimal import Decimal
 from pathlib import Path
 from urllib.parse import urlsplit, urlunsplit
 
@@ -103,6 +104,11 @@
     / "migrations"
     / "0210_global_ask_evidence_search_indexes.sql"
 )
+_ONET_RATING_MIGRATION = (
+    Path(__file__).resolve().parents[1]
+    / "migrations"
+    / "0222_onet_rating_observation_store.sql"
+)
 _CHANNEL_EVIDENCE_MIGRATION = (
     Path(__file__).resolve().parents[1] / "migrations" / "0174_post_lineage_edge_signal.sql"
 )
@@ -173,6 +179,7 @@ def schema_db():
                 cur.execute(_LEFTOVER_MAP_CROSS_SHARE_MIGRATION.read_text())
                 cur.execute(_LEFTOVER_MAP_RECONSTRUCTION_MIGRATION.read_text())
                 cur.execute(_SOURCE_EVENT_TIME_MIGRATION.read_text())
+                cur.execute(_ONET_RATING_MIGRATION.read_text())
                 # psql sends each statement independently, which is required
                 # by CREATE INDEX CONCURRENTLY. psycopg2 treats a multi-
                 # statement execute as one transaction even with autocommit.
@@ -235,10 +242,242 @@ def test_migration_applies_cleanly(schema_db) -> None:
         "post_summary_action",
         "post_chat_result",
         "post_chat_citation",
+        "occupational_data_release",
+        "occupational_source_table",
+        "occupational_scale_definition",
+        "occupational_classification_entry",
+        "occupational_element_definition",
+        "occupational_rating_observation",
     }
     assert expected <= tables
 
 
+def test_onet_rating_store_partitions_upserts_and_rejects_invalid_error(schema_db) -> None:
+    """Persist one realistic rating without collapsing missing category or uncertainty."""
+    with schema_db.cursor() as cur:
+        cur.execute(
+            """
+            insert into occupational_data_release
+                (data_release_code, release_version, source_publisher_name, source_license_url)
+            values ('onet-31.0', '31.0', 'National Center for O*NET Development',
+                    'https://creativecommons.org/licenses/by/4.0/')
+            """
+        )
+        cur.execute(
+            """
+            insert into occupational_source_table
+                (data_release_code, source_table_code, source_table_name,
+                 source_artifact_url, source_artifact_sha256, source_row_count)
+            values ('onet-31.0', 'abilities', 'Abilities', 'https://example.test/abilities.json',
+                    %s, 94640)
+            """,
+            ("a" * 64,),
+        )
+        cur.execute(
+            """
+            insert into occupational_source_table
+                (data_release_code, source_table_code, source_table_name,
+                 source_artifact_url, source_artifact_sha256, source_row_count)
+            values ('onet-31.0', 'scales_reference', 'Scales Reference',
+                    'https://example.test/scales-reference.json', %s, 33)
+            """,
+            ("b" * 64,),
+        )
+        cur.execute(
+            """
+            insert into occupational_source_table
+                (data_release_code, source_table_code, source_table_name,
+                 source_artifact_url, source_artifact_sha256, source_row_count)
+            values ('onet-31.0', 'work_context', 'Work Context',
+                    'https://example.test/work-context.json', %s, 305389)
+            """,
+            ("c" * 64,),
+        )
+        cur.execute(
+            """
+            insert into occupational_scale_definition
+                (data_release_code, source_table_code, scale_id, scale_name,
+                 minimum_value, maximum_value)
+            values ('onet-31.0', 'scales_reference', 'IM', 'Importance', 1, 5)
+            """
+        )
+        cur.execute(
+            """
+            insert into occupational_scale_definition
+                (data_release_code, source_table_code, scale_id, scale_name,
+                 minimum_value, maximum_value)
+            values ('onet-31.0', 'scales_reference', 'CXP',
+                    'Context (Categories 1-5)', 0, 100)
+            """
+        )
+        cur.execute(
+            """
+            insert into occupational_classification_entry
+                (data_release_code, onetsoc_code, occupation_title)
+            values ('onet-31.0', '15-1252.00', 'Synthetic software occupation')
+            """
+        )
+        cur.execute(
+            """
+            insert into occupational_element_definition
+                (data_release_code, element_id, element_name)
+            values ('onet-31.0', '1.A.1.a.1', 'Oral Comprehension')
+            """
+        )
+        cur.execute(
+            """
+            insert into occupational_element_definition
+                (data_release_code, element_id, element_name)
+            values ('onet-31.0', '4.C.1.a.2.f', 'Telephone Conversations')
+            """
+        )
+        cur.execute("savepoint missing_partition")
+        with pytest.raises(psycopg2.errors.CheckViolation):
+            cur.execute(
+                """
+                insert into occupational_rating_observation
+                    (data_release_code, source_table_code, onetsoc_code, element_id,
+                     scale_id, data_value, source_updated_month, domain_source_code)
+                values ('onet-31.0', 'abilities', '15-1252.00', '1.A.1.a.1',
+                        'IM', 4.10, '08/2026', 'Analyst')
+                """
+            )
+        cur.execute("rollback to savepoint missing_partition")
+        cur.execute(
+            """
+            create table occupational_rating_observation_onet_31
+                partition of occupational_rating_observation
+                for values in ('onet-31.0') partition by list (source_table_code)
+            """
+        )
+        cur.execute(
+            """
+            create table occupational_rating_observation_onet_31_abilities
+                partition of occupational_rating_observation_onet_31
+                for values in ('abilities')
+            """
+        )
+        cur.execute(
+            """
+            create table occupational_rating_observation_onet_31_work_context
+                partition of occupational_rating_observation_onet_31
+                for values in ('work_context')
+            """
+        )
+        cur.execute(
+            """
+            insert into occupational_rating_observation
+                (data_release_code, source_table_code, onetsoc_code, element_id,
+                 scale_id, category_value, data_value, source_updated_month,
+                 domain_source_code)
+            values ('onet-31.0', 'work_context', '15-1252.00', '4.C.1.a.2.f',
+                    'CXP', 5, 63.25, '08/2026', 'Incumbent')
+            """
+        )
+        cur.execute(
+            """
+            select category_value, data_value
+              from occupational_rating_observation
+             where scale_id = 'CXP'
+            """
+        )
+        assert cur.fetchone() == (5, Decimal("63.25"))
+        cur.execute("savepoint cxp_outside_percentage")
+        with pytest.raises(psycopg2.errors.CheckViolation):
+            cur.execute(
+                """
+                insert into occupational_rating_observation
+                    (data_release_code, source_table_code, onetsoc_code, element_id,
+                     scale_id, category_value, data_value, source_updated_month,
+                     domain_source_code)
+                values ('onet-31.0', 'work_context', '15-1252.00', '4.C.1.a.2.f',
+                        'CXP', 4, 100.01, '08/2026', 'Incumbent')
+                """
+            )
+        cur.execute("rollback to savepoint cxp_outside_percentage")
+        statement = """
+            insert into occupational_rating_observation
+                (data_release_code, source_table_code, onetsoc_code, element_id,
+                 scale_id, category_value, data_value, sample_size, standard_error,
+                 lower_ci_bound, upper_ci_bound, recommend_suppress, not_relevant,
+                 source_updated_month, domain_source_code)
+            values ('onet-31.0', 'abilities', '15-1252.00', '1.A.1.a.1',
+                    'IM', null, %s, 120, 0.08, 3.94, 4.26, false, false,
+                    '08/2026', 'Analyst')
+            on conflict on constraint occupational_rating_identity_key
+            do nothing
+        """
+        cur.execute(statement, (4.10,))
+        cur.execute("savepoint divergent_duplicate")
+        with pytest.raises(psycopg2.errors.CheckViolation):
+            cur.execute(statement, (4.25,))
+        cur.execute("rollback to savepoint divergent_duplicate")
+        cur.execute(statement, (4.10,))
+        cur.execute(
+            """
+            select count(*), max(data_value), max(source_updated_month)
+            from occupational_rating_observation
+            where data_release_code = 'onet-31.0' and scale_id = 'IM'
+            """
+        )
+        assert cur.fetchone() == (1, Decimal("4.10"), "08/2026")
+        cur.execute("savepoint immutable_update")
+        with pytest.raises(psycopg2.errors.CheckViolation):
+            cur.execute(
+                """
+                update occupational_rating_observation
+                set data_value = 4.25
+                where data_release_code = 'onet-31.0'
+                """
+            )
+        cur.execute("rollback to savepoint immutable_update")
+        cur.execute("savepoint immutable_delete")
+        with pytest.raises(psycopg2.errors.CheckViolation):
+            cur.execute(
+                "delete from occupational_rating_observation where data_release_code = 'onet-31.0'"
+            )
+        cur.execute("rollback to savepoint immutable_delete")
+        cur.execute("savepoint immutable_truncate")
+        with pytest.raises(psycopg2.errors.CheckViolation):
+            cur.execute("truncate occupational_rating_observation")
+        cur.execute("rollback to savepoint immutable_truncate")
+        for savepoint, value_sql, month_sql in (
+            ("outside_scale", "6.00", "'08/2026'"),
+            (
+                "future_source_month",
+                "4.00",
+                "to_char(current_date + interval '1 month', 'MM/YYYY')",
+            ),
+            ("malformed_source_month", "4.00", "'13/2026'"),
+        ):
+            cur.execute(f"savepoint {savepoint}")
+            with pytest.raises(psycopg2.errors.CheckViolation):
+                cur.execute(
+                    f"""
+                    insert into occupational_rating_observation
+                        (data_release_code, source_table_code, onetsoc_code, element_id,
+                         scale_id, category_value, data_value, source_updated_month,
+                         domain_source_code)
+                    values ('onet-31.0', 'abilities', '15-1252.00', '1.A.1.a.1',
+                            'IM', 1, {value_sql}, {month_sql}, 'Analyst')
+                    """
+                )
+            cur.execute(f"rollback to savepoint {savepoint}")
+        cur.execute("savepoint invalid_standard_error")
+        with pytest.raises(psycopg2.errors.CheckViolation):
+            cur.execute(
+                """
+                insert into occupational_rating_observation
+                    (data_release_code, source_table_code, onetsoc_code, element_id,
+                     scale_id, category_value, data_value, standard_error,
+                     source_updated_month, domain_source_code)
+                values ('onet-31.0', 'abilities', '15-1252.00', '1.A.1.a.1',
+                        'IM', 2, 4.00, -0.01, '08/2026', 'Analyst')
+                """
+            )
+        cur.execute("rollback to savepoint invalid_standard_error")
+
+
 def test_global_ask_evidence_search_indexes_exist_on_normalized_tables(schema_db) -> None:
     """The real PostgreSQL schema owns all nine evidence-search indexes."""
     with schema_db.cursor() as cur:

From e9cdf42730ce07bd39370a4144d3c28c7d433498 Mon Sep 17 00:00:00 2001
From: Seongho Bae <me@seonghobae.me>
Date: Thu, 27 Aug 2026 01:30:12 -0700
Subject: [PATCH 17/17] feat(ontology): catalog imported rating sources (#743)

* feat(ontology): catalog imported rating sources

* fix(ui): validate occupation evidence links

* fix(ui): reset rating catalog across authentication

* test(ui): isolate occupation response fencing

* fix(ui): clear occupation evidence across auth changes

* feat(ontology): select represented occupations (#746)

* feat(ontology): select represented occupations

* fix(ui): invalidate rating requests on source change

* fix(ui): clear evidence when occupation changes

* fix(ui): distinguish unavailable occupation catalogs

* fix(ui): align empty evidence guidance with selector

* fix(ui): distinguish unavailable occupation catalogs

---------

Co-authored-by: Codex <codex@localhost>

---------

Co-authored-by: Codex <codex@localhost>
---
 ARCHITECTURE.md                               |   4 +-
 CHANGELOG.md                                  |   7 +
 backend/app/main.py                           |  36 ++-
 backend/app/occupation_rating_ingestion.py    |  71 ++++++
 .../0260-occupation-rating-source-catalog.md  |  41 ++++
 .../0261-rating-source-occupation-selector.md |  45 ++++
 docs/adr/README.md                            |   2 +
 docs/product-requirements.md                  |  28 +++
 docs/product-technical-gap-baseline.md        |   2 +-
 docs/storybook-inventory.md                   |   2 +-
 frontend/src/App.css                          |   5 +-
 frontend/src/api.test.ts                      |  30 +++
 frontend/src/api.ts                           |  40 ++++
 .../OccupationRatingProfile.stories.tsx       |  72 +++++-
 .../OccupationRatingProfile.test.tsx          | 206 +++++++++++++++---
 .../components/OccupationRatingProfile.tsx    | 153 ++++++++++---
 tests/test_occupation_rating_ingestion.py     | 101 ++++++++-
 tests/test_schema.py                          |  23 +-
 18 files changed, 795 insertions(+), 73 deletions(-)
 create mode 100644 docs/adr/0260-occupation-rating-source-catalog.md
 create mode 100644 docs/adr/0261-rating-source-occupation-selector.md

diff --git a/ARCHITECTURE.md b/ARCHITECTURE.md
index 1a112853d..dcfab0b7d 100644
--- a/ARCHITECTURE.md
+++ b/ARCHITECTURE.md
@@ -80,8 +80,8 @@ flowchart LR
 | `temporal_expressions.py` | Pure Korean relative-time resolver for Global Ask (ADR 0150) |
 | `ask_time_axis.py` | Event-time vs ingestion-time clock choice for that window (ADR 0202) |
 | `ontology.py` | Loads the governed Turtle source tree (`lineageweave-kg.ttl` plus generated fragments), the formal OWL 2/RDFS/SKOS vocabulary for the Knowledge Graph's node/edge types, source taxonomies, and published O*NET linkages (ADR 0004, ADR 0252, ADR 0255, ADR 0256) |
-| `backend/app/occupation_rating_ingestion.py` | Projects authenticated, bounded occupation-rating source evidence with exact decimal and artifact provenance semantics (ADR 0258) |
-| `frontend/src/components/OccupationRatingProfile.tsx` | Reads exact occupation evidence in the existing Dashboard and preserves absence, uncertainty, and warning semantics (ADR 0259) |
+| `backend/app/occupation_rating_ingestion.py` | Projects authenticated occupation-rating evidence plus persisted source and represented-occupation catalogs (ADR 0258, ADR 0260, ADR 0261) |
+| `frontend/src/components/OccupationRatingProfile.tsx` | Selects imported source and stored occupation title before reading exact Dashboard evidence, preserving absence, uncertainty, and warning semantics (ADR 0259–0261) |
 | `ontology_neighborhood.py` | Bounded typed ontology/provenance neighborhood (ADR 0184); PostgreSQL stays authoritative, OWL subclass is not an instance edge |
 | `ontology_source_cursor.py` | Opaque HMAC source-window continuation (ADR 0124); keyset pagination, never OFFSET |
 | `period_report.py` | Fit GRM/GPCM on persisted IRT rows, FIPC-select, EAP-score a period (ADR 0003 slice 3; Bock & Mislevy, 1982) |
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 01055a341..a3d970686 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -8,6 +8,13 @@ All notable changes to this project are documented here. Format follows
 
 ### Added
 
+- Each imported rating source now exposes its exact represented O*NET-SOC
+  code/title catalog, and the Dashboard uses that catalog instead of requiring
+  users to know or type an occupation code (ADR 0261).
+- Occupation evidence source selection now comes from an authenticated catalog
+  of actually imported rating artifacts, with release, publisher, license,
+  digest, URL, and row-count provenance and fail-closed loading/empty/error
+  states (ADR 0260).
 - The existing Dashboard now includes an authenticated occupation-evidence
   view with exact source selection, sample/error context, textual suppression
   warnings, artifact links, responsive table access, and distinct unavailable
diff --git a/backend/app/main.py b/backend/app/main.py
index 60dd9e1a5..b56a98ed6 100644
--- a/backend/app/main.py
+++ b/backend/app/main.py
@@ -98,7 +98,11 @@
     upsert_commitment_ticket,
 )
 from backend.app.operations_dashboard import fetch_operations_dashboard
-from backend.app.occupation_rating_ingestion import fetch_occupation_ratings
+from backend.app.occupation_rating_ingestion import (
+    fetch_occupation_rating_sources,
+    fetch_occupation_ratings,
+    fetch_rating_source_occupations,
+)
 from backend.app.keyman_ingestion import ingest_post_keymen
 from backend.app.knowledge_graph import (
     corporate_entity_exists,
@@ -2302,6 +2306,36 @@ async def read_occupation_ratings(
         )
 
 
+@app.get("/api/occupation-rating-sources")
+async def read_occupation_rating_sources(
+    _account: CurrentAccount = Depends(get_current_account),
+    pool: asyncpg.Pool = Depends(get_pool),
+) -> dict[str, list[dict[str, object]]]:
+    """Return the authenticated catalog of imported occupation-rating sources."""
+    async with pool.acquire() as conn:
+        return await fetch_occupation_rating_sources(conn)
+
+
+@app.get("/api/occupation-rating-occupations")
+async def read_rating_source_occupations(
+    data_release_code: str = Query(
+        ..., min_length=1, max_length=63, pattern=r"^[a-z0-9][a-z0-9.-]*$"
+    ),
+    source_table_code: str = Query(
+        ..., min_length=1, max_length=63, pattern=r"^[a-z][a-z0-9_]*$"
+    ),
+    _account: CurrentAccount = Depends(get_current_account),
+    pool: asyncpg.Pool = Depends(get_pool),
+) -> dict[str, object]:
+    """Return occupations represented in one imported rating source."""
+    async with pool.acquire() as conn:
+        return await fetch_rating_source_occupations(
+            conn,
+            data_release_code=data_release_code,
+            source_table_code=source_table_code,
+        )
+
+
 @app.get("/api/posts/{post_id}/counterparties")
 async def read_post_counterparties(
     post_id: str,
diff --git a/backend/app/occupation_rating_ingestion.py b/backend/app/occupation_rating_ingestion.py
index 320fe248d..f7d2d10a1 100644
--- a/backend/app/occupation_rating_ingestion.py
+++ b/backend/app/occupation_rating_ingestion.py
@@ -125,3 +125,74 @@ async def fetch_occupation_ratings(
         "items": items,
         "next_offset": offset + limit if len(rows) > limit else None,
     }
+
+
+async def fetch_occupation_rating_sources(
+    conn: RatingReadConnection,
+) -> dict[str, list[dict[str, object]]]:
+    """Return imported rating artifacts that contain at least one observation."""
+    rows = await conn.fetch(
+        """select source.data_release_code, release.release_version,
+                  release.source_publisher_name, release.source_license_url,
+                  source.source_table_code, source.source_table_name,
+                  source.source_artifact_url, source.source_artifact_sha256,
+                  source.source_row_count
+             from occupational_source_table source
+             join occupational_data_release release
+               on release.data_release_code = source.data_release_code
+            where source.source_table_code <> 'scales_reference'
+              and exists (
+                    select 1
+                      from occupational_rating_observation observation
+                     where observation.data_release_code = source.data_release_code
+                       and observation.source_table_code = source.source_table_code
+              )
+            order by release.imported_at desc, source.data_release_code,
+                     source.source_table_name, source.source_table_code"""
+    )
+    return {"sources": [dict(row) for row in rows]}
+
+
+async def fetch_rating_source_occupations(
+    conn: RatingReadConnection,
+    *,
+    data_release_code: str,
+    source_table_code: str,
+) -> dict[str, object]:
+    """Return occupations with observations in one exact imported source."""
+    source = await conn.fetchrow(
+        """select 1
+             from occupational_source_table
+            where data_release_code = $1 and source_table_code = $2""",
+        data_release_code,
+        source_table_code,
+    )
+    if source is None:
+        return {
+            "data_release_code": data_release_code,
+            "source_table_code": source_table_code,
+            "source_available": False,
+            "occupations": [],
+        }
+    rows = await conn.fetch(
+        """select classification.onetsoc_code, classification.occupation_title
+             from occupational_classification_entry classification
+            where classification.data_release_code = $1
+              and exists (
+                    select 1
+                      from occupational_rating_observation observation
+                     where observation.data_release_code = classification.data_release_code
+                       and observation.source_table_code = $2
+                       and observation.onetsoc_code = classification.onetsoc_code
+              )
+            order by classification.occupation_title,
+                     classification.onetsoc_code""",
+        data_release_code,
+        source_table_code,
+    )
+    return {
+        "data_release_code": data_release_code,
+        "source_table_code": source_table_code,
+        "source_available": True,
+        "occupations": [dict(row) for row in rows],
+    }
diff --git a/docs/adr/0260-occupation-rating-source-catalog.md b/docs/adr/0260-occupation-rating-source-catalog.md
new file mode 100644
index 000000000..93f80ce54
--- /dev/null
+++ b/docs/adr/0260-occupation-rating-source-catalog.md
@@ -0,0 +1,41 @@
+# ADR 0260: Imported occupation-rating source catalog
+
+- Status: Accepted
+- Date: 2026-08-27
+- Extends: ADR 0257, ADR 0258, ADR 0259
+
+## Context
+
+ADR 0259 initially requires users to type internal release and source-table
+codes. That makes a valid product action depend on repository knowledge and
+allows a user to request a source that was never imported. The normalized
+rating store already owns the exact imported artifact catalog and therefore is
+the only authoritative selector source.
+
+## Decision
+
+1. Add an authenticated read endpoint that lists rating artifacts containing
+   at least one persisted occupation observation. Exclude the Scales Reference
+   support artifact from selectable rating sources.
+2. Return release code/version, publisher and license, source code/name, URL,
+   SHA-256, and declared row count. Order releases by persisted import time and
+   sources by stored name/code; do not infer recency from a version string.
+3. The Dashboard selects only an entry returned by this endpoint. If the
+   catalog is loading, empty, or unavailable, disable profile submission and
+   give the user a next action. Do not retain a hidden hand-written fallback.
+4. Authentication matches ADR 0258: imported O*NET artifacts are public
+   reference data, while the catalog still requires a valid workspace account.
+5. The catalog does not claim that all official O*NET artifacts are imported.
+   It describes only current database state with immutable artifact provenance.
+
+## Consequences
+
+The occupation evidence workflow no longer asks users to know storage codes,
+and an unavailable artifact cannot masquerade as a selectable source. Adding
+an official artifact remains an importer operation with digest and row-count
+validation rather than a UI-created catalog row.
+
+## References
+
+National Center for O*NET Development. (2026). *O*NET 31.0 database* [Data
+set]. https://www.onetcenter.org/database.html
diff --git a/docs/adr/0261-rating-source-occupation-selector.md b/docs/adr/0261-rating-source-occupation-selector.md
new file mode 100644
index 000000000..8a3c73ddf
--- /dev/null
+++ b/docs/adr/0261-rating-source-occupation-selector.md
@@ -0,0 +1,45 @@
+# ADR 0261: Occupations represented by an imported rating source
+
+- Status: Accepted
+- Date: 2026-08-27
+- Extends: ADR 0257, ADR 0258, ADR 0260
+
+## Context
+
+The source catalog removes internal release/source entry, but ADR 0260 still
+leaves users to type an O*NET-SOC code. The normalized store already preserves
+the source occupation title and code. A release may contain occupations that
+are absent from one rating artifact, so the release classification alone is
+not sufficient evidence that a profile exists for the selected source.
+
+## Decision
+
+1. Add an authenticated read endpoint returning stored O*NET-SOC code/title
+   pairs that have at least one observation in one exact imported rating
+   source. Keep unavailable source distinct from an available empty source.
+2. Join by normalized release/code identity and an observation-existence
+   predicate. Do not bind occupations by title similarity, keyword inference,
+   external search, or a locally reconstructed classification.
+3. Order by the stored occupation title and then code. Return the complete
+   represented set because the official imported classification is the
+   authoritative finite selector domain; do not introduce an arbitrary result
+   cutoff that makes valid occupations disappear.
+4. Replace free-text occupation-code entry with a native select whose visible
+   label begins with the stored title and retains the exact code. Changing the
+   rating source clears both occupation selection and displayed evidence;
+   changing the occupation clears displayed evidence.
+5. While the occupation catalog is loading, empty, or unavailable, disable
+   profile submission and state the next action. Pagination remains bound to
+   the identifiers returned by the loaded profile under ADR 0259.
+
+## Consequences
+
+Users choose an occupation by its authoritative title without knowing an
+internal code, while API requests continue to carry exact stable identifiers.
+Employer job families and series remain outside this selector until their
+separate authorized import contract exists.
+
+## References
+
+National Center for O*NET Development. (2026). *O*NET 31.0 database* [Data
+set]. https://www.onetcenter.org/database.html
diff --git a/docs/adr/README.md b/docs/adr/README.md
index 27911acd1..3ff700e68 100644
--- a/docs/adr/README.md
+++ b/docs/adr/README.md
@@ -36,6 +36,8 @@ decision from them.
 | [`ONET_RATING_STORE_REFERENCES.md`](../doctoring/ONET_RATING_STORE_REFERENCES.md) | [0257](0257-onet-occupation-rating-observation-store.md) |
 | Occupation-rating authenticated read projection | [0258](0258-occupation-rating-read-api.md) |
 | Occupation-rating Dashboard evidence view | [0259](0259-occupation-rating-evidence-ui.md) |
+| Imported occupation-rating source catalog | [0260](0260-occupation-rating-source-catalog.md) |
+| Rating-source occupation selector | [0261](0261-rating-source-occupation-selector.md) |
 
 [0011](0011-prov-o-standard-relations.md) and [0065](0065-prov-o-provenance-boundary.md) cite the dated W3C PROV-O and PROV-DM Recommendations (https://www.w3.org/TR/2013/REC-prov-o-20130430/ and https://www.w3.org/TR/2013/REC-prov-dm-20130430/).
 
diff --git a/docs/product-requirements.md b/docs/product-requirements.md
index 181f7c755..2ed4d772f 100644
--- a/docs/product-requirements.md
+++ b/docs/product-requirements.md
@@ -175,6 +175,34 @@ scrollable table; narrow layouts retain complete values; suppression remains
 visible beside its value; and Storybook covers populated, narrow, unavailable,
 and empty states using synthetic data.
 
+### PRD-FR-2G — Imported rating-source catalog
+
+- Populate the occupation evidence selector only from imported artifacts that
+  contain observations, preserving release and artifact provenance (ADR 0260).
+- Exclude the scale-definition support artifact from the rating-source selector.
+- Disable profile submission and state the next action while the catalog is
+  loading, empty, or unavailable.
+
+Acceptance: a user never types an internal release/source code; the selector
+order follows persisted import time rather than parsed version heuristics; and
+the real PostgreSQL integration test proves an imported synthetic artifact is
+listed while its supporting scale artifact is not.
+
+### PRD-FR-2H — Occupations represented in a rating source
+
+- Populate the occupation selector with exact stored code/title pairs that
+  have observations in the selected imported source (ADR 0261).
+- Clear the current occupation and profile when the source changes, and clear
+  the profile when the occupation changes; never mix continuation rows across
+  occupations or sources.
+- Keep unavailable source, available-empty source, loading, and transport
+  failure distinct and actionable.
+
+Acceptance: a user selects a stored title rather than typing an internal code;
+the PostgreSQL integration test proves the source membership predicate; and
+component tests prove selector changes clear prior evidence and pagination
+stays bound to the loaded profile identifiers.
+
 ### PRD-FR-3 — Bounded ontology exploration
 
 - Apply RBAC/ABAC, source eligibility, and knowledge cutoff before graph
diff --git a/docs/product-technical-gap-baseline.md b/docs/product-technical-gap-baseline.md
index 1632ad818..63b15f75d 100644
--- a/docs/product-technical-gap-baseline.md
+++ b/docs/product-technical-gap-baseline.md
@@ -29,7 +29,7 @@ explicit unavailable state, not a reason to infer mappings from labels.
 | Occupation-to-construct relations | ADR 0257 defines a candidate 3NF, release/source-partitioned immutable observation store and deterministic pinned-CSV importer preserving value, optional category, sample/error/CI, suppression, relevance, exact `MM/YYYY` source update month, source digest, and domain provenance. The official O*NET 31.0 Abilities file (94,640 rows, 910 occupations, 52 elements; SHA-256 `7e9cd79791ce6014e1d26d0a449ae5b1e7aa7ef52d39b3934c3bb8d438104b88`) and all 33 Scales Reference rows (SHA-256 `bcba23858ce21ecaacbde303a8993e35d46724b4afb8c9ec2b10e04f42adcfc9`) imported into a throwaway local PostgreSQL database with all 94,640 observations, 55 suppression flags, 7,572 not-relevant flags, and source months from `12/2004` through `08/2026`; every scale retained `scales_reference` artifact provenance, the database was dropped afterward, and no corpus is committed or claimed deployed | Pass exact-head review/checks and protected merge; validate and import every selected official rating artifact through an authorized runtime, returning only aggregate evidence; never invent or locally normalize a weight |
 | Job-family and job-series semantics | No authoritative employer-specific job architecture is present | Define an organization-neutral import contract that preserves the authorized source hierarchy and distinguishes standard occupation codes from employer job families/series; no label-based binding |
 | Temporal and multilevel interpretation | Static vocabulary only; no person-level inference is asserted | Version valid and transaction time, preserve occupation/organization/unit nesting and multiple membership, and require TEPP or the owning Rust psychometric service before any calibrated temporal or multilevel result |
-| Product consumption | ADR 0258 defines a candidate authenticated occupation-rating API; ADR 0259 adds a candidate Dashboard evidence view with exact source selection, sample/error context, text warnings, artifact links, responsive overflow, and distinct unavailable/empty states. Component/API tests and Storybook build pass; synthetic populated scenes were visually audited at 1440×900 and 390×844, including the mobile horizontal-scroll next action. Protected delivery and authenticated runtime evidence remain absent | Pass exact-head review/checks and protected merge; verify the authenticated API and rendered Dashboard against an authorized imported source using only aggregate/non-identifying evidence, and add a source catalog selector after an authoritative catalog read contract exists |
+| Product consumption | ADR 0258 defines a candidate authenticated occupation-rating API; ADR 0259 adds a candidate Dashboard evidence view; ADR 0260 replaces release/source code entry with persisted artifact selection; ADR 0261 replaces occupation-code entry with stored titles represented in that source. Component/API/PostgreSQL tests and Storybook scenes cover value/error/warning, absence, source/occupation selection, stale-response, pagination, and safe-link contracts; populated synthetic scenes were visually audited at 1440×900 and 390×844. Protected delivery and authenticated runtime evidence remain absent | Pass exact-head review/checks and protected merge; verify the authenticated catalogs, profile API, and rendered Dashboard against an authorized imported source using only aggregate/non-identifying evidence |
 
 ### Current exact-head PR queue
 
diff --git a/docs/storybook-inventory.md b/docs/storybook-inventory.md
index c9e1570bf..46aa4d467 100644
--- a/docs/storybook-inventory.md
+++ b/docs/storybook-inventory.md
@@ -6,7 +6,7 @@ operator-facing control you can click before changing product CSS.
 | Story | Operator next action | Token / module |
 |---|---|---|
 | `Workspace/OperationsDashboard` | Compare Event and post counts, inspect external-information coverage, then open the cited source behind a claim, handover, or repeat-issue fact. `EvidenceReady`, `NarrowViewport`, `AnalysisPendingAndMissingEvidence`, `AnalysisFailed`, and `LoadError` cover populated, mobile, unavailable-evidence, analysis-pending, retryable failure, and transport-error states. | `--color-dashboard-*`, `OperationsDashboard` |
-| `Ontology/OccupationRatingProfile` | Enter an exact O*NET-SOC code, inspect the published value beside its sample/error and warning, then open the rating or scale artifact. `InteractiveEvidenceReady`, `EvidenceReady`, `NarrowViewport`, `SourceUnavailable`, and `EmptyOccupation` cover the full form, populated table, horizontal mobile access, and honest absence states. | `OccupationRatingProfile`, native table, `--color-border`, `--size-control-min` |
+| `Ontology/OccupationRatingProfile` | Select an imported release/source and stored occupation title, inspect the published value beside its sample/error and warning, then open the rating or scale artifact. `InteractiveEvidenceReady`, `EvidenceReady`, `NarrowViewport`, `CatalogEmpty`, `CatalogUnavailable`, `OccupationsEmpty`, `SourceUnavailable`, and `EmptyOccupation` cover both selectors, populated table, horizontal mobile access, and honest catalog/profile absence states. | `OccupationRatingProfile`, native select/table, `--color-border`, `--size-control-min` |
 | `Post/SimilarVocPanel` | Compare ontology/semantic similar VOC and prior action evidence, then open the source; unavailable states show no fabricated TEPP theta or weight. | `SimilarVocPanel.css`, `SimilarVocPanel` |
 | `Evidence/CitationChip` | Click a cited title to open that source post. | `--color-chip-border`, `--radius-chip`, `CitationChip` |
 | `Evidence/OrganizationAliasChip` | Click a cataloged org; the parenthetical is the unique corroborated SKOS companion. | `--color-chip-border`, `--radius-chip`, `OrganizationAliasChip` |
diff --git a/frontend/src/App.css b/frontend/src/App.css
index 33adf4221..692bdf10e 100644
--- a/frontend/src/App.css
+++ b/frontend/src/App.css
@@ -1448,7 +1448,7 @@
 
 .occupation-rating-form {
   display: grid;
-  grid-template-columns: repeat(3, minmax(10rem, 1fr)) auto;
+  grid-template-columns: minmax(14rem, 1fr) minmax(16rem, 1fr) auto;
   align-items: end;
   gap: var(--space-control-gap);
   margin: 1rem 0;
@@ -1460,7 +1460,8 @@
   gap: var(--space-control-gap);
 }
 
-.occupation-rating-form input { min-height: var(--size-control-min); }
+.occupation-rating-form input,
+.occupation-rating-form select { min-height: var(--size-control-min); }
 .occupation-rating-source { grid-template-columns: repeat(auto-fit, minmax(12rem, 1fr)); margin: 1rem 0; }
 .occupation-rating-scroll-hint { display: none; }
 .occupation-rating-table { overflow-x: auto; border: 1px solid var(--color-border); }
diff --git a/frontend/src/api.test.ts b/frontend/src/api.test.ts
index cf0b30d40..9495dc241 100644
--- a/frontend/src/api.test.ts
+++ b/frontend/src/api.test.ts
@@ -2,8 +2,10 @@ import { afterEach, describe, expect, it, vi } from "vitest";
 import {
   BackendError,
   fetchMe,
+  fetchOccupationRatingSources,
   fetchOccupationRatings,
   fetchOperationsDashboard,
+  fetchRatingSourceOccupations,
   updateTenantConfig,
 } from "./api";
 
@@ -44,6 +46,34 @@ describe("backendFetch provider-error boundary", () => {
     );
   });
 
+  it("reads the authenticated occupation source catalog", async () => {
+    const fetchMock = vi.fn().mockResolvedValue(
+      new Response(JSON.stringify({ sources: [] }), {
+        headers: { "Content-Type": "application/json" },
+      }),
+    );
+    vi.stubGlobal("fetch", fetchMock);
+
+    await fetchOccupationRatingSources("access-token");
+
+    expect(fetchMock.mock.calls[0][0]).toContain("/api/occupation-rating-sources");
+  });
+
+  it("reads occupations for one exact imported source", async () => {
+    const fetchMock = vi.fn().mockResolvedValue(
+      new Response(JSON.stringify({ occupations: [] }), {
+        headers: { "Content-Type": "application/json" },
+      }),
+    );
+    vi.stubGlobal("fetch", fetchMock);
+
+    await fetchRatingSourceOccupations("access-token", "onet-31.0", "abilities");
+
+    expect(fetchMock.mock.calls[0][0]).toContain(
+      "/api/occupation-rating-occupations?data_release_code=onet-31.0&source_table_code=abilities",
+    );
+  });
+
   it("does not expose provider details from server failures", async () => {
     vi.stubGlobal(
       "fetch",
diff --git a/frontend/src/api.ts b/frontend/src/api.ts
index 1e904bbc5..6972a6542 100644
--- a/frontend/src/api.ts
+++ b/frontend/src/api.ts
@@ -997,6 +997,46 @@ export interface OccupationRatingProfile {
   next_offset: number | null;
 }
 
+export interface OccupationRatingSource {
+  data_release_code: string;
+  release_version: string;
+  source_publisher_name: string;
+  source_license_url: string;
+  source_table_code: string;
+  source_table_name: string;
+  source_artifact_url: string;
+  source_artifact_sha256: string;
+  source_row_count: number;
+}
+
+export function fetchOccupationRatingSources(
+  accessToken: string,
+): Promise<{ sources: OccupationRatingSource[] }> {
+  return backendFetch("/api/occupation-rating-sources", accessToken);
+}
+
+export interface RatingSourceOccupation {
+  onetsoc_code: string;
+  occupation_title: string;
+}
+
+export function fetchRatingSourceOccupations(
+  accessToken: string,
+  dataReleaseCode: string,
+  sourceTableCode: string,
+): Promise<{
+  data_release_code: string;
+  source_table_code: string;
+  source_available: boolean;
+  occupations: RatingSourceOccupation[];
+}> {
+  const params = new URLSearchParams({
+    data_release_code: dataReleaseCode,
+    source_table_code: sourceTableCode,
+  });
+  return backendFetch(`/api/occupation-rating-occupations?${params.toString()}`, accessToken);
+}
+
 export function fetchOccupationRatings(
   accessToken: string,
   query: {
diff --git a/frontend/src/components/OccupationRatingProfile.stories.tsx b/frontend/src/components/OccupationRatingProfile.stories.tsx
index d12114844..aba75d34b 100644
--- a/frontend/src/components/OccupationRatingProfile.stories.tsx
+++ b/frontend/src/components/OccupationRatingProfile.stories.tsx
@@ -29,14 +29,33 @@ export const InteractiveEvidenceReady: Story = {
   render: () => <OccupationRatingProfile accessToken="synthetic-token" />,
   beforeEach: () => {
     const previousFetch = globalThis.fetch;
-    globalThis.fetch = async () => new Response(JSON.stringify(ready), {
-      headers: { "Content-Type": "application/json" },
-    });
+    globalThis.fetch = async (input) => new Response(JSON.stringify(
+      String(input).includes("occupation-rating-sources")
+        ? { sources: [{
+            data_release_code: "onet-31.0", release_version: "31.0",
+            source_publisher_name: "Synthetic publisher", source_license_url: "https://example.test/license",
+            source_table_code: "abilities", source_table_name: "Abilities",
+            source_artifact_url: "https://example.test/abilities.csv", source_artifact_sha256: "a".repeat(64),
+            source_row_count: 94640,
+          }] }
+        : String(input).includes("occupation-rating-occupations")
+          ? {
+              data_release_code: "onet-31.0", source_table_code: "abilities",
+              source_available: true,
+              occupations: [
+                { onetsoc_code: "11-1011.00", occupation_title: "Chief Executives" },
+                { onetsoc_code: "15-1252.00", occupation_title: "Software Developers" },
+              ],
+            }
+          : ready,
+    ), { headers: { "Content-Type": "application/json" } });
     return () => { globalThis.fetch = previousFetch; };
   },
   play: async ({ canvasElement }) => {
     const canvas = within(canvasElement);
-    await userEvent.type(canvas.getByLabelText("O*NET-SOC 직업 코드"), "15-1252.00");
+    const occupation = await canvas.findByLabelText("직업");
+    await canvas.findByRole("option", { name: "Software Developers · 15-1252.00" });
+    await userEvent.selectOptions(occupation, "15-1252.00");
     await userEvent.click(canvas.getByRole("button", { name: "직업 근거 열기" }));
     await expect(canvas.findByText("4.10")).resolves.toBeVisible();
   },
@@ -46,5 +65,50 @@ export const NarrowViewport: Story = {
   ...InteractiveEvidenceReady,
   parameters: { viewport: { defaultViewport: "mobile1" } },
 };
+export const CatalogEmpty: Story = {
+  render: () => <OccupationRatingProfile accessToken="synthetic-token" />,
+  beforeEach: () => {
+    const previousFetch = globalThis.fetch;
+    globalThis.fetch = async () => new Response(JSON.stringify({ sources: [] }), {
+      headers: { "Content-Type": "application/json" },
+    });
+    return () => { globalThis.fetch = previousFetch; };
+  },
+  play: async ({ canvasElement }) => {
+    await expect(within(canvasElement).findByText(/가져온 직업 근거 표가 없습니다/)).resolves.toBeVisible();
+  },
+};
+export const CatalogUnavailable: Story = {
+  render: () => <OccupationRatingProfile accessToken="synthetic-token" />,
+  beforeEach: () => {
+    const previousFetch = globalThis.fetch;
+    globalThis.fetch = async () => { throw new Error("synthetic catalog failure"); };
+    return () => { globalThis.fetch = previousFetch; };
+  },
+  play: async ({ canvasElement }) => {
+    await expect(within(canvasElement).findByRole("alert")).resolves.toHaveTextContent("잠시 후 다시 열어 보세요");
+  },
+};
+export const OccupationsEmpty: Story = {
+  render: () => <OccupationRatingProfile accessToken="synthetic-token" />,
+  beforeEach: () => {
+    const previousFetch = globalThis.fetch;
+    globalThis.fetch = async (input) => new Response(JSON.stringify(
+      String(input).includes("occupation-rating-sources")
+        ? { sources: [{
+            data_release_code: "onet-31.0", release_version: "31.0",
+            source_publisher_name: "Synthetic publisher", source_license_url: "https://example.test/license",
+            source_table_code: "abilities", source_table_name: "Abilities",
+            source_artifact_url: "https://example.test/abilities.csv", source_artifact_sha256: "a".repeat(64),
+            source_row_count: 2,
+          }] }
+        : { data_release_code: "onet-31.0", source_table_code: "abilities", source_available: true, occupations: [] },
+    ), { headers: { "Content-Type": "application/json" } });
+    return () => { globalThis.fetch = previousFetch; };
+  },
+  play: async ({ canvasElement }) => {
+    await expect(within(canvasElement).findByText(/선택할 수 있는 직업이 없습니다/)).resolves.toBeVisible();
+  },
+};
 export const SourceUnavailable: Story = { args: { profile: { ...ready, source_available: false, source: null, items: [] } } };
 export const EmptyOccupation: Story = { args: { profile: { ...ready, items: [] } } };
diff --git a/frontend/src/components/OccupationRatingProfile.test.tsx b/frontend/src/components/OccupationRatingProfile.test.tsx
index 11334c948..7c3f881c2 100644
--- a/frontend/src/components/OccupationRatingProfile.test.tsx
+++ b/frontend/src/components/OccupationRatingProfile.test.tsx
@@ -1,12 +1,19 @@
 import { fireEvent, render, screen } from "@testing-library/react";
 import userEvent from "@testing-library/user-event";
-import { describe, expect, it, vi } from "vitest";
-import { fetchOccupationRatings, type OccupationRatingProfile as Payload } from "../api";
+import { beforeEach, describe, expect, it, vi } from "vitest";
+import {
+  fetchOccupationRatingSources,
+  fetchOccupationRatings,
+  fetchRatingSourceOccupations,
+  type OccupationRatingProfile as Payload,
+} from "../api";
 import { OccupationRatingProfile, OccupationRatingProfileView } from "./OccupationRatingProfile";
 
 vi.mock("../api", async (importOriginal) => ({
   ...(await importOriginal<typeof import("../api")>()),
+  fetchOccupationRatingSources: vi.fn(),
   fetchOccupationRatings: vi.fn(),
+  fetchRatingSourceOccupations: vi.fn(),
 }));
 
 const ready: Payload = {
@@ -33,11 +40,46 @@ const ready: Payload = {
   next_offset: null,
 };
 
+beforeEach(() => {
+  vi.mocked(fetchOccupationRatingSources).mockResolvedValue({
+    sources: [{
+      data_release_code: "onet-31.0", release_version: "31.0",
+      source_publisher_name: "Synthetic publisher", source_license_url: "https://example.test/license",
+      source_table_code: "abilities", source_table_name: "Abilities",
+      source_artifact_url: "https://example.test/abilities.csv", source_artifact_sha256: "a".repeat(64),
+      source_row_count: 2,
+    }],
+  });
+  vi.mocked(fetchRatingSourceOccupations).mockResolvedValue({
+    data_release_code: "onet-31.0",
+    source_table_code: "abilities",
+    source_available: true,
+    occupations: [
+      { onetsoc_code: "11-1011.00", occupation_title: "Chief Executives" },
+      { onetsoc_code: "15-1252.00", occupation_title: "Software Developers" },
+    ],
+  });
+});
+
 describe("OccupationRatingProfile", () => {
   it("submits exact identifiers and renders warnings beside the retained value", async () => {
+    vi.mocked(fetchOccupationRatingSources).mockResolvedValue({
+      sources: [{
+        data_release_code: "onet-31.0", release_version: "31.0",
+        source_publisher_name: "Synthetic publisher", source_license_url: "https://example.test/license",
+        source_table_code: "abilities", source_table_name: "Abilities",
+        source_artifact_url: "https://example.test/abilities.csv", source_artifact_sha256: "a".repeat(64),
+        source_row_count: 2,
+      }],
+    });
     vi.mocked(fetchOccupationRatings).mockResolvedValue(ready);
     render(<OccupationRatingProfile accessToken="synthetic-token" />);
-    await userEvent.type(screen.getByLabelText("O*NET-SOC 직업 코드"), "15-1252.00");
+    expect(await screen.findByRole("option", { name: "31.0 · Abilities" })).toBeInTheDocument();
+    await screen.findByRole("option", { name: "Software Developers · 15-1252.00" });
+    await userEvent.selectOptions(
+      await screen.findByLabelText("직업"),
+      "15-1252.00",
+    );
     await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
     expect(fetchOccupationRatings).toHaveBeenCalledWith("synthetic-token", {
       onetsocCode: "15-1252.00", dataReleaseCode: "onet-31.0", sourceTableCode: "abilities", offset: 0,
@@ -48,39 +90,144 @@ describe("OccupationRatingProfile", () => {
     expect(screen.getByText(/표를 가로로 밀어/)).toBeInTheDocument();
   });
 
-  it("hides stale pagination after form edits and loads the new profile", async () => {
-    vi.mocked(fetchOccupationRatings)
-      .mockResolvedValueOnce({ ...ready, next_offset: 100 })
-      .mockResolvedValueOnce({ ...ready, items: [{ ...ready.items[0], scale_id: "LV" }] });
+  it("fails closed when no imported rating source exists", async () => {
+    vi.mocked(fetchOccupationRatingSources).mockResolvedValue({ sources: [] });
     render(<OccupationRatingProfile accessToken="synthetic-token" />);
-    const occupation = screen.getByLabelText("O*NET-SOC 직업 코드");
-    await userEvent.type(occupation, "15-1252.00");
-    await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
-    await screen.findByText("4.10");
 
-    await userEvent.clear(occupation);
-    await userEvent.type(occupation, "11-1011.00");
-    expect(screen.queryByRole("button", { name: "다음 관측값 불러오기" })).not.toBeInTheDocument();
+    expect(await screen.findByText(/가져온 직업 근거 표가 없습니다/)).toBeInTheDocument();
+    expect(screen.getByRole("button", { name: "직업 근거 열기" })).toBeDisabled();
+  });
+
+  it("fails closed when an imported source has no selectable occupation", async () => {
+    vi.mocked(fetchOccupationRatingSources).mockResolvedValue({
+      sources: [{
+        data_release_code: "onet-31.0", release_version: "31.0",
+        source_publisher_name: "Synthetic publisher", source_license_url: "https://example.test/license",
+        source_table_code: "abilities", source_table_name: "Abilities",
+        source_artifact_url: "https://example.test/abilities.csv", source_artifact_sha256: "a".repeat(64),
+        source_row_count: 2,
+      }],
+    });
+    vi.mocked(fetchRatingSourceOccupations).mockResolvedValue({
+      data_release_code: "onet-31.0", source_table_code: "abilities",
+      source_available: true, occupations: [],
+    });
+    render(<OccupationRatingProfile accessToken="synthetic-token" />);
+
+    expect(await screen.findByText(/선택할 수 있는 직업이 없습니다/)).toBeInTheDocument();
+    expect(screen.getByLabelText("직업")).toBeDisabled();
+  });
+
+  it("distinguishes an unavailable occupation catalog from an empty one", async () => {
+    vi.mocked(fetchRatingSourceOccupations).mockResolvedValue({
+      data_release_code: "onet-31.0", source_table_code: "abilities",
+      source_available: false, occupations: [],
+    });
+    render(<OccupationRatingProfile accessToken="synthetic-token" />);
+
+    expect(await screen.findByText(/직업 목록이 아직 준비되지 않았습니다/)).toHaveAttribute("role", "status");
+    expect(screen.queryByRole("alert")).not.toBeInTheDocument();
+    expect(screen.queryByText(/선택할 수 있는 직업이 없습니다/)).not.toBeInTheDocument();
+  });
+
+  it("reports a transport failure separately from an unavailable occupation catalog", async () => {
+    vi.mocked(fetchRatingSourceOccupations).mockRejectedValue(new Error("synthetic transport failure"));
+    render(<OccupationRatingProfile accessToken="synthetic-token" />);
+
+    expect(await screen.findByRole("alert")).toHaveTextContent("직업 목록을 확인하지 못했습니다");
+    expect(screen.queryByText(/직업 목록이 아직 준비되지 않았습니다/)).not.toBeInTheDocument();
+  });
+
+  it("clears a stale catalog error when authentication changes", async () => {
+    vi.mocked(fetchOccupationRatingSources)
+      .mockRejectedValueOnce(new Error("synthetic catalog failure"))
+      .mockResolvedValueOnce({ sources: [{
+        data_release_code: "onet-31.0", release_version: "31.0",
+        source_publisher_name: "Synthetic publisher", source_license_url: "https://example.test/license",
+        source_table_code: "abilities", source_table_name: "Abilities",
+        source_artifact_url: "https://example.test/abilities.csv", source_artifact_sha256: "a".repeat(64),
+        source_row_count: 2,
+      }] });
+    const { rerender } = render(<OccupationRatingProfile accessToken="expired-token" />);
+    expect(await screen.findByRole("alert")).toHaveTextContent("근거 표를 확인하지 못했습니다");
+
+    rerender(<OccupationRatingProfile accessToken="fresh-token" />);
+
+    expect(await screen.findByRole("option", { name: "31.0 · Abilities" })).toBeInTheDocument();
+    expect(screen.queryByRole("alert")).not.toBeInTheDocument();
+  });
+
+  it("clears evidence and ignores an in-flight response when authentication changes", async () => {
+    let finishExpired: ((profile: Payload) => void) | undefined;
+    vi.mocked(fetchOccupationRatingSources).mockResolvedValue({ sources: [{
+      data_release_code: "onet-31.0", release_version: "31.0",
+      source_publisher_name: "Synthetic publisher", source_license_url: "https://example.test/license",
+      source_table_code: "abilities", source_table_name: "Abilities",
+      source_artifact_url: "https://example.test/abilities.csv", source_artifact_sha256: "a".repeat(64),
+      source_row_count: 2,
+    }] });
+    vi.mocked(fetchOccupationRatings).mockImplementation(
+      () => new Promise((resolve) => { finishExpired = resolve; }),
+    );
+    const { rerender } = render(<OccupationRatingProfile accessToken="expired-token" />);
+    await screen.findByRole("option", { name: "31.0 · Abilities" });
+    await screen.findByRole("option", { name: "Software Developers · 15-1252.00" });
+    await userEvent.selectOptions(screen.getByLabelText("직업"), "15-1252.00");
     await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
 
-    expect(fetchOccupationRatings).toHaveBeenLastCalledWith("synthetic-token", {
-      onetsocCode: "11-1011.00", dataReleaseCode: "onet-31.0", sourceTableCode: "abilities", offset: 0,
+    rerender(<OccupationRatingProfile accessToken="fresh-token" />);
+    finishExpired?.(ready);
+
+    expect(await screen.findByRole("option", { name: "31.0 · Abilities" })).toBeInTheDocument();
+    expect(screen.queryByText("4.10")).not.toBeInTheDocument();
+    expect(screen.getByRole("button", { name: "직업 근거 열기" })).toBeDisabled();
+  });
+
+  it("clears loaded evidence when the occupation selection changes", async () => {
+    vi.mocked(fetchOccupationRatingSources).mockResolvedValue({
+      sources: [{
+        data_release_code: "onet-31.0", release_version: "31.0",
+        source_publisher_name: "Synthetic publisher", source_license_url: "https://example.test/license",
+        source_table_code: "abilities", source_table_name: "Abilities",
+        source_artifact_url: "https://example.test/abilities.csv", source_artifact_sha256: "a".repeat(64),
+        source_row_count: 2,
+      }],
     });
-    expect(await screen.findByText("4.10")).toBeInTheDocument();
+    vi.mocked(fetchOccupationRatings).mockResolvedValueOnce({ ...ready, next_offset: 100 });
+    render(<OccupationRatingProfile accessToken="synthetic-token" />);
+    const occupation = await screen.findByLabelText("직업");
+    await screen.findByRole("option", { name: "Software Developers · 15-1252.00" });
+    await userEvent.selectOptions(occupation, "15-1252.00");
+    await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
+    await screen.findByText("4.10");
+
+    await userEvent.selectOptions(occupation, "11-1011.00");
+
+    expect(screen.queryByText("4.10")).not.toBeInTheDocument();
+    expect(screen.queryByRole("button", { name: "다음 관측값 불러오기" })).not.toBeInTheDocument();
   });
 
   it("removes stale evidence while a fresh occupation loads", async () => {
+    vi.mocked(fetchOccupationRatingSources).mockResolvedValue({
+      sources: [{
+        data_release_code: "onet-31.0", release_version: "31.0",
+        source_publisher_name: "Synthetic publisher", source_license_url: "https://example.test/license",
+        source_table_code: "abilities", source_table_name: "Abilities",
+        source_artifact_url: "https://example.test/abilities.csv", source_artifact_sha256: "a".repeat(64),
+        source_row_count: 2,
+      }],
+    });
     vi.mocked(fetchOccupationRatings)
       .mockResolvedValueOnce(ready)
       .mockImplementationOnce(() => new Promise(() => undefined));
     render(<OccupationRatingProfile accessToken="synthetic-token" />);
-    const occupation = screen.getByLabelText("O*NET-SOC 직업 코드");
-    await userEvent.type(occupation, "15-1252.00");
+    const occupation = await screen.findByLabelText("직업");
+    await screen.findByRole("option", { name: "Software Developers · 15-1252.00" });
+    await userEvent.selectOptions(occupation, "15-1252.00");
     await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
     await screen.findByText("4.10");
 
-    await userEvent.clear(occupation);
-    await userEvent.type(occupation, "11-1011.00");
+    await userEvent.selectOptions(occupation, "11-1011.00");
     await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
 
     expect(screen.queryByText("4.10")).not.toBeInTheDocument();
@@ -89,16 +236,23 @@ describe("OccupationRatingProfile", () => {
 
   it("ignores a superseded occupation response that finishes last", async () => {
     let finishFirst: ((profile: Payload) => void) | undefined;
+    vi.mocked(fetchOccupationRatingSources).mockResolvedValue({ sources: [{
+      data_release_code: "onet-31.0", release_version: "31.0",
+      source_publisher_name: "Synthetic publisher", source_license_url: "https://example.test/license",
+      source_table_code: "abilities", source_table_name: "Abilities",
+      source_artifact_url: "https://example.test/abilities.csv", source_artifact_sha256: "a".repeat(64),
+      source_row_count: 2,
+    }] });
     vi.mocked(fetchOccupationRatings)
       .mockImplementationOnce(() => new Promise((resolve) => { finishFirst = resolve; }))
       .mockResolvedValueOnce({ ...ready, onetsoc_code: "11-1011.00", items: [{ ...ready.items[0], data_value: "3.20" }] });
     render(<OccupationRatingProfile accessToken="synthetic-token" />);
-    const occupation = screen.getByLabelText("O*NET-SOC 직업 코드");
-    await userEvent.type(occupation, "15-1252.00");
+    const occupation = await screen.findByLabelText("직업");
+    await screen.findByRole("option", { name: "Software Developers · 15-1252.00" });
+    await userEvent.selectOptions(occupation, "15-1252.00");
     await userEvent.click(screen.getByRole("button", { name: "직업 근거 열기" }));
 
-    await userEvent.clear(occupation);
-    await userEvent.type(occupation, "11-1011.00");
+    await userEvent.selectOptions(occupation, "11-1011.00");
     fireEvent.submit(occupation.closest("form")!);
     expect(await screen.findByText("3.20")).toBeInTheDocument();
 
@@ -111,7 +265,7 @@ describe("OccupationRatingProfile", () => {
     const { rerender } = render(<OccupationRatingProfileView profile={{ ...ready, source_available: false, source: null, items: [] }} />);
     expect(screen.getByRole("status")).toHaveTextContent("아직 준비되지 않았습니다");
     rerender(<OccupationRatingProfileView profile={{ ...ready, items: [] }} />);
-    expect(screen.getByRole("status")).toHaveTextContent("관측값이 없습니다");
+    expect(screen.getByRole("status")).toHaveTextContent("직업이나 근거 표를 바꿔");
   });
 
   it("does not turn a non-http artifact value into a customer link", () => {
diff --git a/frontend/src/components/OccupationRatingProfile.tsx b/frontend/src/components/OccupationRatingProfile.tsx
index 2d1445187..9f733b5b5 100644
--- a/frontend/src/components/OccupationRatingProfile.tsx
+++ b/frontend/src/components/OccupationRatingProfile.tsx
@@ -1,7 +1,11 @@
-import { useRef, useState } from "react";
+import { useEffect, useRef, useState } from "react";
 import {
+  fetchOccupationRatingSources,
   fetchOccupationRatings,
+  fetchRatingSourceOccupations,
   type OccupationRatingProfile as OccupationRatingProfilePayload,
+  type OccupationRatingSource,
+  type RatingSourceOccupation,
 } from "../api";
 
 type Props = { accessToken: string };
@@ -19,26 +23,94 @@ function safeHttpUrl(value: string | null | undefined): string | null {
 /** Lets an authenticated user inspect one exact imported occupation profile. */
 export function OccupationRatingProfile({ accessToken }: Props) {
   const [onetsocCode, setOnetsocCode] = useState("");
-  const [releaseCode, setReleaseCode] = useState("onet-31.0");
-  const [sourceCode, setSourceCode] = useState("abilities");
+  const [sources, setSources] = useState<OccupationRatingSource[] | null>(null);
+  const [selectedSource, setSelectedSource] = useState("");
+  const [sourceCatalogError, setSourceCatalogError] = useState(false);
+  const [occupations, setOccupations] = useState<RatingSourceOccupation[] | null>(null);
+  const [occupationCatalogUnavailable, setOccupationCatalogUnavailable] = useState(false);
+  const [occupationCatalogError, setOccupationCatalogError] = useState(false);
   const [profile, setProfile] = useState<OccupationRatingProfilePayload | null>(null);
   const [status, setStatus] = useState<"idle" | "loading" | "error">("idle");
   const requestSequence = useRef(0);
+  const selectedSourceRecord = sources?.find(
+    (item) => `${item.data_release_code}|${item.source_table_code}` === selectedSource,
+  );
   const profileMatchesForm = profile != null
     && profile.onetsoc_code === onetsocCode
-    && profile.data_release_code === releaseCode
-    && profile.source_table_code === sourceCode;
+    && profile.data_release_code === selectedSourceRecord?.data_release_code
+    && profile.source_table_code === selectedSourceRecord?.source_table_code;
+
+  useEffect(() => {
+    let active = true;
+    requestSequence.current += 1;
+    setSourceCatalogError(false);
+    setSources(null);
+    setSelectedSource("");
+    setProfile(null);
+    setStatus("idle");
+    fetchOccupationRatingSources(accessToken)
+      .then(({ sources: loaded }) => {
+        if (!active) return;
+        setSources(loaded);
+        setSelectedSource(
+          loaded[0] ? `${loaded[0].data_release_code}|${loaded[0].source_table_code}` : "",
+        );
+      })
+      .catch(() => active && setSourceCatalogError(true));
+    return () => { active = false; };
+  }, [accessToken]);
+
+  useEffect(() => {
+    requestSequence.current += 1;
+    setStatus("idle");
+    const source = sources?.find(
+      (item) => `${item.data_release_code}|${item.source_table_code}` === selectedSource,
+    );
+    setOnetsocCode("");
+    setProfile(null);
+    setOccupationCatalogUnavailable(false);
+    setOccupationCatalogError(false);
+    if (!source) {
+      setOccupations(null);
+      return;
+    }
+    let active = true;
+    setOccupations(null);
+    fetchRatingSourceOccupations(
+      accessToken,
+      source.data_release_code,
+      source.source_table_code,
+    )
+      .then((payload) => {
+        if (!active) return;
+        if (!payload.source_available) {
+          setOccupationCatalogUnavailable(true);
+          return;
+        }
+        setOccupations(payload.occupations);
+      })
+      .catch(() => active && setOccupationCatalogError(true));
+    return () => { active = false; };
+  }, [accessToken, selectedSource, sources]);
 
   function load(offset: number | null = null) {
     const requestId = requestSequence.current + 1;
     requestSequence.current = requestId;
-    const request = offset == null
-      ? { onetsocCode, dataReleaseCode: releaseCode, sourceTableCode: sourceCode }
-      : {
-          onetsocCode: profile?.onetsoc_code ?? onetsocCode,
-          dataReleaseCode: profile?.data_release_code ?? releaseCode,
-          sourceTableCode: profile?.source_table_code ?? sourceCode,
-        };
+    const source = selectedSourceRecord;
+    const request = offset == null && source
+      ? {
+          onetsocCode,
+          dataReleaseCode: source.data_release_code,
+          sourceTableCode: source.source_table_code,
+        }
+      : profile
+        ? {
+            onetsocCode: profile.onetsoc_code,
+            dataReleaseCode: profile.data_release_code,
+            sourceTableCode: profile.source_table_code,
+          }
+        : null;
+    if (!request) return;
     if (offset == null) setProfile(null);
     setStatus("loading");
     fetchOccupationRatings(accessToken, {
@@ -64,7 +136,7 @@ export function OccupationRatingProfile({ accessToken }: Props) {
       <header>
         <p className="dashboard-eyebrow">공개 직업 근거</p>
         <h2 id="occupation-rating-heading">직업별 업무 특성 확인</h2>
-        <p>직업 코드와 근거 표를 선택해 관측값, 오차, 사용 주의사항을 함께 확인하세요.</p>
+        <p>직업과 근거 표를 선택해 관측값, 오차, 사용 주의사항을 함께 확인하세요.</p>
       </header>
       <form
         className="occupation-rating-form"
@@ -74,29 +146,52 @@ export function OccupationRatingProfile({ accessToken }: Props) {
         }}
       >
         <label>
-          O*NET-SOC 직업 코드
-          <input
+          직업
+          <select
             required
-            pattern="[0-9]{2}-[0-9]{4}\.[0-9]{2}"
-            placeholder="15-1252.00"
             value={onetsocCode}
-            onChange={(event) => setOnetsocCode(event.target.value)}
-          />
-        </label>
-        <label>
-          데이터 릴리스
-          <input required value={releaseCode} onChange={(event) => setReleaseCode(event.target.value)} />
+            onChange={(event) => {
+              requestSequence.current += 1;
+              setOnetsocCode(event.target.value);
+              setProfile(null);
+              setStatus("idle");
+            }}
+            disabled={occupations === null || occupations.length === 0}
+          >
+            <option value="">직업 선택</option>
+            {(occupations ?? []).map((occupation) => (
+              <option key={occupation.onetsoc_code} value={occupation.onetsoc_code}>
+                {occupation.occupation_title} · {occupation.onetsoc_code}
+              </option>
+            ))}
+          </select>
         </label>
-        <label>
-          근거 표
-          <input required value={sourceCode} onChange={(event) => setSourceCode(event.target.value)} />
+        <label className="occupation-rating-source-select">
+          근거 릴리스·표
+          <select required value={selectedSource} onChange={(event) => setSelectedSource(event.target.value)}>
+            {(sources ?? []).map((source) => (
+              <option
+                key={`${source.data_release_code}|${source.source_table_code}`}
+                value={`${source.data_release_code}|${source.source_table_code}`}
+              >
+                {source.release_version} · {source.source_table_name}
+              </option>
+            ))}
+          </select>
         </label>
-        <button className="btn-secondary" type="submit" disabled={status === "loading"}>
+        <button className="btn-secondary" type="submit" disabled={status === "loading" || !selectedSource || !onetsocCode}>
           {status === "loading" ? "근거를 불러오는 중" : "직업 근거 열기"}
         </button>
       </form>
+      {sources === null && !sourceCatalogError ? <p role="status">사용 가능한 근거 표를 확인하는 중입니다.</p> : null}
+      {sources?.length === 0 ? <p role="status">가져온 직업 근거 표가 없습니다. 데이터 담당자에게 근거 가져오기를 요청하세요.</p> : null}
+      {sourceCatalogError ? <p role="alert">사용 가능한 근거 표를 확인하지 못했습니다. 잠시 후 다시 열어 보세요.</p> : null}
+      {selectedSource && occupations === null && !occupationCatalogUnavailable && !occupationCatalogError ? <p role="status">이 근거 표의 직업 목록을 확인하는 중입니다.</p> : null}
+      {selectedSource && occupations?.length === 0 ? <p role="status">이 근거 표에 선택할 수 있는 직업이 없습니다. 다른 근거 표를 선택하세요.</p> : null}
+      {occupationCatalogUnavailable ? <p role="status">이 근거 표의 직업 목록이 아직 준비되지 않았습니다. 다른 근거 표를 선택하거나 데이터 담당자에게 가져오기를 요청하세요.</p> : null}
+      {occupationCatalogError ? <p role="alert">직업 목록을 확인하지 못했습니다. 잠시 후 다시 열어 보세요.</p> : null}
       {status === "error" ? (
-        <p role="alert">직업 근거를 불러오지 못했습니다. 코드와 접근 권한을 확인한 뒤 다시 시도하세요.</p>
+        <p role="alert">직업 근거를 불러오지 못했습니다. 선택 항목과 접근 권한을 확인한 뒤 다시 시도하세요.</p>
       ) : null}
       {profile ? <OccupationRatingProfileView profile={profile} /> : null}
       {profileMatchesForm && profile.next_offset != null ? (
@@ -129,7 +224,7 @@ export function OccupationRatingProfileView({
   if (profile.items.length === 0) {
     return (
       <p role="status">
-        이 근거 표에는 선택한 직업의 관측값이 없습니다. 직업 코드나 근거 표를 바꿔 확인하세요.
+        이 근거 표에는 선택한 직업의 관측값이 없습니다. 직업이나 근거 표를 바꿔 확인하세요.
       </p>
     );
   }
diff --git a/tests/test_occupation_rating_ingestion.py b/tests/test_occupation_rating_ingestion.py
index 3f2244d7f..3f8ec21dd 100644
--- a/tests/test_occupation_rating_ingestion.py
+++ b/tests/test_occupation_rating_ingestion.py
@@ -3,8 +3,16 @@
 import asyncio
 from decimal import Decimal
 
-from backend.app.main import read_occupation_ratings
-from backend.app.occupation_rating_ingestion import fetch_occupation_ratings
+from backend.app.main import (
+    read_occupation_rating_sources,
+    read_occupation_ratings,
+    read_rating_source_occupations,
+)
+from backend.app.occupation_rating_ingestion import (
+    fetch_occupation_rating_sources,
+    fetch_occupation_ratings,
+    fetch_rating_source_occupations,
+)
 
 
 class FakeConnection:
@@ -14,14 +22,16 @@ def __init__(self, source, rows=()):
         self.source = source
         self.rows = list(rows)
         self.fetch_called = False
+        self.last_fetch_query = ""
 
     async def fetchrow(self, _query: str, *_args: object):
         """Return configured source metadata."""
         return self.source
 
-    async def fetch(self, _query: str, *_args: object):
+    async def fetch(self, query: str, *_args: object):
         """Return configured observation rows."""
         self.fetch_called = True
+        self.last_fetch_query = query
         return self.rows
 
 
@@ -160,3 +170,88 @@ def test_authenticated_route_delegates_to_bounded_projection() -> None:
     )
 
     assert result["source_available"] is False
+
+
+def test_source_catalog_returns_only_query_selected_imports() -> None:
+    source = {
+        "data_release_code": "onet-31.0",
+        "release_version": "31.0",
+        "source_publisher_name": "National Center for O*NET Development",
+        "source_license_url": "https://example.test/license",
+        "source_table_code": "abilities",
+        "source_table_name": "Abilities",
+        "source_artifact_url": "https://example.test/abilities.csv",
+        "source_artifact_sha256": "a" * 64,
+        "source_row_count": 94640,
+    }
+
+    conn = FakeConnection(None, (source,))
+    result = asyncio.run(fetch_occupation_rating_sources(conn))
+
+    assert result == {"sources": [source]}
+    assert "source_table_code <> 'scales_reference'" in conn.last_fetch_query
+    assert "and exists" in conn.last_fetch_query
+
+
+def test_authenticated_source_catalog_route_uses_shared_projection() -> None:
+    result = asyncio.run(
+        read_occupation_rating_sources(
+            _account=object(),
+            pool=FakePool(FakeConnection(None)),
+        )
+    )
+
+    assert result == {"sources": []}
+
+
+def test_source_occupation_catalog_distinguishes_unavailable_from_empty() -> None:
+    unavailable = asyncio.run(
+        fetch_rating_source_occupations(
+            FakeConnection(None),
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+        )
+    )
+    empty = asyncio.run(
+        fetch_rating_source_occupations(
+            FakeConnection({"exists": 1}),
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+        )
+    )
+
+    assert unavailable["source_available"] is False
+    assert empty["source_available"] is True
+    assert empty["occupations"] == []
+
+
+def test_source_occupation_catalog_returns_authoritative_codes_and_titles() -> None:
+    rows = (
+        {"onetsoc_code": "11-1011.00", "occupation_title": "Chief Executives"},
+        {"onetsoc_code": "15-1252.00", "occupation_title": "Software Developers"},
+    )
+    conn = FakeConnection({"exists": 1}, rows)
+
+    result = asyncio.run(
+        fetch_rating_source_occupations(
+            conn,
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+        )
+    )
+
+    assert result["occupations"] == list(rows)
+    assert "and exists" in conn.last_fetch_query
+
+
+def test_authenticated_source_occupation_route_uses_shared_projection() -> None:
+    result = asyncio.run(
+        read_rating_source_occupations(
+            data_release_code="onet-31.0",
+            source_table_code="abilities",
+            _account=object(),
+            pool=FakePool(FakeConnection({"exists": 1})),
+        )
+    )
+
+    assert result["source_available"] is True
diff --git a/tests/test_schema.py b/tests/test_schema.py
index d12fa59de..391367af5 100644
--- a/tests/test_schema.py
+++ b/tests/test_schema.py
@@ -28,7 +28,11 @@
 import psycopg2.errors
 import pytest
 
-from backend.app.occupation_rating_ingestion import fetch_occupation_ratings
+from backend.app.occupation_rating_ingestion import (
+    fetch_occupation_rating_sources,
+    fetch_occupation_ratings,
+    fetch_rating_source_occupations,
+)
 from backend.app.post_chat_ingestion import gather_global_chat_sources
 from scripts.import_onet_ratings import import_ratings
 
@@ -528,10 +532,10 @@ def test_onet_rating_importer_is_idempotent_against_postgresql(
         )
         assert cur.fetchone() == (1, Decimal("4.10"), True)
 
-    async def read_imported_profile() -> dict[str, object]:
+    async def read_imported_profile() -> tuple[dict[str, object], dict[str, object], dict[str, object]]:
         conn = await asyncpg.connect(args.target_dsn)
         try:
-            return await fetch_occupation_ratings(
+            profile = await fetch_occupation_ratings(
                 conn,
                 data_release_code=args.release_code,
                 source_table_code=args.source_table_code,
@@ -539,13 +543,24 @@ async def read_imported_profile() -> dict[str, object]:
                 limit=100,
                 offset=0,
             )
+            catalog = await fetch_occupation_rating_sources(conn)
+            occupations = await fetch_rating_source_occupations(
+                conn,
+                data_release_code=args.release_code,
+                source_table_code=args.source_table_code,
+            )
+            return profile, catalog, occupations
         finally:
             await conn.close()
 
-    profile = asyncio.run(read_imported_profile())
+    profile, catalog, occupations = asyncio.run(read_imported_profile())
     assert profile["source_available"] is True
     assert profile["items"][0]["data_value"] == "4.10"
     assert profile["source"]["scale_artifact_sha256"] == args.scales_sha256
+    assert catalog["sources"][0]["source_table_code"] == "abilities"
+    assert occupations["occupations"] == [
+        {"onetsoc_code": "15-1252.00", "occupation_title": "Synthetic occupation"}
+    ]
 
 
 def test_global_ask_evidence_search_indexes_exist_on_normalized_tables(schema_db) -> None: