feat(ask): bind relative-time filters to event time (v2.13.0)

AGENTS.md

@@ -276,6 +276,12 @@ and are not a leftover score. Complete-case coverage (ADR 0168) persists to
 `report_leftover_map_coverage` and captions the pair list with how
 many scored posts entered the map.
 
+Global Ask relative-time filters (ADR 0150 / 0202) bind to
+`source_post.event_occurred_at` and fall back to `created_at` only
+when the event instant is missing. Cited evidence names **Time
+axis** so the reader can open that post and see which clock
+matched. Do not invent an event date or a theta.
+
 Period leftover pairs (ADR 0017 / 0018 / 0048 / 0049 / 0149) are computed in
 `lineageweave/leftover_pairs.py` from the residual after a real
 GRM/GPCM score, never invented. Missing cells stay out of the

ARCHITECTURE.md

@@ -78,6 +78,8 @@ flowchart LR
 | `post_summary.py` | Pluggable LLM Korean summary + key events + R&R derivation for a post |
 | `post_chat.py` | Pluggable in-popup chat's reason-and-cite step (retrieve step lives in `backend/app/post_chat_ingestion.py`) |
 | `commitment_extraction.py` | Pluggable LLM derivation of a customer commitment (promise + deadline) from a post; `Null` default, `ContextualOrchestrator` real impl |
+| `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 `docs/ontology/lineageweave-kg.ttl`, the formal OWL 2/RDFS/SKOS vocabulary for the Knowledge Graph's node/edge types (ADR 0004) |
 | `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 |

CHANGELOG.d/2.17.0-ask-event-time-filter.md

@@ -0,0 +1,3 @@
+Global Ask relative-time filters bind to event_occurred_at with a
+created_at fallback and name the matching time axis on cited evidence
+(ADR 0202 / #569).

CHANGELOG.md

@@ -90,6 +90,15 @@ All notable changes to this project are documented here. Format follows
   session token is removed, hides live refocus on static catalog snapshots,
   and looks up corporate-parent visibility independently of the child. OWL-Time
   is cited as a W3C Candidate Recommendation Draft.
+- Global Ask relative-time filters bind to `event_occurred_at` (the
+  source-system event instant) and fall back to `created_at` only when
+  that event clock is missing (ADR 0202 / #569). After `make seed`,
+  leftover closest/farthest pairs still sit above the member list; a
+  click still opens that post. Ask **어제 무슨 일이 있었나요?** on
+  bulk-imported fixtures that share one ingest day keeps the post whose
+  event fell yesterday, drops last week's event, and names **Time axis**
+  on the cited evidence. Open that cited post to read which clock
+  matched. Never invent a theta or an event date.
 
 ## [2.15.0] - 2026-08-25
 

CLAUDE.md

@@ -54,60 +54,3 @@ migration replay (ADR 0166), docstring coverage, and the measurement
 boundary are all stated in [AGENTS.md](AGENTS.md) -- read it before
 changing code, tests, or runtime policy rather than restating anything
 here.
-=======
-period-report run on the same snapshot (ADR 0013 / ADR 0024). The TEPP path goes through `tepp_client`. A missing
-transport or an unused accepted envelope is Failed
-(`tepp_not_available` / `tepp_result_not_persisted`). Do not invent a
-theta or a local psychometric substitute. The home list caption stays
-`kind · status · entity`; the machine failure code is detail-only
-(ADR 0014). Open a Failed TEPP row, then connect a live TEPP
-transport. A failed lineage row retries reconstruction -- it does not
-mention TEPP. A failed period-report row rebuilds the report. A
-pending TEPP row does not claim a calibrated measurement. A pending
-lineage row says reconstruction has not started yet.
-Digest prefixes stay audible; hover a prefix to read the full digest.
-Opening a cutoff title shows the live post. Titles marked updated
-after cutoff were rewritten after the run; the opened body names
-both clocks and shows **Body this run knew** beside the live
-rewrite. Compare those two texts before treating the live body as
-reconstructed evidence (ADR 0016 / 0025).
-`POST /api/analysis-runs` records Pending lineage only on an
-authorized cutoff capture (ADR 0017). TEPP and period-report kinds
-are 422. The Request button waits until affiliated corps load; choose
-a corp if the token walks more than one. `POST /api/analysis-runs/{id}/start`
-commits Running plus a durable outbox row, then reconstructs that
-frozen cutoff bag (ADR 0021 / ADR 0023) or submits TEPP through
-`tepp_client` (ADR 0022). A missing transport or unused accepted
-envelope is Failed. Failed TEPP is terminal — connect a TEPP
-transport from that Failed row. Create does not invent a Pending
-TEPP row. Do not invent a theta. Hover the Result prefix to read
-the parent-choice digest.
-After `make seed`, open **Period report · Succeeded · Demo Corp**,
-then **Open period report 2026-W02**. The home week is already
-2026-W02, so the grouping comparison strip lands on Demo Corp. Report
-grouping is Corporate entity and Demo Corp is current. The focused
-chip name contains `Corporate entity: Demo Corp` and the persisted
-mean θ. The period-report panel says Demo Corp is the opened grouping
-and to read its mean θ and member posts, then open a post. Those
-members land immediately under that next action, ahead of Other Corp
-and the week strip. After `make seed`, leftover closest/farthest pairs
-sit above the member list with leftover-map rank (rank 0 names no
-leftover structure) and unexplained leftover `U` next to leftover-map
-distance `d`. Leftover-map axis share badges name Gabriel inertia
-of axes 1 and 2; open a leftover pair to read the post–criterion cell.
-The shares do not invent a leftover score. Opening Public post names the next action: read
-Event Lineage, Keyman, and evaluation on that post. The popup Event
-Lineage DAG marks that post current. After that current node, the
-popup names Keyman and evaluation as the next read. After landed
-evaluation, the popup names the first Keyman as the next read. After
-landed Ada West related, the popup names the first related node as
-the next read. After that next action, the popup lands Priya Nair
-related nodes. After those related nodes land, the popup names Ask
-about this lineage as the next read. After that next action, the
-popup lands Ask about this lineage. After landed chat, the popup names
-the first Ask. After that next action, the popup lands the first Ask
-answer. After landed first Ask answer, the popup names the first
-cited source. After that next action, the popup lands the first cited
-evidence. Changing the week first still
-focuses the report period field. Mean θ stays on the period-report
-panel.

backend/app/post_chat_ingestion.py

@@ -26,6 +26,7 @@
 
 import asyncpg
 
+from lineageweave.ask_time_axis import row_matches_time_range, time_axis_evidence_fact
 from lineageweave.image_content import ImageContentClient, NullImageContentClient
 from lineageweave.knowledge_graph import (
     NODE_POST,
@@ -406,9 +407,11 @@ async def gather_global_chat_sources(
 
     When `question` contains a Korean relative-time expression ("어제",
     "작년 이맘때쯤", "3일 전", ...; see `lineageweave.temporal_expressions`),
-    the final candidate set is also bounded to posts whose `created_at`
-    falls in the resolved date range -- an unbounded expression ("언젠가")
-    or no expression at all applies no date filter.
+    the final candidate set is bounded to posts whose event instant
+    (`event_occurred_at`, falling back to `created_at`) falls in the
+    resolved Seoul date range -- an unbounded expression ("언젠가")
+    or no expression at all applies no date filter. Cited sources name
+    which clock matched (ADR 0202).
 
     The source set is intentionally bounded until retrieval/reranking is
     needed for a much larger corpus; every selected body still uses the same
@@ -419,8 +422,9 @@ async def gather_global_chat_sources(
     if vision_client is None:
         vision_client = NullImageContentClient()
     # A relative-time expression ("어제", "작년 이맘때쯤", ...) narrows the
-    # candidate window by created_at below; it must not also become a
-    # near-meaningless literal keyword search term (see TEMPORAL_STOPWORDS).
+    # candidate window by event time (fallback: created_at) below; it must
+    # not also become a near-meaningless literal keyword search term
+    # (see TEMPORAL_STOPWORDS).
     resolved_time_range = resolve_korean_relative_time(
         question or "", today=today or _seoul_today()
     )
@@ -475,24 +479,34 @@ async def gather_global_chat_sources(
             """
             select post_id, matched_in
               from (
-                   (select post_id, created_at, 'title' as matched_in
+                   (select post_id, coalesce(event_occurred_at, created_at) as event_clock,
+                           'title' as matched_in
                       from source_post
                      where post_title ilike '%' || $1 || '%'
+                       and ($2::date is null or (coalesce(event_occurred_at, created_at) at time zone 'Asia/Seoul')::date >= $2)
+                       and ($3::date is null or (coalesce(event_occurred_at, created_at) at time zone 'Asia/Seoul')::date <= $3)
                      limit 32)
                     union all
-                   (select post_id, created_at, 'body' as matched_in
+                   (select post_id, coalesce(event_occurred_at, created_at) as event_clock,
+                           'body' as matched_in
                       from source_post
                      where lower(left(source_post_search_text(post_body), 16384))
                                like '%' || lower($1) || '%'
+                       and ($2::date is null or (coalesce(event_occurred_at, created_at) at time zone 'Asia/Seoul')::date >= $2)
+                       and ($3::date is null or (coalesce(event_occurred_at, created_at) at time zone 'Asia/Seoul')::date <= $3)
                      limit 32)
                     union all
-                   (select post_id, created_at, 'body' as matched_in
+                   (select post_id, coalesce(event_occurred_at, created_at) as event_clock,
+                           'body' as matched_in
                       from source_post
                      where to_tsvector('simple', source_post_search_text(post_body))
                                @@ plainto_tsquery('simple', $1)
+                       and ($2::date is null or (coalesce(event_occurred_at, created_at) at time zone 'Asia/Seoul')::date >= $2)
+                       and ($3::date is null or (coalesce(event_occurred_at, created_at) at time zone 'Asia/Seoul')::date <= $3)
                      limit 32)
                     union all
-                   (select post_id, created_at, 'source_field' as matched_in
+                   (select post_id, coalesce(event_occurred_at, created_at) as event_clock,
+                           'source_field' as matched_in
                       from source_post
                      where concat_ws(' ', source_system_code, source_record_key,
                                       source_author_code, source_author_name,
@@ -502,12 +516,16 @@ async def gather_global_chat_sources(
                                       source_customer_code, source_customer_name,
                                       source_project_code, source_project_name)
                                ilike '%' || $1 || '%'
+                       and ($2::date is null or (coalesce(event_occurred_at, created_at) at time zone 'Asia/Seoul')::date >= $2)
+                       and ($3::date is null or (coalesce(event_occurred_at, created_at) at time zone 'Asia/Seoul')::date <= $3)
                      limit 32)
                    ) matches
-             order by created_at desc, post_id desc
+             order by event_clock desc, post_id desc
             limit 32
             """,
             term,
+            resolved_time_range[0] if resolved_time_range else None,
+            resolved_time_range[1] if resolved_time_range else None,
         )
         for row in candidate_rows:
             post_id = str(row["post_id"])
@@ -551,14 +569,15 @@ async def gather_global_chat_sources(
                source_company_code, source_company_name, source_process_unit_code,
                source_process_unit_name, source_sales_pool_code, source_sales_pool_name,
                source_customer_code, source_customer_name,
-               source_project_code, source_project_name
+               source_project_code, source_project_name,
+               created_at, event_occurred_at
           from source_post
          where (visibility_code = 'public'
             or corporate_entity_id::text = any($1::text[]))
-           and ($4::date is null or (created_at at time zone 'Asia/Seoul')::date >= $4)
-           and ($5::date is null or (created_at at time zone 'Asia/Seoul')::date <= $5)
+           and ($4::date is null or (coalesce(event_occurred_at, created_at) at time zone 'Asia/Seoul')::date >= $4)
+           and ($5::date is null or (coalesce(event_occurred_at, created_at) at time zone 'Asia/Seoul')::date <= $5)
          order by array_position($2::uuid[], post_id) nulls last,
-                  created_at desc, post_id desc
+                  coalesce(event_occurred_at, created_at) desc, post_id desc
          limit $3
         """,
         list(authorized_corporate_entity_ids),
@@ -567,11 +586,16 @@ async def gather_global_chat_sources(
         resolved_time_range[0] if resolved_time_range else None,
         resolved_time_range[1] if resolved_time_range else None,
     )
-    visible_rows = [row for row in rows if can_see_post(row)][:limit]
+    visible_rows = [
+        row
+        for row in rows
+        if can_see_post(row) and row_matches_time_range(row, resolved_time_range)
+    ][:limit]
     visible_ids = [str(row["post_id"]) for row in visible_rows]
     anchor_is_visible = lineage_anchor_id in visible_ids
     semantic_facts = await _semantic_facts_for_posts(conn, visible_ids)
     graph_facts = (await _graph_facts_for_posts(conn, visible_ids))[:16]
+    time_filter_active = resolved_time_range is not None
     sources: list[ChatSourceDocument] = []
     for index, row in enumerate(visible_rows):
         normalized_body = await _normalize_post_body_text(row["post_body"], vision_client)
@@ -594,7 +618,8 @@ async def gather_global_chat_sources(
                 graph_facts=graph_facts if index == 0 else (),
                 evidence_facts=_source_hint_facts(row)
                 + semantic_facts.get(post_id, ())
-                + lineage_fact,
+                + lineage_fact
+                + time_axis_evidence_fact(row, time_filter_active=time_filter_active),
             )
         )
     return sources

backend/tests/test_api.py

@@ -162,6 +162,11 @@
     / "migrations"
     / "0182_report_leftover_map_unexplained.sql"
 )
+_EVENT_OCCURRED_AT_MIGRATION = (
+    Path(__file__).resolve().parents[2]
+    / "migrations"
+    / "0183_source_post_event_occurred_at.sql"
+)
 
 
 def _postgres_available() -> bool:
@@ -303,6 +308,7 @@ def seeded_db(demo_analyst_token):
             cur.execute(_LEFTOVER_MAP_RANK_MIGRATION.read_text())
             cur.execute(_LEFTOVER_MAP_COVERAGE_MIGRATION.read_text())
             cur.execute(_GLOBAL_ASK_JOB_MIGRATION.read_text())
+            cur.execute(_EVENT_OCCURRED_AT_MIGRATION.read_text())
             cur.execute(_LEFTOVER_MAP_AXIS_MIGRATION.read_text())
             cur.execute(_LEFTOVER_MAP_UNEXPLAINED_MIGRATION.read_text())
             cur.execute(

docs/adr/0150-korean-relative-time-retrieval.md

@@ -2,7 +2,7 @@
 
 - Status: Accepted
 - Date: 2026-08-22
-- Related: [0047](0047-global-ask-semantic-retrieval.md), [0090](0090-global-ask-lineage-timeline-expansion.md)
+- Related: [0047](0047-global-ask-semantic-retrieval.md), [0090](0090-global-ask-lineage-timeline-expansion.md), [0202](0202-ask-event-time-filter.md)
 
 ## Context
 
@@ -27,12 +27,13 @@ bound -- the reader has explicitly declined to name one, which is the same
 retrieval behavior as finding no expression at all.
 
 `gather_global_chat_sources` applies the resolved window as an additional
-`created_at` bound on its final ABAC-filtered candidate query, additive to
-the existing keyword-match ranking -- it narrows the already-ranked
-candidate set, it does not replace ranking with a date filter. Matched
-temporal literals are excluded from keyword-term extraction
-(`TEMPORAL_STOPWORDS`) so a resolved expression does not also become a
-near-meaningless literal search term.
+event-time bound on its final ABAC-filtered candidate query (ADR 0202:
+`coalesce(event_occurred_at, created_at)`), additive to the existing
+keyword-match ranking -- it narrows the already-ranked candidate set, it
+does not replace ranking with a date filter. Cited sources name which
+clock matched. Matched temporal literals are excluded from keyword-term
+extraction (`TEMPORAL_STOPWORDS`) so a resolved expression does not also
+become a near-meaningless literal search term.
 
 ## Considered alternatives