feat(retrieval): a task's work logs join the document it is embedded as (#4251)

Step 1 of #4251. A work log is the richest prose Scribe holds about WHY
something is the way it is — written during the work, recording what was tried
and ruled out. #4241 made it readable from the agent's door. It was still not
findable, so "has anyone tried this approach?" — precisely the question a log
answers — could not reach one. The cost is not hypothetical: #4208 was rebuilt
in this session because its logs were unreachable.

THE DESIGN QUESTION the issue left open was whether logs embed as part of their
task's document or as rows of their own. As their own rows, a hit has to be
resolved back to a task to be worth anything, and it needs a fourth search, a
fourth result shape and a fourth arm. As part of the task, the objection is
that a long log drowns a short title.

That objection was true before #280 and is not true now. Chunking made one
record into one vector per section, so each log becomes its own title-anchored
chunk, scored separately, and the task's own prose keeps the chunk it always
had — a task is as findable as its best-matching log rather than as the average
of everything in it. The other half is this session's other build: a search
hands back the chunk that won (#4243), so a hit earned by a log shows that
log's passage under the task's title. Without that a reader would have got
body[:240] of the task — the opening of a record whose relevance lives three
hundred lines further down.

So `task_document(title, body, logs)` sits beside `rule_document`: a
synthesised embed-time shape, because the stored record is the task row and the
logs live in their own table, so the document that should be searchable exists
nowhere until it is built. A task with no logs is returned untouched — most
notes are not tasks and most tasks carry no log, and their vectors are the
corpus every tuned number here was measured against.

CHUNKER_VERSION 1 → 2, and its comment now says what the version actually
means. It used to read "whenever chunk_document's output can change for the
same input", which this change would slip past: `chunk_document` is untouched
and every task with a log now embeds differently while its title, body and the
chunker all stand still. The invariant is the document a record is embedded as.
The startup backfill re-embeds on that.

Create, edit and delete of a log all refresh the task through `embed_note`, the
one path every writer shares — an edited log whose vectors still carry its old
wording keeps matching what it no longer says.

No new kind enters the auto-inject menu: tasks were always in it, and this
makes recall on them better rather than changing what the menu spans. The
calibration stamp will now report shape_version 2 against numbers measured at
1, which is exactly the report #4104 built it to make.

Also promoted `session_returning` into tests/helpers beside `make_mock_session`
(#2834) — two files had spelled it out identically and a third was about to.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01821k5B3Ysecp9fNYs92Kuy

src/scribe/services/embeddings.py

@@ -23,6 +23,7 @@ from sqlalchemy import delete, or_, select
 from scribe.models import async_session
 from scribe.models.embedding import NoteEmbedding, RuleEmbedding
 from scribe.models.note import Note
+from scribe.models.task_log import TaskLog
 from scribe.services.access import can_read_project, notes_visibility_clause
 
 if TYPE_CHECKING:  # resolves forward refs without importing at runtime
@@ -271,11 +272,20 @@ def untrigger_title(title: str | None, trigger: str | None) -> str:
 # search. The fix is the document shape: one vector per meaningful chunk, and a
 # record is as findable as its best-matching section.
 
-# Bumped whenever chunk_document's output can change for the same input. Stored
-# on every note_embeddings row so the startup backfill can re-embed exactly the
-# notes whose stored shape is stale — a version comparison instead of the table
-# wipe migrations 0067/0077 had to do.
-CHUNKER_VERSION = 1
+# Bumped whenever THE DOCUMENT A RECORD IS EMBEDDED AS can change for a record
+# that itself has not changed. Stored on every note_embeddings row so the
+# startup backfill re-embeds exactly the stale notes — a version comparison
+# instead of the table wipe migrations 0067/0077 had to do.
+#
+# Stated that way rather than as "chunk_document's output for the same input",
+# which is what it used to say: `chunk_document` is only the last step, and
+# version 2 moves without touching it. A task's document now carries its work
+# logs (#4251), so every task that has one embeds differently than it did while
+# its title, body and the chunker are all untouched — exactly the case the
+# narrower wording would have read as "nothing to re-embed".
+#
+#   1 → 2: work logs joined the task document.
+CHUNKER_VERSION = 2
 
 
 # The public name of the space every score lives in, and the two facts that
@@ -483,6 +493,47 @@ async def _claim_parent_row(session, id_column, row_id: int, label: str) -> bool
     return True
 
 
+async def _work_log_sections(note_id: int) -> list[tuple[object, str | None]]:
+    """A task's work logs, oldest first, for its embedded document (#4251).
+
+    Reads the table directly rather than through `task_logs.logs_for_task`,
+    because that function asks a PERMISSION question — may this user read this
+    task — and there is no user here. An index build acts for the record, and
+    the record's vectors carry the owner's `user_id`, so the access decision is
+    made once at search time by the clause that already scopes every hit.
+
+    That also settles what happens on a shared task: a collaborator's log is
+    part of the task's document, so it becomes findable by everyone who can
+    read the task and by nobody else. The same answer `logs_for_task` gives a
+    reader (#4241), which is the point — a log that can be read and not found
+    is the half-surface that issue was about.
+
+    Asked for every note, not only tasks. `upsert_note_embedding` is handed a
+    note_id and no kind — and the three writers that call it would each have to
+    learn to pass one — so a non-task simply has no rows and gets []. One
+    indexed lookup beside an ONNX forward pass over every chunk is not the
+    expense worth adding a parameter to three call sites for.
+
+    Returns [] on any failure. A task whose logs could not be read should embed
+    as its own prose rather than not embed at all: less findable is recoverable
+    at the next write, unindexed is not.
+    """
+    try:
+        async with async_session() as session:
+            result = await session.execute(
+                select(TaskLog.created_at, TaskLog.content)
+                .where(TaskLog.task_id == note_id)
+                .order_by(TaskLog.created_at.asc(), TaskLog.id.asc())
+            )
+            return list(result.all())
+    except Exception:
+        logger.warning(
+            "Could not read work logs for note %d; embedding its own prose only",
+            note_id, exc_info=True,
+        )
+        return []
+
+
 async def upsert_note_embedding(
     note_id: int, user_id: int, title: str | None, body: str | None
 ) -> None:
@@ -496,6 +547,7 @@ async def upsert_note_embedding(
     inserted in one transaction, so a concurrent read sees the old shape or the
     new one, never a mixture.
     """
+    title, body = task_document(title, body, await _work_log_sections(note_id))
     chunks = chunk_document(title, body)
     try:
         if not chunks:
@@ -908,6 +960,71 @@ async def backfill_note_embeddings() -> None:
 
 # ── Rules (milestone 307, note 3026) ────────────────────────────────────
 
+# The heading a work log gets inside its task's embedded document. A CONSTANT
+# because it is load-bearing twice over: `_split_sections` splits on it, so it
+# is what keeps a log from being merged into the task's own prose, and it is
+# what a reader sees at the top of a matched passage — "this is a log entry,
+# not the task's description". Changing it changes the chunk boundaries of
+# every task that has one, which is a CHUNKER_VERSION move.
+WORK_LOG_HEADING = "## Work log"
+
+
+def task_document(
+    title: str | None,
+    body: str | None,
+    logs: "Sequence[tuple[object, str | None]]" = (),
+) -> tuple[str | None, str | None]:
+    """The (title, body) a TASK is EMBEDDED as — its prose plus its work logs.
+
+    A synthesised embed-time shape, like `rule_document` and unlike a lesson:
+    the stored record is the task row, and the logs live in their own table, so
+    the document that should be searchable exists nowhere until it is built
+    here (#4251).
+
+    WHY THE LOGS BELONG IN THE TASK'S DOCUMENT rather than in rows of their
+    own. "Has anyone tried this before?" is answered by a log and asked of a
+    task — a hit on a bare log would have to be resolved back to its task to be
+    worth anything, so the useful result is the task either way. The objection
+    to folding them in is that a long log drowns a short title, and that was
+    true before #280: one vector per record meant a 2,000-word log averaged the
+    task's own subject away, and everything past ~400 words was truncated
+    unread. Chunking removed both. Each log becomes its own section, each
+    section its own title-anchored vector, each scored separately — so a task
+    is as findable as its best-matching log, and the task's own prose keeps the
+    chunk it always had.
+
+    That the result is LEGIBLE is the other half, and it is this session's
+    other build: a search now hands back the chunk that won (#4243), so a hit
+    earned by a log shows that log's passage under the task's title. Without it
+    the reader would get `body[:240]` of the task — the opening of a record
+    whose relevance lives three hundred lines further down.
+
+    Ordering is oldest-first, matching how the web renders the narrative. Only
+    the heading date distinguishes the sections, so it is part of the shape:
+    "when was this tried" is half of what a log answers.
+
+    An entry with no content is skipped rather than emitted as a bare heading —
+    an empty section is a vector with nothing in it but the task's title, which
+    competes with the task's real chunk and says nothing.
+    """
+    sections = []
+    for created_at, content in logs:
+        text = (content or "").strip()
+        if not text:
+            continue
+        stamp = getattr(created_at, "date", None)
+        heading = (
+            f"{WORK_LOG_HEADING} — {stamp()}" if callable(stamp)
+            else WORK_LOG_HEADING
+        )
+        sections.append(f"{heading}\n\n{text}")
+    if not sections:
+        return title, body
+    prose = (body or "").strip()
+    joined = "\n\n".join(sections)
+    return title, f"{prose}\n\n{joined}" if prose else joined
+
+
 def rule_document(
     title: str | None, statement: str | None, when_to_apply: str | None,
 ) -> tuple[str | None, str | None]:

src/scribe/services/task_logs.py

@@ -14,6 +14,36 @@ logger = logging.getLogger(__name__)
 _UNSET = object()
 
 
+async def _refresh_task_document(session, task_id: int) -> None:
+    """Re-embed the task whose work log just changed (#4251).
+
+    A task's embedded document carries its logs, so a log written and not
+    indexed is #4241's half-surface wearing different clothes: the entry is
+    readable and unfindable, and the next session rebuilds what this one ruled
+    out. Create, edit and delete all go through here — an edited log that keeps
+    matching its old wording is the stale-vector problem `upsert_note_embedding`
+    already refuses to leave behind for a body.
+
+    Loads the NOTE rather than synthesising one. `embed_note` reads
+    `title`, `body` and the OWNER's `user_id` off what it is handed, so a
+    stand-in row would index the logs under an empty title — throwing away the
+    per-chunk topical anchor that makes any of this discriminative — and file
+    the vectors under the wrong user.
+
+    Failure is swallowed the way `embed_note`'s own is: a log that saved must
+    not fail on its index refresh. The startup backfill is the backstop.
+    """
+    from scribe.services.notes import embed_note
+
+    try:
+        result = await session.execute(select(Note).where(Note.id == task_id))
+        note = result.scalars().first()
+        if note is not None:
+            embed_note(note)
+    except Exception:  # noqa: BLE001 - indexing never breaks a write
+        logger.exception("embedding refresh failed for task %s", task_id)
+
+
 async def create_log(
     user_id: int,
     task_id: int,
@@ -36,6 +66,7 @@ async def create_log(
         session.add(log)
         await session.commit()
         await session.refresh(log)
+        await _refresh_task_document(session, task_id)
         return log
 
 
@@ -144,6 +175,7 @@ async def update_log(
         log.updated_at = datetime.now(timezone.utc)
         await session.commit()
         await session.refresh(log)
+        await _refresh_task_document(session, log.task_id)
         return log
 
 
@@ -155,6 +187,8 @@ async def delete_log(user_id: int, log_id: int) -> bool:
         log = result.scalars().first()
         if log is None:
             return False
+        task_id = log.task_id
         await session.delete(log)
         await session.commit()
+        await _refresh_task_document(session, task_id)
         return True

tests/helpers.py

@@ -89,6 +89,24 @@ def make_mock_session() -> AsyncMock:
     return s
 
 
+def session_returning(note):
+    """``make_mock_session`` whose every ``execute()`` yields `note` (or None).
+
+    The commonest shape above the bare session: a service that loads one
+    record by id and acts on it. Two files had spelled this out identically
+    before a third was about to (#3207 — derive before the third copy), and
+    it belongs beside `make_mock_session` for the same reason that one does.
+
+    For a service making SEVERAL different reads, set
+    ``session.execute.side_effect`` to a list of results instead.
+    """
+    session = make_mock_session()
+    result = MagicMock()
+    result.scalars.return_value.first.return_value = note
+    session.execute = AsyncMock(return_value=result)
+    return session
+
+
 async def ensure_user(session, username: str, role: str = "user"):
     """Get-or-create a User by username inside an open session (flushed, not
     committed).

tests/test_chunking.py

@@ -185,11 +185,17 @@ async def test_search_collapses_chunk_rows_to_best_chunk_per_note():
 # --- the write path: one row per chunk (#280 step 3) -------------------------
 
 
-def _session_ctx():
+def _session_ctx(log_rows=()):
+    """A session stand-in. `log_rows` answers the work-log read a task's
+    document now needs (#4251) — answered explicitly rather than left to
+    autovivify, because `list(result.all())` on a bare MagicMock raises and is
+    swallowed, which would quietly make every one of these a no-log test."""
     from unittest.mock import AsyncMock, MagicMock
 
     session = MagicMock()
-    session.execute = AsyncMock()
+    result = MagicMock()
+    result.all.return_value = list(log_rows)
+    session.execute = AsyncMock(return_value=result)
     session.commit = AsyncMock()
     ctx = MagicMock()
     ctx.__aenter__ = AsyncMock(return_value=session)
@@ -226,6 +232,32 @@ async def test_upsert_stores_one_versioned_row_per_chunk():
     session.execute.assert_awaited()  # the delete that makes replacement atomic
 
 
+async def test_a_tasks_work_logs_reach_the_rows_it_is_stored_as():
+    """The wiring half of #4251: the shaper is pure and tested next door, so
+    what is pinned here is that the WRITER actually asks for the logs and
+    embeds what comes back. A log that is written, readable (#4241) and absent
+    from the index is the same half-surface one layer down."""
+    import datetime
+    from unittest.mock import AsyncMock, patch
+
+    from scribe.services import embeddings as emb
+
+    session, ctx = _session_ctx(
+        log_rows=[(datetime.datetime(2026, 9, 20), "ruled out the cache theory")]
+    )
+    with (
+        patch.object(emb, "async_session", return_value=ctx),
+        patch.object(emb, "get_embeddings", AsyncMock(return_value=[[0.0] * 384])),
+    ):
+        await emb.upsert_note_embedding(7, 42, "A task", "Its own prose.")
+
+    rows = [call.args[0] for call in session.add.call_args_list]
+    stored = "\n".join(r.chunk_text for r in rows)
+    assert "ruled out the cache theory" in stored
+    assert "Its own prose." in stored
+    assert emb.WORK_LOG_HEADING in stored
+
+
 async def test_upsert_of_an_emptied_record_clears_rows_instead_of_embedding():
     """An empty embedding is worse than none, and a STALE one is worse than
     that — a record emptied of content must stop being findable by what it no

tests/test_embedding_yields_to_a_delete.py

@@ -32,6 +32,11 @@ def _mock_session(lock_result: object = 7, execute_side_effect=None):
     session = MagicMock()
     claimed = MagicMock()
     claimed.scalar_one_or_none.return_value = lock_result
+    # A task's document carries its work logs (#4251), so the refresh reads
+    # task_logs before it builds chunks. Answered explicitly — left to
+    # autovivify, `list(result.all())` would raise and be swallowed, and these
+    # tests would be exercising the failure path without saying so.
+    claimed.all.return_value = []
     if execute_side_effect is not None:
         session.execute = AsyncMock(side_effect=execute_side_effect)
     else:
@@ -65,10 +70,19 @@ async def test_a_note_refresh_claims_the_row_before_rewriting_its_vectors():
     ):
         await emb.upsert_note_embedding(7, 42, "T", "a short body")
 
-    claim, replace = [c.args[0] for c in session.execute.call_args_list][:2]
-    assert compiled_sql(claim, dialect=PG).startswith("SELECT notes.id")
-    assert "FOR KEY SHARE NOWAIT" in compiled_sql(claim, dialect=PG)
-    assert compiled_sql(replace, dialect=PG).startswith("DELETE FROM note_embeddings")
+    sqls = [compiled_sql(c.args[0], dialect=PG) for c in session.execute.call_args_list]
+    # Located by what they ARE rather than by position: the work-log read a
+    # task's document needs (#4251) runs before any of this, and an index is
+    # not what the claim is about.
+    claim = next(i for i, sql in enumerate(sqls) if "FOR KEY SHARE NOWAIT" in sql)
+    replace = next(
+        i for i, sql in enumerate(sqls) if sql.startswith("DELETE FROM note_embeddings")
+    )
+    assert sqls[claim].startswith("SELECT notes.id")
+    assert claim < replace, "the claim goes first — that is the whole fix"
+    assert not any("note_embeddings" in sql for sql in sqls[:claim]), (
+        "nothing may touch the chunk rows before the parent is claimed"
+    )
     session.add.assert_called()
 
 
@@ -113,6 +127,8 @@ async def test_a_record_already_gone_is_not_re_embedded():
     ):
         await emb.upsert_note_embedding(7, 42, "T", "a short body")
 
-    assert session.execute.await_count == 1
+    sqls = [compiled_sql(c.args[0], dialect=PG) for c in session.execute.call_args_list]
+    assert any("FOR KEY SHARE NOWAIT" in sql for sql in sqls), "it did reach the claim"
+    assert not any("note_embeddings" in sql for sql in sqls), "and stopped there"
     session.add.assert_not_called()
     session.commit.assert_not_awaited()

tests/test_services_dedup.py

@@ -14,23 +14,14 @@ from scribe.services.dedup import (
     plan_candidate_text,
     plan_match_response,
 )
-from tests.helpers import fake_note, make_mock_session
-
-
-def _session_returning(note):
-    """A mocked async_session() whose single execute() yields `note` (or None)."""
-    s = make_mock_session()
-    result = MagicMock()
-    result.scalars.return_value.first.return_value = note
-    s.execute = AsyncMock(return_value=result)
-    return s
+from tests.helpers import fake_note, make_mock_session, session_returning
 
 
 @pytest.mark.asyncio
 async def test_title_exact_match_returns_title_duplicate():
     note = fake_note(id=10, title="Setup CI")
     with patch("scribe.services.dedup.async_session",
-               return_value=_session_returning(note)):
+               return_value=session_returning(note)):
         # whitespace/case differences are normalized away
         dup = await find_duplicate_note(7, "  setup   ci ", project_id=2, is_task=True)
     assert dup is not None
@@ -43,7 +34,7 @@ async def test_title_exact_match_returns_title_duplicate():
 async def test_short_body_skips_semantic_check():
     sem = AsyncMock()
     with patch("scribe.services.dedup.async_session",
-               return_value=_session_returning(None)), \
+               return_value=session_returning(None)), \
          patch("scribe.services.dedup.embeddings_svc.semantic_search_notes", sem):
         dup = await find_duplicate_note(7, "Unique", body="too short", project_id=2)
     assert dup is None
@@ -55,7 +46,7 @@ async def test_semantic_match_when_body_substantial():
     hit = fake_note(id=20, title="Existing", note_type="note")
     sem = AsyncMock(return_value=[(0.93, hit)])
     with patch("scribe.services.dedup.async_session",
-               return_value=_session_returning(None)), \
+               return_value=session_returning(None)), \
          patch("scribe.services.dedup.embeddings_svc.semantic_search_notes", sem):
         dup = await find_duplicate_note(
             7, "Title", body="x" * 250, project_id=2, is_task=False, note_type="note",
@@ -84,7 +75,7 @@ async def test_gate_catches_a_duplicate_hiding_in_a_later_chunk():
     # Every chunk misses except the LAST one the gate will ask about.
     sem = AsyncMock(side_effect=[[] for _ in range(n_chunks - 1)] + [[(0.94, hit)]])
     with patch("scribe.services.dedup.async_session",
-               return_value=_session_returning(None)), \
+               return_value=session_returning(None)), \
          patch("scribe.services.dedup.embeddings_svc.semantic_search_notes", sem):
         dup = await find_duplicate_note(
             7, "Title", body=body, project_id=2, is_task=False, note_type="note",
@@ -98,7 +89,7 @@ async def test_semantic_match_of_other_note_type_is_ignored():
     other = fake_note(id=21, title="X", note_type="process")
     sem = AsyncMock(return_value=[(0.97, other)])
     with patch("scribe.services.dedup.async_session",
-               return_value=_session_returning(None)), \
+               return_value=session_returning(None)), \
          patch("scribe.services.dedup.embeddings_svc.semantic_search_notes", sem):
         dup = await find_duplicate_note(7, "Title", body="x" * 250, note_type="note")
     assert dup is None  # type mismatch must not block
@@ -108,7 +99,7 @@ async def test_semantic_match_of_other_note_type_is_ignored():
 async def test_rule_title_match_in_topic():
     rule = fake_note(id=47, title="Honor the multi-user sharing ACL")
     with patch("scribe.services.dedup.async_session",
-               return_value=_session_returning(rule)):
+               return_value=session_returning(rule)):
         dup = await find_duplicate_rule(
             "honor the   multi-user sharing acl", topic_id=7,
         )
@@ -361,7 +352,7 @@ async def test_plan_title_match_short_circuits_the_semantic_arm():
     ms = MagicMock(id=415, title="Plan gate")
     sem = AsyncMock()
     with patch("scribe.services.dedup.can_read_project", AsyncMock(return_value=True)), \
-         patch("scribe.services.dedup.async_session", return_value=_session_returning(ms)), \
+         patch("scribe.services.dedup.async_session", return_value=session_returning(ms)), \
          patch("scribe.services.dedup.embeddings_svc.semantic_search_milestones", sem):
         dup = await find_matching_plan(7, 2, " plan  GATE", "x" * 300)
     assert (dup.id, dup.reason, dup.similarity) == (415, "title", 1.0)
@@ -373,7 +364,7 @@ async def test_plan_semantic_arm_asks_for_active_plans_in_the_project_at_the_set
     ms = MagicMock(id=9, title="Metadata")
     sem = AsyncMock(return_value=[(0.912345, ms)])
     with patch("scribe.services.dedup.can_read_project", AsyncMock(return_value=True)), \
-         patch("scribe.services.dedup.async_session", return_value=_session_returning(None)), \
+         patch("scribe.services.dedup.async_session", return_value=session_returning(None)), \
          patch("scribe.services.dedup.embeddings_svc.semantic_search_milestones", sem), \
          patch("scribe.services.settings.get_setting", AsyncMock(return_value="0.8")):
         dup = await find_matching_plan(7, 2, "Book metadata", "x" * 300)
@@ -396,7 +387,7 @@ async def test_plan_gate_fails_open():
 @pytest.mark.asyncio
 async def test_plan_gate_says_nothing_about_a_project_the_caller_cannot_read():
     ms = MagicMock(id=415, title="Their plan")
-    session = MagicMock(return_value=_session_returning(ms))
+    session = MagicMock(return_value=session_returning(ms))
     with patch("scribe.services.dedup.can_read_project", AsyncMock(return_value=False)), \
          patch("scribe.services.dedup.async_session", session):
         assert await find_matching_plan(8, 2, "Their plan") is None

tests/test_task_document_shape.py

@@ -0,0 +1,200 @@
+"""task_document — the document a TASK is embedded as, work logs included (#4251).
+
+A work log is the richest record Scribe holds of *why* something is the way it
+is: prose written during the work, saying what was tried and ruled out. #4241
+made it readable. It was still not findable, so "has anyone tried this
+approach?" — the question a log answers — could not reach one, and #4208 was
+rebuilt in this very session because its logs were unreachable.
+
+These tests pin the shape and the two properties the decision rested on: a task
+without logs embeds EXACTLY as it always did, and a task with them stays as
+discriminative as it was, because each log is its own title-anchored chunk
+rather than prose averaged into one vector.
+"""
+import datetime
+
+from scribe.services.embeddings import (
+    WORK_LOG_HEADING,
+    chunk_document,
+    embedding_text,
+    task_document,
+)
+
+D1 = datetime.datetime(2026, 9, 20, 10, 0)
+D2 = datetime.datetime(2026, 9, 21, 11, 0)
+
+
+# --- the identity half: a task with no logs is untouched ---------------------
+
+
+def test_a_task_with_no_logs_embeds_exactly_as_it_did_before():
+    """Most notes are not tasks and most tasks carry no log. Their vectors are
+    the corpus this instance's thresholds were measured against, and changing
+    them for nothing would move every tuned number underneath itself (#4225)."""
+    assert task_document("A title", "A body") == ("A title", "A body")
+    assert task_document("A title", "A body", []) == ("A title", "A body")
+    assert task_document("T", None) == ("T", None)
+
+
+def test_an_entry_with_no_content_is_skipped_rather_than_emitted_empty():
+    """A bare heading is a vector containing nothing but the task's title — it
+    would compete with the task's real chunk and say nothing."""
+    _t, body = task_document("T", "prose", [(D1, "   "), (D2, None), (D1, "real")])
+    assert body.count(WORK_LOG_HEADING) == 1
+    assert "real" in body
+
+
+def test_a_task_that_is_only_logs_still_yields_a_document():
+    """A task opened with a title and filled in entirely through its log — the
+    shape every step of a milestone starts as."""
+    title, body = task_document("T", "", [(D1, "the only content")])
+    assert title == "T"
+    assert body.startswith(WORK_LOG_HEADING)
+    assert "the only content" in body
+
+
+# --- the shape: the logs are sections, oldest first --------------------------
+
+
+def test_logs_are_appended_as_dated_sections_after_the_tasks_own_prose():
+    title, body = task_document(
+        "Fix the thing", "The task prose.",
+        [(D1, "Tried X, ruled out."), (D2, "Y worked.")],
+    )
+    assert title == "Fix the thing"
+    assert body.index("The task prose.") < body.index("Tried X") < body.index("Y worked.")
+    assert f"{WORK_LOG_HEADING} — 2026-09-20" in body
+    assert f"{WORK_LOG_HEADING} — 2026-09-21" in body
+
+
+def test_an_entry_with_no_timestamp_still_gets_its_own_section():
+    """Degrades to an undated heading rather than dropping the entry or
+    crashing — a restored row or a hand-built one is not a reason to lose the
+    richest prose in the record."""
+    _t, body = task_document("T", "p", [(None, "content from nowhere")])
+    assert WORK_LOG_HEADING in body
+    assert "content from nowhere" in body
+
+
+# --- why folding them in is safe: chunking, not averaging --------------------
+
+
+def test_each_log_becomes_its_own_title_anchored_chunk():
+    """THE decision this build turns on. The objection to putting logs in the
+    task's document is that a long log drowns a short title — true before #280,
+    when one vector per record meant a 2,000-word log averaged the task's
+    subject away and everything past ~400 words was truncated unread. Chunking
+    answers both: separate vectors, each carrying the title as its anchor."""
+    def _entry(subject: str) -> str:
+        para = " ".join([f"This log entry discusses the {subject} in detail."] * 8)
+        return "\n\n".join(f"{para} (p{i})" for i in range(6))
+
+    title, body = task_document(
+        "Short task title", "Short body.",
+        [(D1, _entry("approach")), (D2, _entry("alternative"))],
+    )
+    chunks = chunk_document(title, body)
+    assert len(chunks) > 1, "a long log must split rather than truncate"
+    for chunk in chunks:
+        assert chunk.startswith("Short task title\n")
+
+    # The task's own prose still has a chunk of its own — it is not merged into
+    # a log section and scored as part of it.
+    assert any("Short body." in c and "alternative" not in c for c in chunks)
+
+    # And nothing is lost: this is what #280 exists for. Paragraph-shaped,
+    # because that is what a log is and because a single unbroken 3,000-char
+    # line is hard-split mid-line by design — `test_chunking` owns that case.
+    joined = "\n".join(chunks)
+    for line in body.splitlines():
+        if line.strip():
+            assert line.strip() in joined, f"content dropped: {line[:60]!r}"
+
+
+def test_a_short_task_with_a_short_log_is_still_one_sharp_chunk():
+    """Over-splitting is not the goal either. A task and one short log stay a
+    single document, the shape note #2485 measured as the sharpest."""
+    title, body = task_document("T", "body", [(D1, "a short note about it")])
+    assert chunk_document(title, body) == [embedding_text(title, body)]
+
+
+# --- keeping it current: a log write refreshes the task's vectors ------------
+#
+# The shaper above is pure. These pin that the writers actually call it — a
+# task whose logs are in the document but never re-indexed after one is written
+# is #4241's half-surface one layer down: the entry is readable, and the search
+# still answers as though it were never written.
+
+from unittest.mock import MagicMock, patch          # noqa: E402
+
+import pytest                                       # noqa: E402
+
+from scribe.services import task_logs as svc        # noqa: E402
+from tests.helpers import (                      # noqa: E402
+    fake_note, make_mock_session, session_returning,
+)
+
+
+@pytest.mark.asyncio
+async def test_writing_a_work_log_refreshes_the_tasks_embedding():
+    note = fake_note(id=7, title="A task", body="prose", is_task=True)
+    session = session_returning(note)
+    with (
+        patch.object(svc, "async_session", return_value=session),
+        patch("scribe.services.notes.embed_note") as embed,
+    ):
+        await svc.create_log(42, 7, "what I tried")
+    embed.assert_called_once()
+    assert embed.call_args.args[0] is note
+
+
+@pytest.mark.asyncio
+async def test_editing_a_work_log_refreshes_it_too():
+    """An edited log whose vectors still carry the old wording keeps matching
+    what it no longer says — the stale-vector case `upsert_note_embedding`
+    already refuses to leave behind for a body."""
+    note = fake_note(id=7, title="A task", body="prose", is_task=True)
+    log = MagicMock(id=3, task_id=7)
+    session = make_mock_session()
+    found_log, found_note = MagicMock(), MagicMock()
+    found_log.scalars.return_value.first.return_value = log
+    found_note.scalars.return_value.first.return_value = note
+    session.execute.side_effect = [found_log, found_note]
+    with (
+        patch.object(svc, "async_session", return_value=session),
+        patch("scribe.services.notes.embed_note") as embed,
+    ):
+        await svc.update_log(42, 3, content="reworded")
+    embed.assert_called_once_with(note)
+
+
+@pytest.mark.asyncio
+async def test_deleting_a_work_log_refreshes_it_too():
+    """And reads the task id BEFORE the row goes — after the delete there is
+    nothing left to ask which task it belonged to."""
+    note = fake_note(id=7, title="A task", body="prose", is_task=True)
+    log = MagicMock(id=3, task_id=7)
+    session = make_mock_session()
+    found_log, found_note = MagicMock(), MagicMock()
+    found_log.scalars.return_value.first.return_value = log
+    found_note.scalars.return_value.first.return_value = note
+    session.execute.side_effect = [found_log, found_note]
+    with (
+        patch.object(svc, "async_session", return_value=session),
+        patch("scribe.services.notes.embed_note") as embed,
+    ):
+        assert await svc.delete_log(42, 3) is True
+    embed.assert_called_once_with(note)
+
+
+@pytest.mark.asyncio
+async def test_a_failed_refresh_does_not_fail_the_log_that_saved():
+    """Indexing never breaks a write. The startup backfill is the backstop."""
+    session = session_returning(fake_note(id=7, title="T", body="b", is_task=True))
+    with (
+        patch.object(svc, "async_session", return_value=session),
+        patch("scribe.services.notes.embed_note", side_effect=RuntimeError("boom")),
+    ):
+        log = await svc.create_log(42, 7, "what I tried")
+    assert log is not None
+    session.commit.assert_awaited()