fix(cleanup): library scans survive navigation, reconnect on return

The transparency / single-color audit cards held the run + poll timer in
local component state, so navigating away destroyed both and onMounted never
reconnected — the Celery scan kept running and writing LibraryAuditRun, but
the UI forgot it. Now each card, on mount, fetches its rule's latest run
(GET /api/cleanup/audit?rule=<rule>&limit=1) and rehydrates: shows progress +
resumes polling if still running, or shows the completed result (ready/applied/
error) so the operator can act on it after returning. Adds the ?rule= filter
to the audit-history endpoint + cleanup store latestAuditForRule().

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

tests/test_api_cleanup.py

@@ -155,6 +155,27 @@ async def test_audit_history_returns_recent_runs(client, db):
     assert len(body["runs"]) >= 3
 
 
+@pytest.mark.asyncio
+async def test_audit_history_filters_by_rule(client, db):
+    db.add(LibraryAuditRun(
+        rule="transparency", params={"threshold": 0.9},
+        status="applied", matched_ids=[], finished_at=datetime.now(UTC),
+    ))
+    db.add(LibraryAuditRun(
+        rule="single_color", params={"threshold": 0.95, "tolerance": 30},
+        status="ready", matched_count=2, matched_ids=[1, 2],
+        finished_at=datetime.now(UTC),
+    ))
+    await db.commit()
+    # ?rule=&limit=1 → just THIS rule's latest run (the card-reconnect query).
+    resp = await client.get("/api/cleanup/audit?rule=single_color&limit=1")
+    assert resp.status_code == 200
+    runs = (await resp.get_json())["runs"]
+    assert len(runs) == 1
+    assert runs[0]["rule"] == "single_color"
+    assert runs[0]["matched_count"] == 2
+
+
 @pytest.mark.asyncio
 async def test_audit_apply_with_token_deletes(client, db, tmp_path):
     rec = await _seed_image(db, tmp_path, w=100, h=100, name="apply.png")