diff --git a/scripts/build_featured_data.py b/scripts/build_featured_data.py
index e0b3d72cebf..f7fc92b2a56 100644
--- a/scripts/build_featured_data.py
+++ b/scripts/build_featured_data.py
@@ -100,6 +100,12 @@ def extract_doi(url: str) -> str | None:
     return None
 
 
+def clean_apa_reference(apa: str) -> str:
+    """Remove the Portico archive-provider artefact immediately before a DOI URL."""
+    # Same narrow repair as parse_clusters_to_sheet.tidy_lookup_citation (#984).
+    return re.sub(r'\. Portico\. (?=https://doi\.org/)', '. ', apa.strip())
+
+
 def make_short_ref(apa: str) -> str:
     """Generate a compact citation like 'Munafo et al. (2017)' from an APA string."""
     if not apa:
@@ -275,7 +281,7 @@ def main():
 
         oa_url = row[COL_OA_LINK].strip()
         is_oa = bool(oa_url)
-        apa = row[COL_APA].strip()
+        apa = clean_apa_reference(row[COL_APA])
 
         card = {
             "title": row[COL_TITLE].strip(),
diff --git a/scripts/tests/test_build_featured_data.py b/scripts/tests/test_build_featured_data.py
new file mode 100644
index 00000000000..575d143028c
--- /dev/null
+++ b/scripts/tests/test_build_featured_data.py
@@ -0,0 +1,71 @@
+"""Regression tests for archive-provider artefacts in featured APA references."""
+
+import contextlib
+import io
+import json
+import sys
+import tempfile
+import unittest
+from pathlib import Path
+from unittest.mock import patch
+
+sys.path.insert(0, str(Path(__file__).resolve().parents[1]))
+
+import build_featured_data as featured
+
+
+class FeaturedReferences(unittest.TestCase):
+    APA = "Doe, A. (2026). Title. Current Protocols, 6(1). Portico. https://doi.org/10.1002/cpz1.70299"
+    CLEAN_APA = "Doe, A. (2026). Title. Current Protocols, 6(1). https://doi.org/10.1002/cpz1.70299"
+
+    def test_removes_archive_provider_before_doi(self):
+        self.assertEqual(featured.clean_apa_reference(self.APA), self.CLEAN_APA)
+
+    def test_preserves_other_portico_mentions(self):
+        for apa in (
+            "Doe, A. (2026). The Portico archive. Journal. https://doi.org/10.1/x",
+            "Doe, A. (2026). Title. Portico. https://example.org/article",
+            "Doe, A. (2026). Title. Portico.",
+            self.CLEAN_APA,
+            "",
+        ):
+            with self.subTest(apa=apa):
+                self.assertEqual(featured.clean_apa_reference(apa), apa)
+
+    def test_both_outputs_use_cleaned_publication_reference(self):
+        doi = "10.1002/cpz1.70299"
+        publication = ["Sub-cluster", doi, self.APA] + [""] * 15
+        publication[6] = "Title"
+        rows = {
+            featured.RECOMMENDATIONS_CSV_URL: [["", "Recommender", "", "", "", "", doi, "Useful"]],
+            featured.PUBLICATIONS_CSV_URL: [publication],
+            featured.CLUSTERS_CSV_URL: [["Cluster"]],
+            featured.SUBCLUSTERS_CSV_URL: [["Cluster", "Sub-cluster"]],
+            featured.TAGS_CSV_URL: [],
+            featured.VOTES_CSV_URL: [],
+        }
+
+        with tempfile.TemporaryDirectory() as directory:
+            outputs = {
+                name: str(Path(directory) / filename)
+                for name, filename in (
+                    ("OUTPUT_PATH", "featured_resources.json"),
+                    ("PUB_CARDS_PATH", "pub_cards.json"),
+                    ("FILTER_TAGS_PATH", "filter_tags.json"),
+                )
+            }
+            with patch.multiple(featured, **outputs), patch.object(
+                featured, "fetch_csv", side_effect=lambda url, skip_header=False: rows[url]
+            ), contextlib.redirect_stdout(io.StringIO()):
+                featured.main()
+
+            cards = json.loads(Path(outputs["PUB_CARDS_PATH"]).read_text())
+            resources = json.loads(Path(outputs["OUTPUT_PATH"]).read_text())
+
+        self.assertEqual(cards[doi]["apa"], self.CLEAN_APA)
+        self.assertEqual(resources["clusters"]["1"][0]["apa"], self.CLEAN_APA)
+        self.assertEqual(cards[doi]["short_ref"], "Doe (2026)")
+
+
+if __name__ == "__main__":
+    unittest.main()