fix: stop naming the operator view in robots.txt

fintick/dashboard.py

@@ -148,13 +148,16 @@ def read_feed(database: str | Path, *, limit: int = DEFAULT_LIMIT) -> dict[str,
 
 SITE_ORIGIN = os.environ.get("FINTICK_SITE_ORIGIN", "https://fintick.fyi").rstrip("/")
 
-# Crawlable: the board itself. Not crawlable: the JSON API (no prose to index) and the
-# ?ops operator view, which is the same page plus telemetry and would read as duplicate
-# content. The canonical link handles ?ops for engines that ignore the query rule.
+# Crawlable: the board itself. Not crawlable: the JSON API, which carries no prose to
+# index.
+#
+# The operator view is deliberately NOT named here. robots.txt is a public file that
+# scanners fetch first, so a Disallow line advertises a path rather than protecting it.
+# Consolidating the operator view onto the board is the canonical link's job, and it
+# does that without publishing anything.
 ROBOTS_TXT = """User-agent: *
 Allow: /$
 Disallow: /api/
-Disallow: /*?ops
 
 Sitemap: {origin}/sitemap.xml
 """

tests/test_dashboard.py

@@ -264,15 +264,21 @@ def test_structured_data_parses(self) -> None:
         graph = json.loads(block)["@graph"]
         self.assertEqual({node["@type"] for node in graph}, {"WebSite", "WebApplication"})
 
-    def test_robots_points_at_sitemap_and_shields_api_and_ops(self) -> None:
+    def test_robots_points_at_sitemap_and_disallows_the_api(self) -> None:
         status, headers, body = self._get("/robots.txt")
         text = body.decode()
         self.assertEqual(status, 200)
         self.assertTrue(headers["Content-Type"].startswith("text/plain"))
         self.assertIn("Disallow: /api/", text)
-        self.assertIn("Disallow: /*?ops", text)
         self.assertIn("Sitemap: https://fintick.fyi/sitemap.xml", text)
 
+    def test_public_documents_never_name_the_operator_view(self) -> None:
+        # robots.txt and llms.txt are fetched by scanners. Naming the operator view in
+        # either advertises it; a Disallow line is a signpost, not a shield.
+        for path in ("/robots.txt", "/llms.txt", "/sitemap.xml"):
+            _, _, body = self._get(path)
+            self.assertNotIn("ops", body.decode().lower(), f"{path} names the operator view")
+
     def test_sitemap_is_well_formed_xml(self) -> None:
         import xml.etree.ElementTree as ET