feat(openapi): complete upload preview and UI evidence

tests/unit/test_collect_capability_baseline.py

@@ -10,14 +10,17 @@ COLLECTOR = ROOT / "scripts" / "collect-capability-baseline.py"
 
 
 class CapabilityBaselineCollectorTests(unittest.TestCase):
-    def run_playwright(self, report: object) -> tuple[subprocess.CompletedProcess[str], Path, tempfile.TemporaryDirectory[str]]:
+    def run_playwright(self, report: object, required: list[str] | None = None) -> tuple[subprocess.CompletedProcess[str], Path, tempfile.TemporaryDirectory[str]]:
         temporary = tempfile.TemporaryDirectory()
         root = Path(temporary.name)
         report_path = root / "report.json"
         output_path = root / "candidate.json"
         report_path.write_text(json.dumps(report), encoding="utf-8")
+        command = ["python3", str(COLLECTOR), "playwright", "--report", str(report_path), "--output", str(output_path), "--source-revision", "0" * 40, "--environment-class", "community-test", "--flow-id", "ui-operations"]
+        for title in required or []:
+            command.extend(["--required-test", title])
         result = subprocess.run(
-            ["python3", str(COLLECTOR), "playwright", "--report", str(report_path), "--output", str(output_path), "--source-revision", "0" * 40, "--environment-class", "community-test", "--flow-id", "ui-operations"],
+            command,
             text=True,
             capture_output=True,
             check=False,
@@ -51,6 +54,22 @@ class CapabilityBaselineCollectorTests(unittest.TestCase):
         self.assertEqual(result.returncode, 0, result.stderr)
         self.assertEqual(json.loads(output.read_text(encoding="utf-8"))["execution_verdict"], "skipped")
 
+    def test_required_openapi_tests_fail_closed_when_missing_skipped_or_flaky(self) -> None:
+        required = ["OpenAPI required scenario"]
+        reports = [
+            {"suites": [{"specs": [{"title": "another scenario", "tests": [{"status": "expected", "results": [{"status": "passed", "retry": 0}]}]}]}]},
+            {"suites": [{"specs": [{"title": required[0], "tests": [{"status": "skipped", "results": []}]}]}]},
+            {"suites": [{"specs": [{"title": required[0], "tests": [{"status": "flaky", "results": [{"status": "failed", "retry": 0}, {"status": "passed", "retry": 1}]}]}]}]},
+        ]
+        for report in reports:
+            with self.subTest(report=report):
+                result, output, temporary = self.run_playwright(report, required)
+                self.addCleanup(temporary.cleanup)
+                self.assertEqual(result.returncode, 0, result.stderr)
+                candidate = json.loads(output.read_text(encoding="utf-8"))
+                self.assertEqual(candidate["execution_verdict"], "fail")
+                self.assertFalse(candidate["accepted"])
+
     def test_raw_report_content_never_reaches_candidate_or_error(self) -> None:
         canary = "Bearer secret-canary /home/private/workspace https://private.invalid?q=secret"
         report = {"suites": [], "errors": [{"message": canary}], "stdout": [canary]}