ci(ananke): publish per-test case metrics for flaky tracking

scripts/publish_quality_metrics.py

@@ -87,6 +87,7 @@ def _build_payload(
     tests_failed: int,
     tests_errors: int,
     tests_skipped: int,
+    test_cases: list[tuple[str, str]],
     coverage_percent: float,
     source_lines_over_500: int,
     checks: dict[str, str],
@@ -114,6 +115,12 @@ def _build_payload(
         f'ananke_quality_gate_checks_total{{suite="{suite}",check="{check_name}",result="{check_status}"}} 1'
         for check_name, check_status in checks.items()
     )
+    if test_cases:
+        lines.append("# TYPE platform_quality_gate_test_case_result gauge")
+        lines.extend(
+            f'platform_quality_gate_test_case_result{{suite="{suite}",test="{_escape_label(test_name)}",status="{_escape_label(test_status)}"}} 1'
+            for test_name, test_status in test_cases
+        )
     return "\n".join(lines) + "\n"
 
 
@@ -159,6 +166,25 @@ def _parse_go_test_counts(output_path: Path) -> dict[str, int]:
     }
 
 
+def _parse_go_test_cases(output_path: Path) -> list[tuple[str, str]]:
+    """Parse per-test status records from go test output text."""
+    if not output_path.exists():
+        return []
+    text = output_path.read_text(encoding="utf-8", errors="ignore")
+    cases: list[tuple[str, str]] = []
+    patterns = {
+        "passed": re.compile(r"^--- PASS: ([^\s(]+)", flags=re.M),
+        "failed": re.compile(r"^--- FAIL: ([^\s(]+)", flags=re.M),
+        "skipped": re.compile(r"^--- SKIP: ([^\s(]+)", flags=re.M),
+    }
+    for status, pattern in patterns.items():
+        for test_name in pattern.findall(text):
+            cleaned = str(test_name).strip()
+            if cleaned:
+                cases.append((cleaned, status))
+    return cases
+
+
 def _read_exit_code(path: Path) -> int:
     if not path.exists():
         return 1
@@ -288,7 +314,9 @@ def main(argv: list[str] | None = None) -> int:
     resolved_failed = max(args.local_failed, remote_failed)
     coverage_percent = _read_coverage_percent(args.coverage_percent_file)
     source_lines_over_500 = _count_source_files_over_limit(repo_root, max_lines=500)
-    tests = _parse_go_test_counts(Path(os.getenv("ANANKE_QUALITY_OUTPUT_FILE", str(build_dir / "quality-gate.out"))))
+    quality_output = Path(os.getenv("ANANKE_QUALITY_OUTPUT_FILE", str(build_dir / "quality-gate.out")))
+    tests = _parse_go_test_counts(quality_output)
+    test_cases = _parse_go_test_cases(quality_output)
     gate_rc = _read_exit_code(Path(os.getenv("ANANKE_QUALITY_EXIT_CODE_PATH", str(build_dir / "quality-gate.rc"))))
     docs_status = _read_status(Path(os.getenv("ANANKE_QUALITY_DOCS_STATUS_PATH", str(build_dir / "docs-naming.status"))))
     gate_failed = gate_rc != 0
@@ -310,6 +338,7 @@ def main(argv: list[str] | None = None) -> int:
         tests_failed=tests["failed"],
         tests_errors=tests["errors"],
         tests_skipped=tests["skipped"],
+        test_cases=test_cases,
         coverage_percent=coverage_percent,
         source_lines_over_500=source_lines_over_500,
         checks=checks,