test: correct mislabelled test in src/audit/analyzers/jobs.rs (#883)

timeline_to_jobs_parses_real_ado_shape claimed to validate "real ADO
shape" parsing but only asserted on jobs.len() and name ordering — both
already covered by the filtering and sorting tests. The test setup
provides complete job data (result, startTime, finishTime, id, parentId)
but none of those fields were verified, meaning regressions in field
parsing would go undetected.

Add assertions for status, result, duration, started_at, and finished_at
on the first parsed job, and add a comment clarifying that extra ADO
fields (id, parentId) are silently ignored.

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Author: github-actions[bot]

src/audit/analyzers/jobs.rs

@@ -237,6 +237,8 @@ mod tests {
 
     #[test]
     fn timeline_to_jobs_parses_real_ado_shape() {
+        // Simulates a real ADO timeline response with extra fields (id, parentId)
+        // that the parser should silently ignore.
         let timeline = json!({
             "records": [
                 {
@@ -279,5 +281,13 @@ mod tests {
             jobs.iter().map(|job| job.name.as_str()).collect::<Vec<_>>(),
             vec!["Agent", "Detection", "SafeOutputs"]
         );
+
+        // Verify all fields are correctly parsed for the first job; extra ADO
+        // fields (id, parentId) must be silently ignored.
+        assert_eq!(jobs[0].status, "completed");
+        assert_eq!(jobs[0].result.as_deref(), Some("succeeded"));
+        assert_eq!(jobs[0].duration.as_deref(), Some("1m 0s"));
+        assert_eq!(jobs[0].started_at.as_deref(), Some("2026-01-01T00:00:00Z"));
+        assert_eq!(jobs[0].finished_at.as_deref(), Some("2026-01-01T00:01:00Z"));
     }
 }