app/webui: replace generic "see summary.txt" SAT failure text with the real reason

Every place that surfaced a FAILED SAT result — task error messages,
component-status.json detail, and the hardware snapshot's ErrorDescription/
StatusHistory — used to say only "SAT overall_status=FAILED (see
summary.txt)" or "<label> failed", forcing an engineer to go dig through the
run directory to find out what actually broke.

nvidia-config's summary.txt now carries a "warnings" field with the specific
GPU/NVLink finding. A new SATFailureDetail/satFailureDetailFromKV in
component_status_db.go reads that field, or falls back to naming whichever
generic SAT sub-job(s) reported non-OK/UNSUPPORTED status along with their
exit code. This feeds both the task-runner error message and the component
status DB. sat_overlay.go's satKeyStatus (which drives ErrorDescription on
the exported hardware snapshot) now does the same, with storage kept
per-device so one drive's rc doesn't get attributed to another's card.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

audit/internal/platform/nvidia_config_check.go

@@ -470,6 +470,11 @@ func renderNvidiaConfigCheckSummary(status NvidiaConfigCheckStatus) string {
 		fmt.Fprintln(&b, "overall_status=OK")
 	} else {
 		fmt.Fprintln(&b, "overall_status=FAILED")
+		// warnings carries the actual reason(s) in prose so a reader of
+		// summary.txt (or anything parsing it, like sat.DescribeFailure)
+		// never has to fall back to "see the report" — each entry already
+		// names the specific GPU or NVLink pair and what's wrong with it.
+		fmt.Fprintf(&b, "warnings=%s\n", strings.Join(status.Warnings, "; "))
 	}
 	return b.String()
 }

audit/internal/platform/nvidia_config_check_test.go

@@ -166,3 +166,25 @@ func TestRenderNvidiaConfigCheckSummaryOverallStatus(t *testing.T) {
 		t.Fatalf("status with warnings missing overall_status=FAILED:\n%s", got)
 	}
 }
+
+// TestRenderNvidiaConfigCheckSummaryIncludesWarningsField guards that a
+// failed run's summary.txt carries the actual reason in prose (a "warnings"
+// field), not just a bare overall_status=FAILED — app.SATFailureDetail reads
+// this field so a failed task's error message names the specific GPU/NVLink
+// problem instead of telling the engineer to go read summary.txt themselves.
+func TestRenderNvidiaConfigCheckSummaryIncludesWarningsField(t *testing.T) {
+	status := NvidiaConfigCheckStatus{Warnings: []string{
+		"GPU 0 (H100): ECC is disabled",
+		"NVLink GPU0<->GPU1: 2/36 NVLinks inactive on a bonded pair",
+	}}
+	got := renderNvidiaConfigCheckSummary(status)
+	want := "warnings=GPU 0 (H100): ECC is disabled; NVLink GPU0<->GPU1: 2/36 NVLinks inactive on a bonded pair\n"
+	if !strings.Contains(got, want) {
+		t.Fatalf("summary missing combined warnings field:\ngot:\n%s\nwant substring:\n%s", got, want)
+	}
+
+	clean := NvidiaConfigCheckStatus{}
+	if got := renderNvidiaConfigCheckSummary(clean); strings.Contains(got, "warnings=") {
+		t.Fatalf("clean status summary should omit warnings field entirely:\n%s", got)
+	}
+}