fix: preserve reviewer results after tool failures

internal/reviewplan/summary.go

@@ -375,7 +375,7 @@ func ReviewersProducedResults(coverage []ReviewerCoverageSummary) map[string]boo
 
 func coverageResultProduced(status string) bool {
 	switch strings.TrimSpace(status) {
-	case "complete_broad", "complete_constrained", "incomplete_skipped":
+	case "complete_broad", "complete_constrained", "incomplete_skipped", "incomplete_tool":
 		return true
 	default:
 		return false

internal/reviewplan/summary_test.go

@@ -456,6 +456,32 @@ func TestRollupSummaryRendering(t *testing.T) {
 		}
 	})
 
+	t.Run("incomplete tool preserves produced results", func(t *testing.T) {
+		req := summaryRequest()
+		req.RunSummary.ReviewerCoverage = []ReviewerCoverageSummary{
+			{AgentID: "go:implementation-tests", Status: "incomplete_tool", Scope: []string{"main.go"}, InspectedFiles: []string{"main.go"}},
+			{AgentID: "policies:conventions", Status: "incomplete_tool", Scope: []string{"main.go"}, InspectedFiles: []string{"main.go"}},
+			{AgentID: "failed", Status: "incomplete_failed", Scope: []string{"main.go"}},
+		}
+		req.RunSummary.SelectedReviewers = append(req.RunSummary.SelectedReviewers, "failed")
+		plan, err := Build(req)
+		if err != nil {
+			t.Fatalf("Build: %v", err)
+		}
+		for _, want := range []string{
+			"| go:implementation-tests | 2 |",
+			"| policies:conventions | 0 |",
+			"| failed | ⚠️ did not run |",
+			"- `go:implementation-tests` — ⚠️ incomplete (tool failure); skipped: none; constraints: none\n",
+			"- `policies:conventions` — ⚠️ incomplete (tool failure); skipped: none; constraints: none\n",
+			"- `failed` — ⚠️ failed\n",
+		} {
+			if !strings.Contains(plan.RollupMarkdown, want) {
+				t.Errorf("rollup missing %q:\n%s", want, plan.RollupMarkdown)
+			}
+		}
+	})
+
 	t.Run("incomplete tool reviewer coverage force comment", func(t *testing.T) {
 		req := baseRequest()
 		req.Findings = nil

internal/view/review_reviewer_ran_test.go

@@ -18,11 +18,15 @@ func TestReviewSummaryJSONDistinguishesFailedReviewer(t *testing.T) {
 			{Name: "security:code-auditor", Findings: 0},
 			{Name: "documentation:docs", Findings: 0},
 			{Name: "policies:conventions", Findings: 0},
+			{Name: "tool-zero", Findings: 0},
+			{Name: "tool-findings", Findings: 2},
 		},
 		Run: reviewplan.RunSummary{
 			ReviewerCoverage: []reviewplan.ReviewerCoverageSummary{
 				{AgentID: "security:code-auditor", Status: "incomplete_failed"},
 				{AgentID: "documentation:docs", Status: "complete_broad"},
+				{AgentID: "tool-zero", Status: "incomplete_tool"},
+				{AgentID: "tool-findings", Status: "incomplete_tool"},
 				// policies:conventions absent: status unknown.
 			},
 		},
@@ -40,6 +44,14 @@ func TestReviewSummaryJSONDistinguishesFailedReviewer(t *testing.T) {
 	if !strings.Contains(got, `"name":"documentation:docs","findings":0,"ran":true`) {
 		t.Fatalf("completed reviewer must serialize ran:true, got:\n%s", got)
 	}
+	for _, want := range []string{
+		`"name":"tool-zero","findings":0,"ran":true`,
+		`"name":"tool-findings","findings":2,"ran":true`,
+	} {
+		if !strings.Contains(got, want) {
+			t.Errorf("tool-incomplete reviewer must preserve its result %s, got:\n%s", want, got)
+		}
+	}
 	// Unknown coverage omits the field rather than guessing a failure.
 	if !strings.Contains(got, `"name":"policies:conventions","findings":0}`) {
 		t.Fatalf("unknown coverage must omit ran, got:\n%s", got)