fix(ui): correct usage bar color logic

src/kimi_cli/ui/shell/usage.py

@@ -250,15 +250,21 @@ def _build_usage_panel(summary: UsageRow | None, limits: list[UsageRow]) -> Pane
 
 
 def _format_row(row: UsageRow, label_width: int) -> RenderableType:
-    ratio = (row.limit - row.used) / row.limit if row.limit > 0 else 0
-    color = _ratio_color(ratio)
+    if row.limit > 0:
+        usage_ratio = row.used / row.limit
+        color = _ratio_color(usage_ratio)
+        percent = (1 - usage_ratio) * 100
+        percent_text = f"{percent:.0f}% left"
+    else:
+        usage_ratio = 0
+        color = "grey50"
+        percent_text = "N/A"
 
     label = Text(f"{row.label:<{label_width}}  ", style="cyan")
     bar = ProgressBar(total=row.limit or 1, completed=row.used, width=20, complete_style=color)
 
     detail = Text()
-    percent = ratio * 100
-    detail.append(f"  {percent:.0f}% left", style="bold")
+    detail.append(f"  {percent_text}", style="bold")
     if row.reset_hint:
         detail.append(f"  ({row.reset_hint})", style="grey50")
 

tests/ui_and_conv/test_usage.py

@@ -0,0 +1,74 @@
+"""Tests for usage display color logic."""
+
+from unittest.mock import patch
+
+import pytest
+
+from kimi_cli.ui.shell.usage import _format_row, _ratio_color, UsageRow
+
+
+class TestRatioColor:
+    """Tests for _ratio_color function."""
+
+    @pytest.mark.parametrize(
+        "ratio,expected,description",
+        [
+            # Low usage -> green
+            (0.0, "green", "0% used"),
+            (0.07, "green", "7% used, plenty left"),
+            (0.5, "green", "50% used, still ok"),
+            (0.69, "green", "69% used, boundary"),
+            # Medium usage -> yellow
+            (0.7, "yellow", "70% used, warning zone"),
+            (0.75, "yellow", "75% used"),
+            (0.89, "yellow", "89% used, boundary"),
+            # High usage -> red
+            (0.9, "red", "90% used, danger zone"),
+            (0.95, "red", "95% used"),
+            (1.0, "red", "100% used, exhausted"),
+        ],
+    )
+    def test_ratio_color(self, ratio: float, expected: str, description: str):
+        """Test color for various usage ratios."""
+        assert _ratio_color(ratio) == expected, f"Failed for {description}"
+
+
+class TestFormatRow:
+    """Tests for _format_row function."""
+
+    @pytest.mark.parametrize(
+        "used,limit,expected_ratio",
+        [
+            (0, 100, 0.0),     # 0% used -> boundary (green)
+            (7, 100, 0.07),    # 7% used -> low usage (green)
+            (50, 100, 0.5),    # 50% used -> low usage (green)
+            (69, 100, 0.69),   # 69% used -> boundary (green)
+            (70, 100, 0.7),    # 70% used -> boundary (yellow)
+            (75, 100, 0.75),   # 75% used -> medium usage (yellow)
+            (89, 100, 0.89),   # 89% used -> boundary (yellow)
+            (90, 100, 0.9),    # 90% used -> boundary (red)
+            (95, 100, 0.95),   # 95% used -> high usage (red)
+            (100, 100, 1.0),   # 100% used -> boundary (red)
+        ],
+    )
+    def test_format_row_passes_usage_ratio(self, used: int, limit: int, expected_ratio: float):
+        """Test that _format_row passes usage ratio (used/limit) to _ratio_color."""
+        row = UsageRow(label="Test", used=used, limit=limit)
+
+        with patch("kimi_cli.ui.shell.usage._ratio_color") as mock_ratio_color:
+            mock_ratio_color.return_value = "green"
+            _format_row(row, label_width=10)
+
+            mock_ratio_color.assert_called_once()
+            actual_ratio = mock_ratio_color.call_args[0][0]
+            assert abs(actual_ratio - expected_ratio) < 0.001, (
+                f"Expected usage ratio {expected_ratio}, got {actual_ratio}"
+            )
+
+    def test_format_row_zero_limit(self):
+        """Test _format_row handles zero limit gracefully."""
+        row = UsageRow(label="Test", used=0, limit=0)
+
+        # Should not raise and should display "N/A"
+        result = _format_row(row, label_width=10)
+        assert result is not None