fix(vscode): make cache hit rate write-aware

.changeset/write-aware-cache-rate.md

@@ -0,0 +1,5 @@
+---
+"kilo-code": patch
+---
+
+Show a write-aware cache hit rate in the VS Code session usage bar.

packages/kilo-vscode/tests/unit/model-usage.test.ts

@@ -1,6 +1,7 @@
 import { describe, expect, test } from "bun:test"
 import type { Provider, SessionModelUsage } from "../../webview-ui/src/types/messages"
 import {
+  cacheRate,
   groupModelUsage,
   hasModelUsage,
   isSameSessionTree,
@@ -44,6 +45,8 @@ describe("model usage", () => {
       }),
     ).toBeFalse()
     expect(tokenSummary(usage)).toEqual({ input: 10, output: 2, cached: 20 })
+    expect(cacheRate(models[0])).toBe("57.1%")
+    expect(cacheRate({ ...models[0], tokens: { ...tokens, input: 0, cache: { read: 0, write: 0 } } })).toBe("-")
     expect(groupModelUsage(models, providers).map((group) => group.providerName)).toEqual(["Kilo Gateway", "MiniMax"])
     expect(modelUsageName(models[0], providers)).toBe("Qwen 3.7 Plus")
     expect(modelUsageName(models[1], providers)).toBe("MiniMax M3")

packages/kilo-vscode/webview-ui/src/components/chat/TaskUsage.tsx

@@ -5,7 +5,7 @@ import { Icon } from "@kilocode/kilo-ui/icon"
 import { useLanguage } from "../../context/language"
 import { useProvider } from "../../context/provider"
 import type { SessionModelUsage } from "../../types/messages"
-import { groupModelUsage, modelUsageName, type TokenSummary } from "../../context/model-usage"
+import { cacheRate, groupModelUsage, modelUsageName, type TokenSummary } from "../../context/model-usage"
 import { formatCompactCount } from "../../utils/format"
 
 interface TaskUsageProps {
@@ -35,12 +35,6 @@ export const TaskUsage: Component<TaskUsageProps> = (props) => {
     if (value > 0 && value < 0.000001) return "<$0.000001"
     return money().format(value)
   }
-  const rate = (model: SessionModelUsage["models"][number]) => {
-    const total = model.tokens.input + model.tokens.cache.read
-    if (total === 0) return "-"
-    return `${((model.tokens.cache.read / total) * 100).toFixed(1)}%`
-  }
-
   const Summary = () => (
     <>
       <span class="task-header-tokens-label">Tokens</span>
@@ -102,7 +96,7 @@ export const TaskUsage: Component<TaskUsageProps> = (props) => {
                         </div>
                         <div class="task-header-usage-meta">
                           Cache R {count(model.tokens.cache.read)} · W {count(model.tokens.cache.write)} · Hit Rate{" "}
-                          {rate(model)}
+                          {cacheRate(model)}
                         </div>
                       </div>
                     )}

packages/kilo-vscode/webview-ui/src/context/model-usage.ts

@@ -46,6 +46,12 @@ export function tokenSummary(usage: SessionModelUsage): TokenSummary {
   }
 }
 
+export function cacheRate(model: SessionModelUsage["models"][number]) {
+  const total = model.tokens.input + model.tokens.cache.read + model.tokens.cache.write
+  if (total === 0) return "-"
+  return `${((model.tokens.cache.read / total) * 100).toFixed(1)}%`
+}
+
 export function groupModelUsage(models: SessionModelUsage["models"], providers: Record<string, Provider>) {
   const groups = new Map<string, { providerID: string; providerName: string; models: SessionModelUsage["models"] }>()
   for (const model of models) {