fix(native-chat): require complete image echo evidence

(cherry picked from commit 94f1cc66ee)

src/renderer/src/components/native-chat/native-chat-literal-marker-echo.test.ts

@@ -124,4 +124,38 @@ describe('literal image-marker turns in native chat', () => {
     expect(rowText(rows[0]!)).toBe('')
     expect(rowImagePaths(rows[0]!)).toEqual(['/tmp/a.png'])
   })
+
+  it('keeps a captioned multi-image echo until the transcript accounts for every image', () => {
+    const pending: NativeChatPendingSend[] = [
+      {
+        id: 'p1',
+        text: 'compare these',
+        imagePaths: ['/tmp/a.png', '/tmp/b.png', '/tmp/c.png'],
+        sentAt: 100
+      }
+    ]
+    const transcript = [
+      userMessage('source-a', '[Image: source: /tmp/a.png]'),
+      userMessage('prompt', 'compare these [Image #1]')
+    ]
+
+    const rows = renderedUserRows(transcript, pending)
+
+    expect(rows.map(rowText)).toEqual(['compare these', 'compare these'])
+    expect(rows.map(rowImagePaths)).toEqual([
+      ['/tmp/a.png'],
+      ['/tmp/a.png', '/tmp/b.png', '/tmp/c.png']
+    ])
+
+    const complete = [
+      userMessage('source-a', '[Image: source: /tmp/a.png]'),
+      userMessage('source-b', '[Image: source: /tmp/b.png]'),
+      userMessage('source-c', '[Image: source: /tmp/c.png]'),
+      userMessage('complete-prompt', 'compare these [Image #1] [Image #2] [Image #3]')
+    ]
+    const completeRows = renderedUserRows(complete, pending)
+    expect(completeRows).toHaveLength(1)
+    expect(rowText(completeRows[0]!)).toBe('compare these')
+    expect(rowImagePaths(completeRows[0]!)).toEqual(['/tmp/a.png', '/tmp/b.png', '/tmp/c.png'])
+  })
 })

src/renderer/src/components/native-chat/native-chat-pending-occurrence.ts

@@ -1,7 +1,9 @@
 import {
+  nativeChatUserMessageImageEvidenceCount,
   nativeChatUserMessageMatchText,
+  nativeChatUserTextMatchText,
   normalizeNativeChatUserText,
-  normalizeNativeChatUserTextWithLiteralFallback
+  normalizedNativeChatUserMessageText
 } from '../../../../shared/native-chat-image-transcript-markers'
 import { isImageRefBlock, type NativeChatMessage } from '../../../../shared/native-chat-types'
 
@@ -25,36 +27,47 @@ export function normalizeNativeChatPendingText(text: string): string {
 function nativeChatPendingMatchText(
   pending: Pick<NativeChatPendingOccurrence, 'text' | 'imagePaths'>
 ): string {
-  return pending.imagePaths?.some(Boolean)
-    ? normalizeNativeChatPendingText(pending.text)
-    : normalizeNativeChatUserTextWithLiteralFallback(pending.text)
+  return nativeChatUserTextMatchText(pending.text, Boolean(pending.imagePaths?.some(Boolean)))
+}
+
+function nativeChatTextContentKey(text: string, imageCount: number): string {
+  return imageCount > 0 ? `text:${text}\0images:${imageCount}` : `text:${text}`
 }
 
 export function nativeChatPendingContentKey(
   pending: Pick<NativeChatPendingOccurrence, 'text' | 'imagePaths'>
 ): string {
   const text = nativeChatPendingMatchText(pending)
-  if (text) {
-    return `text:${text}`
-  }
   const imagePaths = pending.imagePaths?.filter(Boolean) ?? []
+  if (text) {
+    return nativeChatTextContentKey(text, imagePaths.length)
+  }
   return imagePaths.length > 0 ? `images:${JSON.stringify(imagePaths)}` : 'empty'
 }
 
-function nativeChatUserMessageContentKey(message: NativeChatMessage): string | null {
+function nativeChatUserMessageContentKeys(message: NativeChatMessage): readonly string[] {
   if (message.role !== 'user') {
-    return null
+    return []
   }
-  const text = nativeChatUserMessageMatchText(message) ?? ''
-  if (text) {
-    return `text:${text}`
+  const keys = new Set<string>()
+  const matchText = nativeChatUserMessageMatchText(message) ?? ''
+  if (matchText) {
+    keys.add(nativeChatTextContentKey(matchText, 0))
   }
-  const imagePaths = message.blocks
-    .filter(isImageRefBlock)
-    .map((block) => block.path)
-    .filter((path): path is string => Boolean(path))
-  const key = nativeChatPendingContentKey({ text: '', imagePaths })
-  return key === 'empty' ? null : key
+  const imageCount = nativeChatUserMessageImageEvidenceCount(message)
+  const imageText = normalizedNativeChatUserMessageText(message)
+  if (imageText && imageCount > 0) {
+    keys.add(nativeChatTextContentKey(imageText, imageCount))
+  } else if (imageCount > 0) {
+    const imagePaths = message.blocks
+      .filter(isImageRefBlock)
+      .map((block) => block.path)
+      .filter((path): path is string => Boolean(path))
+    if (imagePaths.length > 0) {
+      keys.add(`images:${JSON.stringify(imagePaths)}`)
+    }
+  }
+  return [...keys]
 }
 
 export function matchingNativeChatUserContentCounts(
@@ -62,8 +75,7 @@ export function matchingNativeChatUserContentCounts(
 ): Map<string, number> {
   const counts = new Map<string, number>()
   for (const message of messages) {
-    const key = nativeChatUserMessageContentKey(message)
-    if (key) {
+    for (const key of nativeChatUserMessageContentKeys(message)) {
       counts.set(key, (counts.get(key) ?? 0) + 1)
     }
   }
@@ -77,8 +89,7 @@ export function advancedNativeChatUserContentCounts(
   const waiting = new Map<string, number>()
   for (const message of messages) {
     if (message.role === 'user') {
-      const key = nativeChatUserMessageContentKey(message)
-      if (key) {
+      for (const key of nativeChatUserMessageContentKeys(message)) {
         waiting.set(key, (waiting.get(key) ?? 0) + 1)
       }
       continue
@@ -180,10 +191,13 @@ export function selectPendingIndicesRepresentedByUserTexts(
   }
   const remaining = pending.map((entry, index) => ({
     index,
-    text: nativeChatPendingMatchText(entry)
+    text: nativeChatPendingMatchText(entry),
+    hasImages: Boolean(entry.imagePaths?.some(Boolean))
   }))
   for (const userText of userTexts) {
-    const open = remaining.filter((entry) => !represented.has(entry.index) && entry.text.length > 0)
+    const open = remaining.filter(
+      (entry) => !entry.hasImages && !represented.has(entry.index) && entry.text.length > 0
+    )
     const gluedCount = countLeadingPendingTextsGluedToUserText(
       open.map((entry) => entry.text),
       userText

src/shared/native-chat-image-transcript-markers.ts

@@ -48,10 +48,16 @@ export function normalizeNativeChatUserText(text: string): string {
 /** Marker text that stripping would erase entirely is text the user typed
  *  literally — a real marker always rides on an image turn, which normalization
  *  has already folded away by the time echo matching runs. */
-export function normalizeNativeChatUserTextWithLiteralFallback(text: string): string {
+function normalizeNativeChatUserTextWithLiteralFallback(text: string): string {
   return normalizeNativeChatUserText(text) || text.trim().replace(/\s+/g, ' ')
 }
 
+export function nativeChatUserTextMatchText(text: string, hasImages: boolean): string {
+  return hasImages
+    ? normalizeNativeChatUserText(text)
+    : normalizeNativeChatUserTextWithLiteralFallback(text)
+}
+
 function joinedUserText(message: NativeChatMessage): string {
   return message.blocks
     .filter(isTextBlock)
@@ -74,11 +80,7 @@ export function nativeChatUserMessageMatchText(message: NativeChatMessage): stri
     return null
   }
   const joined = joinedUserText(message)
-  const normalized = normalizeNativeChatUserText(joined)
-  if (normalized || message.blocks.some(isImageRefBlock)) {
-    return normalized || null
-  }
-  return normalizeNativeChatUserTextWithLiteralFallback(joined) || null
+  return nativeChatUserTextMatchText(joined, message.blocks.some(isImageRefBlock)) || null
 }
 
 function stripImagePromptMarkersFromTextBlocks(
@@ -132,6 +134,14 @@ export function countImagePromptMarkers(message: NativeChatMessage): number {
   )
 }
 
+export function nativeChatUserMessageImageEvidenceCount(message: NativeChatMessage): number {
+  if (message.role !== 'user') {
+    return 0
+  }
+  const imageRefCount = message.blocks.filter(isImageRefBlock).length
+  return Math.max(imageRefCount, countImagePromptMarkers(message))
+}
+
 /** Claude records image paths as source turns followed by a prompt carrying
  *  image markers. Merge the whole run back into one native user turn. */
 export function normalizeImageTranscriptMessages(