fix(agents): keep structured prompt summaries UTF-16 safe

Alix-007 · Alix-007 · commit b35de5205580 · 2026-07-07T12:40:09.000+08:00
diff --git a/src/agents/embedded-agent-runner/run/attempt.prompt-helpers.test.ts b/src/agents/embedded-agent-runner/run/attempt.prompt-helpers.test.ts
@@ -38,12 +38,31 @@ vi.mock("../../../plugins/host-hook-state.js", () => hostHookStateMocks);
 
 import {
   forgetPromptBuildDrainCacheForRun,
+  mergeOrphanedTrailingUserPrompt,
   resolvePromptSubmissionSkipReason,
   resolveAttemptMediaTaskSystemPromptAddition,
   resolvePromptBuildHookResult,
   shouldInjectHeartbeatPrompt,
 } from "./attempt.prompt-helpers.js";
 
+function hasLoneSurrogate(value: string): boolean {
+  for (let index = 0; index < value.length; index += 1) {
+    const code = value.charCodeAt(index);
+    if (code >= 0xd800 && code <= 0xdbff) {
+      const next = value.charCodeAt(index + 1);
+      if (!(next >= 0xdc00 && next <= 0xdfff)) {
+        return true;
+      }
+      index += 1;
+      continue;
+    }
+    if (code >= 0xdc00 && code <= 0xdfff) {
+      return true;
+    }
+  }
+  return false;
+}
+
 describe("shouldInjectHeartbeatPrompt", () => {
   it("keeps global heartbeat guidance out of commitment-only runs", () => {
     const heartbeatParams = {
@@ -64,6 +83,31 @@ describe("shouldInjectHeartbeatPrompt", () => {
   });
 });
 
+describe("mergeOrphanedTrailingUserPrompt", () => {
+  it("keeps structured media and JSON summaries on UTF-16 boundaries", () => {
+    const result = mergeOrphanedTrailingUserPrompt({
+      prompt: "Continue.",
+      trigger: "user",
+      leafMessage: {
+        content: [
+          {
+            type: "image_url",
+            image_url: { url: `${"u".repeat(299)}😀tail` },
+          },
+          {
+            [`${"k".repeat(997)}😀tail`]: 1,
+          },
+        ],
+      },
+    });
+
+    expect(result.merged).toBe(true);
+    expect(hasLoneSurrogate(result.prompt)).toBe(false);
+    expect(result.prompt).toContain("[image_url]");
+    expect(result.prompt).toContain("chars)");
+  });
+});
+
 describe("resolveAttemptMediaTaskSystemPromptAddition", () => {
   it("joins active media task guidance for user triggers", () => {
     imageGenerationTaskStatusMocks.buildActiveImageGenerationTaskPromptContextForSession.mockReturnValue(
diff --git a/src/agents/embedded-agent-runner/run/attempt.prompt-helpers.ts b/src/agents/embedded-agent-runner/run/attempt.prompt-helpers.ts
@@ -30,6 +30,7 @@ import { buildActiveVideoGenerationTaskPromptContextForSession } from "../../vid
 import { buildEmbeddedCompactionRuntimeContext } from "../compaction-runtime-context.js";
 import { resolveContextEngineCapabilities } from "../context-engine-capabilities.js";
 import { log } from "../logger.js";
+import { truncateUtf16Safe } from "../../../utils.js";
 import { shouldInjectHeartbeatPromptForTrigger } from "./trigger-policy.js";
 import type { EmbeddedRunAttemptParams } from "./types.js";
 
@@ -337,7 +338,7 @@ function summarizeStructuredMediaRef(label: string, value: unknown): string | un
     return `[${label}] inline data URI (${mimeType}, ${trimmed.length} chars)`;
   }
   if (trimmed.length > MAX_STRUCTURED_MEDIA_REF_CHARS) {
-    return `[${label}] ${trimmed.slice(0, MAX_STRUCTURED_MEDIA_REF_CHARS)}... (${trimmed.length} chars)`;
+    return `[${label}] ${truncateUtf16Safe(trimmed, MAX_STRUCTURED_MEDIA_REF_CHARS)}... (${trimmed.length} chars)`;
   }
   return `[${label}] ${trimmed}`;
 }
@@ -349,7 +350,7 @@ function summarizeStructuredJsonString(value: string): string {
   }
   const trimmed = value.trim();
   if (trimmed.length > MAX_STRUCTURED_JSON_STRING_CHARS) {
-    return `${trimmed.slice(0, MAX_STRUCTURED_JSON_STRING_CHARS)}... (${trimmed.length} chars)`;
+    return `${truncateUtf16Safe(trimmed, MAX_STRUCTURED_JSON_STRING_CHARS)}... (${trimmed.length} chars)`;
   }
   return value;
 }
@@ -418,7 +419,7 @@ function stringifyStructuredJsonFallback(part: unknown): string | undefined {
       (match) => `[inline data URI: ${match.length} chars]`,
     );
     return withoutInlineData.length > 1_000
-      ? `${withoutInlineData.slice(0, 1_000)}... (${withoutInlineData.length} chars)`
+      ? `${truncateUtf16Safe(withoutInlineData, 1_000)}... (${withoutInlineData.length} chars)`
       : withoutInlineData;
   } catch {
     return undefined;

Original file line number	Diff line number	Diff line change
`@@ -30,6 +30,7 @@ import { buildActiveVideoGenerationTaskPromptContextForSession } from "../../vid`
`30`	`30`	`import { buildEmbeddedCompactionRuntimeContext } from "../compaction-runtime-context.js";`
`31`	`31`	`import { resolveContextEngineCapabilities } from "../context-engine-capabilities.js";`
`32`	`32`	`import { log } from "../logger.js";`
	`33`	`+import { truncateUtf16Safe } from "../../../utils.js";`
`33`	`34`	`import { shouldInjectHeartbeatPromptForTrigger } from "./trigger-policy.js";`
`34`	`35`	`import type { EmbeddedRunAttemptParams } from "./types.js";`
`35`	`36`
`@@ -337,7 +338,7 @@ function summarizeStructuredMediaRef(label: string, value: unknown): string \| un`
`337`	`338`	return `[${label}] inline data URI (${mimeType}, ${trimmed.length} chars)`;
`338`	`339`	`}`
`339`	`340`	`if (trimmed.length > MAX_STRUCTURED_MEDIA_REF_CHARS) {`
`340`		- return `[${label}] ${trimmed.slice(0, MAX_STRUCTURED_MEDIA_REF_CHARS)}... (${trimmed.length} chars)`;
	`341`	+ return `[${label}] ${truncateUtf16Safe(trimmed, MAX_STRUCTURED_MEDIA_REF_CHARS)}... (${trimmed.length} chars)`;
`341`	`342`	`}`
`342`	`343`	return `[${label}] ${trimmed}`;
`343`	`344`	`}`
`@@ -349,7 +350,7 @@ function summarizeStructuredJsonString(value: string): string {`
`349`	`350`	`}`
`350`	`351`	`const trimmed = value.trim();`
`351`	`352`	`if (trimmed.length > MAX_STRUCTURED_JSON_STRING_CHARS) {`
`352`		- return `${trimmed.slice(0, MAX_STRUCTURED_JSON_STRING_CHARS)}... (${trimmed.length} chars)`;
	`353`	+ return `${truncateUtf16Safe(trimmed, MAX_STRUCTURED_JSON_STRING_CHARS)}... (${trimmed.length} chars)`;
`353`	`354`	`}`
`354`	`355`	`return value;`
`355`	`356`	`}`
`@@ -418,7 +419,7 @@ function stringifyStructuredJsonFallback(part: unknown): string \| undefined {`
`418`	`419`	(match) => `[inline data URI: ${match.length} chars]`,
`419`	`420`	`);`
`420`	`421`	`return withoutInlineData.length > 1_000`
`421`		- ? `${withoutInlineData.slice(0, 1_000)}... (${withoutInlineData.length} chars)`
	`422`	+ ? `${truncateUtf16Safe(withoutInlineData, 1_000)}... (${withoutInlineData.length} chars)`
`422`	`423`	`: withoutInlineData;`
`423`	`424`	`} catch {`
`424`	`425`	`return undefined;`