fix(file-transfer): keep bounded stderr tail UTF-16 safe end-to-end (#104151)

Co-authored-by: NIO <nocodet@mail.com>

extensions/file-transfer/src/shared/append-bounded-text-tail.test.ts

@@ -0,0 +1,25 @@
+// File Transfer tests cover bounded stderr tail UTF-16 safety.
+import { describe, expect, it } from "vitest";
+import { appendBoundedTextTail, projectBoundedTextTail } from "./append-bounded-text-tail.js";
+
+const hasUnpairedUtf16Surrogate = (text: string): boolean =>
+  /[\uD800-\uDBFF](?![\uDC00-\uDFFF])/.test(text);
+
+describe("appendBoundedTextTail", () => {
+  it("keeps stderr tail UTF-16 safe when the boundary bisects an emoji", () => {
+    const stderr = appendBoundedTextTail("prefix", Buffer.from("🤖tail"), 5);
+
+    expect(stderr).toBe("tail");
+    expect(hasUnpairedUtf16Surrogate(stderr)).toBe(false);
+  });
+});
+
+describe("projectBoundedTextTail", () => {
+  it("keeps final error projection UTF-16 safe when the boundary bisects an emoji", () => {
+    const stderr = "p".repeat(5) + "🤖fail";
+    const projected = projectBoundedTextTail(stderr, 5);
+
+    expect(projected).toBe("fail");
+    expect(hasUnpairedUtf16Surrogate(projected)).toBe(false);
+  });
+});

extensions/file-transfer/src/shared/append-bounded-text-tail.ts

@@ -0,0 +1,15 @@
+// Bounded stderr tails must stay UTF-16 safe when tar/dir-fetch paths mention emoji filenames.
+import { sliceUtf16Safe } from "openclaw/plugin-sdk/text-utility-runtime";
+
+export function appendBoundedTextTail(current: string, chunk: Buffer, maxChars: number): string {
+  const next = current + chunk.toString();
+  if (next.length <= maxChars) {
+    return next;
+  }
+  return sliceUtf16Safe(next, Math.max(0, next.length - maxChars));
+}
+
+// Final error projections reuse the ring-buffer tail; raw slice(-N) can still bisect emoji.
+export function projectBoundedTextTail(text: string, maxChars: number): string {
+  return sliceUtf16Safe(text, Math.max(0, text.length - maxChars));
+}

extensions/file-transfer/src/shared/node-invoke-policy.stream-errors.test.ts

@@ -70,4 +70,32 @@ describe("dir.fetch archive policy output lifecycle", () => {
       }),
     ).resolves.toEqual({ ok: true, entries: ["ok.txt"] });
   });
+
+  it("surfaces UTF-16 safe tar stderr tail when archive listing fails with emoji at projection boundary", async () => {
+    const oldNoise = "n".repeat(250);
+    // Length 201: raw slice(-200) would start on the low surrogate of 🤖.
+    const recent = "🤖" + "f".repeat(199);
+    const spawnMock = mockTarSpawn((child) => {
+      child.stderr.emit("data", Buffer.from(oldNoise));
+      child.stderr.emit("data", Buffer.from(recent));
+      child.emit("close", 2);
+    });
+    const { testing } = await importWithSpawn(spawnMock);
+    const { projectBoundedTextTail } = await import("./append-bounded-text-tail.js");
+
+    const result = await testing.listDirFetchArchiveEntries({
+      tarBase64: Buffer.from("archive").toString("base64"),
+    });
+    expect(result.ok).toBe(false);
+    if (!result.ok) {
+      expect(result.reason).toContain(projectBoundedTextTail(recent, 200));
+      expect(result.reason).toContain("f".repeat(199));
+      expect(result.reason).not.toContain("🤖");
+      expect(
+        /[\uD800-\uDBFF](?![\uDC00-\uDFFF])|(?<![\uD800-\uDBFF])[\uDC00-\uDFFF]/.test(
+          result.reason,
+        ),
+      ).toBe(false);
+    }
+  });
 });

extensions/file-transfer/src/shared/node-invoke-policy.ts

@@ -6,6 +6,7 @@ import type {
   OpenClawPluginNodeInvokePolicyContext,
   OpenClawPluginNodeInvokePolicyResult,
 } from "openclaw/plugin-sdk/plugin-entry";
+import { appendBoundedTextTail, projectBoundedTextTail } from "./append-bounded-text-tail.js";
 import { appendFileTransferAudit, type FileTransferAuditOp } from "./audit.js";
 import { consumeChildOutput } from "./child-output.js";
 import {
@@ -22,6 +23,7 @@ const DIR_FETCH_MAX_ENTRIES = 5000;
 const DIR_FETCH_ARCHIVE_LIST_TIMEOUT_MS = 30_000;
 const DIR_FETCH_ARCHIVE_LIST_MAX_OUTPUT_BYTES = 32 * 1024 * 1024;
 const DIR_FETCH_ARCHIVE_LIST_STDERR_TAIL_CHARS = 4096;
+const DIR_FETCH_ARCHIVE_LIST_ERROR_STDERR_CHARS = 200;
 
 type FileTransferCommand = FileTransferNodeInvokeCommand;
 
@@ -31,11 +33,6 @@ function asRecord(value: unknown): Record<string, unknown> {
     : {};
 }
 
-function appendBoundedTextTail(current: string, chunk: Buffer, maxChars: number): string {
-  const next = current + chunk.toString();
-  return next.length > maxChars ? next.slice(-maxChars) : next;
-}
-
 function readPath(params: Record<string, unknown>): string {
   return typeof params.path === "string" ? params.path.trim() : "";
 }
@@ -424,7 +421,7 @@ async function listDirFetchArchiveEntries(
         finish({
           ok: false,
           code: "ARCHIVE_ENTRIES_UNREADABLE",
-          reason: `tar -tzf exited ${code}: ${stderr.slice(-200)}`,
+          reason: `tar -tzf exited ${code}: ${projectBoundedTextTail(stderr, DIR_FETCH_ARCHIVE_LIST_ERROR_STDERR_CHARS)}`,
         });
         return;
       }

extensions/file-transfer/src/tools/dir-fetch-tool.test.ts

@@ -5,6 +5,7 @@ import fs from "node:fs/promises";
 import os from "node:os";
 import path from "node:path";
 import { afterEach, beforeEach, describe, expect, it, vi } from "vitest";
+import { projectBoundedTextTail } from "../shared/append-bounded-text-tail.js";
 import { validateTarUncompressedBudget } from "./dir-fetch-tool.js";
 
 let tmpRoot: string;
@@ -276,7 +277,7 @@ describe("dir.fetch tar validation", () => {
       const result = await testing.preValidateTarball(Buffer.from("x"));
       expect(result.ok).toBe(false);
       if (!result.ok) {
-        expect(result.reason).toContain(recent.slice(-200));
+        expect(result.reason).toContain(projectBoundedTextTail(recent, 200));
         expect(result.reason).not.toContain(oldNoise.slice(0, 40));
       }
     } finally {
@@ -284,4 +285,41 @@ describe("dir.fetch tar validation", () => {
       vi.resetModules();
     }
   });
+
+  it("surfaces UTF-16 safe tar stderr tail when listing fails with emoji at projection boundary", async () => {
+    vi.resetModules();
+    const oldNoise = "n".repeat(250);
+    // Length 201: raw slice(-200) would start on the low surrogate of 🤖.
+    const recent = "🤖" + "f".repeat(199);
+    vi.doMock("node:child_process", async (importOriginal) => {
+      const actual = await importOriginal<typeof import("node:child_process")>();
+      return {
+        ...actual,
+        spawn: mockTarSpawn((child) => {
+          child.stderr.emit("data", Buffer.from(oldNoise));
+          child.stderr.emit("data", Buffer.from(recent));
+          child.emit("close", 2);
+        }),
+      };
+    });
+
+    try {
+      const { testing } = await import("./dir-fetch-tool.js");
+      const result = await testing.preValidateTarball(Buffer.from("x"));
+      expect(result.ok).toBe(false);
+      if (!result.ok) {
+        expect(result.reason).toContain(projectBoundedTextTail(recent, 200));
+        expect(result.reason).toContain("f".repeat(199));
+        expect(result.reason).not.toContain("🤖");
+        expect(
+          /[\uD800-\uDBFF](?![\uDC00-\uDFFF])|(?<![\uD800-\uDBFF])[\uDC00-\uDFFF]/.test(
+            result.reason,
+          ),
+        ).toBe(false);
+      }
+    } finally {
+      vi.doUnmock("node:child_process");
+      vi.resetModules();
+    }
+  });
 });

extensions/file-transfer/src/tools/dir-fetch-tool.ts

@@ -5,6 +5,10 @@ import fs from "node:fs/promises";
 import path from "node:path";
 import type { AnyAgentTool } from "openclaw/plugin-sdk/agent-harness-runtime";
 import { saveMediaBuffer } from "openclaw/plugin-sdk/media-store";
+import {
+  appendBoundedTextTail,
+  projectBoundedTextTail,
+} from "../shared/append-bounded-text-tail.js";
 import { appendFileTransferAudit } from "../shared/audit.js";
 import { consumeChildOutput } from "../shared/child-output.js";
 import { IMAGE_MIME_INLINE_SET, mimeFromExtension } from "../shared/mime.js";
@@ -32,6 +36,8 @@ const TAR_UNPACK_TIMEOUT_MS = 60_000;
 const TAR_UNPACK_MAX_ENTRIES = 5000;
 const TAR_LIST_OUTPUT_MAX_CHARS = 32 * 1024 * 1024;
 const TAR_STDERR_TAIL_CHARS = 4096;
+const TAR_ERROR_REASON_STDERR_CHARS = 200;
+const TAR_UNPACK_ERROR_STDERR_CHARS = 300;
 
 // Hard caps on uncompressed extraction. Defends against decompression-bomb
 // archives that compress to <16MB but expand to gigabytes. Both caps are
@@ -40,11 +46,6 @@ const TAR_STDERR_TAIL_CHARS = 4096;
 const DIR_FETCH_MAX_UNCOMPRESSED_BYTES = 64 * 1024 * 1024;
 const DIR_FETCH_MAX_SINGLE_FILE_BYTES = 16 * 1024 * 1024;
 
-function appendBoundedTextTail(current: string, chunk: Buffer, maxChars: number): string {
-  const next = current + chunk.toString();
-  return next.length > maxChars ? next.slice(-maxChars) : next;
-}
-
 async function listTarOutputLines<T>(input: {
   args: string[];
   label: string;
@@ -137,7 +138,10 @@ async function listTarOutputLines<T>(input: {
         return;
       }
       if (code !== 0) {
-        finish({ ok: false, reason: `${input.label} exited ${code}: ${stderr.slice(-200)}` });
+        finish({
+          ok: false,
+          reason: `${input.label} exited ${code}: ${projectBoundedTextTail(stderr, TAR_ERROR_REASON_STDERR_CHARS)}`,
+        });
         return;
       }
       if (pending) {
@@ -350,7 +354,7 @@ export async function validateTarUncompressedBudget(
       if (code !== 0) {
         finish({
           ok: false,
-          reason: `tar uncompressed budget validation exited ${code}: ${stderr.slice(-200)}`,
+          reason: `tar uncompressed budget validation exited ${code}: ${projectBoundedTextTail(stderr, TAR_ERROR_REASON_STDERR_CHARS)}`,
         });
         return;
       }
@@ -455,7 +459,11 @@ async function unpackTar(tarBuffer: Buffer, destDir: string): Promise<void> {
     });
     child.on("close", (code) => {
       if (code !== 0) {
-        fail(new Error(`tar unpack exited ${code}: ${stderrOut.slice(-300)}`));
+        fail(
+          new Error(
+            `tar unpack exited ${code}: ${projectBoundedTextTail(stderrOut, TAR_UNPACK_ERROR_STDERR_CHARS)}`,
+          ),
+        );
         return;
       }
       succeed();