fix: harden BlueBubbles voice memos (#1477) (thanks @Nicell)

CHANGELOG.md

@@ -9,6 +9,7 @@ Docs: https://docs.clawd.bot
 - Lobster: allow workflow file args via `argsJson` in the plugin tool. https://docs.clawd.bot/tools/lobster
 - Agents: add identity avatar config support and Control UI avatar rendering. (#1329, #1424) Thanks @dlauer.
 - Memory: prevent CLI hangs by deferring vector probes, adding sqlite-vec/embedding timeouts, and showing sync progress early.
+- BlueBubbles: add `asVoice` support for MP3/CAF voice memos in sendAttachment. (#1477) Thanks @Nicell.
 - Docs: add troubleshooting entry for gateway.mode blocking gateway start. https://docs.clawd.bot/gateway/troubleshooting
 - Docs: add /model allowlist troubleshooting note. (#1405)
 - Docs: add per-message Gmail search example for gog. (#1220) Thanks @mbelinky.

docs/channels/bluebubbles.md

@@ -147,7 +147,8 @@ Available actions:
 - **addParticipant**: Add someone to a group (`chatGuid`, `address`)
 - **removeParticipant**: Remove someone from a group (`chatGuid`, `address`)
 - **leaveGroup**: Leave a group chat (`chatGuid`)
-- **sendAttachment**: Send media/files (`to`, `buffer`, `filename`)
+- **sendAttachment**: Send media/files (`to`, `buffer`, `filename`, `asVoice`)
+  - Voice memos: set `asVoice: true` with **MP3** or **CAF** audio to send as an iMessage voice message. BlueBubbles converts MP3 → CAF when sending voice memos.
 
 ### Message IDs (short vs full)
 Clawdbot may surface *short* message IDs (e.g., `1`, `2`) to save tokens.

extensions/bluebubbles/src/actions.test.ts

@@ -521,6 +521,42 @@ describe("bluebubblesMessageActions", () => {
       });
     });
 
+    it("passes asVoice through sendAttachment", async () => {
+      const { sendBlueBubblesAttachment } = await import("./attachments.js");
+
+      const cfg: ClawdbotConfig = {
+        channels: {
+          bluebubbles: {
+            serverUrl: "http://localhost:1234",
+            password: "test-password",
+          },
+        },
+      };
+
+      const base64Buffer = Buffer.from("voice").toString("base64");
+
+      await bluebubblesMessageActions.handleAction({
+        action: "sendAttachment",
+        params: {
+          to: "+15551234567",
+          filename: "voice.mp3",
+          buffer: base64Buffer,
+          contentType: "audio/mpeg",
+          asVoice: true,
+        },
+        cfg,
+        accountId: null,
+      });
+
+      expect(sendBlueBubblesAttachment).toHaveBeenCalledWith(
+        expect.objectContaining({
+          filename: "voice.mp3",
+          contentType: "audio/mpeg",
+          asVoice: true,
+        }),
+      );
+    });
+
     it("throws when buffer is missing for setGroupIcon", async () => {
       const cfg: ClawdbotConfig = {
         channels: {

extensions/bluebubbles/src/actions.ts

@@ -3,7 +3,6 @@ import {
   BLUEBUBBLES_ACTIONS,
   createActionGate,
   jsonResult,
-  readBooleanParam,
   readNumberParam,
   readReactionParams,
   readStringParam,
@@ -51,6 +50,17 @@ function readMessageText(params: Record<string, unknown>): string | undefined {
   return readStringParam(params, "text") ?? readStringParam(params, "message");
 }
 
+function readBooleanParam(params: Record<string, unknown>, key: string): boolean | undefined {
+  const raw = params[key];
+  if (typeof raw === "boolean") return raw;
+  if (typeof raw === "string") {
+    const trimmed = raw.trim().toLowerCase();
+    if (trimmed === "true") return true;
+    if (trimmed === "false") return false;
+  }
+  return undefined;
+}
+
 /** Supported action names for BlueBubbles */
 const SUPPORTED_ACTIONS = new Set<ChannelMessageActionName>(BLUEBUBBLES_ACTION_NAMES);
 

extensions/bluebubbles/src/attachments.test.ts

@@ -1,6 +1,6 @@
 import { describe, expect, it, vi, beforeEach, afterEach } from "vitest";
 
-import { downloadBlueBubblesAttachment } from "./attachments.js";
+import { downloadBlueBubblesAttachment, sendBlueBubblesAttachment } from "./attachments.js";
 import type { BlueBubblesAttachment } from "./types.js";
 
 vi.mock("./accounts.js", () => ({
@@ -238,3 +238,109 @@ describe("downloadBlueBubblesAttachment", () => {
     expect(result.buffer).toEqual(new Uint8Array([1]));
   });
 });
+
+describe("sendBlueBubblesAttachment", () => {
+  beforeEach(() => {
+    vi.stubGlobal("fetch", mockFetch);
+    mockFetch.mockReset();
+  });
+
+  afterEach(() => {
+    vi.unstubAllGlobals();
+  });
+
+  function decodeBody(body: Uint8Array) {
+    return Buffer.from(body).toString("utf8");
+  }
+
+  it("marks voice memos when asVoice is true and mp3 is provided", async () => {
+    mockFetch.mockResolvedValueOnce({
+      ok: true,
+      text: () => Promise.resolve(JSON.stringify({ messageId: "msg-1" })),
+    });
+
+    await sendBlueBubblesAttachment({
+      to: "chat_guid:iMessage;-;+15551234567",
+      buffer: new Uint8Array([1, 2, 3]),
+      filename: "voice.mp3",
+      contentType: "audio/mpeg",
+      asVoice: true,
+      opts: { serverUrl: "http://localhost:1234", password: "test" },
+    });
+
+    const body = mockFetch.mock.calls[0][1]?.body as Uint8Array;
+    const bodyText = decodeBody(body);
+    expect(bodyText).toContain('name="isAudioMessage"');
+    expect(bodyText).toContain("true");
+    expect(bodyText).toContain('filename="voice.mp3"');
+  });
+
+  it("normalizes mp3 filenames for voice memos", async () => {
+    mockFetch.mockResolvedValueOnce({
+      ok: true,
+      text: () => Promise.resolve(JSON.stringify({ messageId: "msg-2" })),
+    });
+
+    await sendBlueBubblesAttachment({
+      to: "chat_guid:iMessage;-;+15551234567",
+      buffer: new Uint8Array([1, 2, 3]),
+      filename: "voice",
+      contentType: "audio/mpeg",
+      asVoice: true,
+      opts: { serverUrl: "http://localhost:1234", password: "test" },
+    });
+
+    const body = mockFetch.mock.calls[0][1]?.body as Uint8Array;
+    const bodyText = decodeBody(body);
+    expect(bodyText).toContain('filename="voice.mp3"');
+    expect(bodyText).toContain('name="voice.mp3"');
+  });
+
+  it("throws when asVoice is true but media is not audio", async () => {
+    await expect(
+      sendBlueBubblesAttachment({
+        to: "chat_guid:iMessage;-;+15551234567",
+        buffer: new Uint8Array([1, 2, 3]),
+        filename: "image.png",
+        contentType: "image/png",
+        asVoice: true,
+        opts: { serverUrl: "http://localhost:1234", password: "test" },
+      }),
+    ).rejects.toThrow("voice messages require audio");
+    expect(mockFetch).not.toHaveBeenCalled();
+  });
+
+  it("throws when asVoice is true but audio is not mp3 or caf", async () => {
+    await expect(
+      sendBlueBubblesAttachment({
+        to: "chat_guid:iMessage;-;+15551234567",
+        buffer: new Uint8Array([1, 2, 3]),
+        filename: "voice.wav",
+        contentType: "audio/wav",
+        asVoice: true,
+        opts: { serverUrl: "http://localhost:1234", password: "test" },
+      }),
+    ).rejects.toThrow("require mp3 or caf");
+    expect(mockFetch).not.toHaveBeenCalled();
+  });
+
+  it("sanitizes filenames before sending", async () => {
+    mockFetch.mockResolvedValueOnce({
+      ok: true,
+      text: () => Promise.resolve(JSON.stringify({ messageId: "msg-3" })),
+    });
+
+    await sendBlueBubblesAttachment({
+      to: "chat_guid:iMessage;-;+15551234567",
+      buffer: new Uint8Array([1, 2, 3]),
+      filename: "../evil.mp3",
+      contentType: "audio/mpeg",
+      opts: { serverUrl: "http://localhost:1234", password: "test" },
+    });
+
+    const body = mockFetch.mock.calls[0][1]?.body as Uint8Array;
+    const bodyText = decodeBody(body);
+    expect(bodyText).toContain('filename="evil.mp3"');
+    expect(bodyText).toContain('name="evil.mp3"');
+  });
+});

extensions/bluebubbles/src/attachments.ts

@@ -1,8 +1,5 @@
 import crypto from "node:crypto";
-import { spawn } from "node:child_process";
-import { writeFile, unlink, mkdtemp, readFile } from "node:fs/promises";
-import { tmpdir } from "node:os";
-import { join } from "node:path";
+import path from "node:path";
 import type { ClawdbotConfig } from "clawdbot/plugin-sdk";
 import { resolveBlueBubblesAccount } from "./accounts.js";
 import { resolveChatGuidForTarget } from "./send.js";
@@ -23,6 +20,30 @@ export type BlueBubblesAttachmentOpts = {
 };
 
 const DEFAULT_ATTACHMENT_MAX_BYTES = 8 * 1024 * 1024;
+const AUDIO_MIME_MP3 = new Set(["audio/mpeg", "audio/mp3"]);
+const AUDIO_MIME_CAF = new Set(["audio/x-caf", "audio/caf"]);
+
+function sanitizeFilename(input: string | undefined, fallback: string): string {
+  const trimmed = input?.trim() ?? "";
+  const base = trimmed ? path.basename(trimmed) : "";
+  return base || fallback;
+}
+
+function ensureExtension(filename: string, extension: string, fallbackBase: string): string {
+  const currentExt = path.extname(filename);
+  if (currentExt.toLowerCase() === extension) return filename;
+  const base = currentExt ? filename.slice(0, -currentExt.length) : filename;
+  return `${base || fallbackBase}${extension}`;
+}
+
+function resolveVoiceInfo(filename: string, contentType?: string) {
+  const normalizedType = contentType?.trim().toLowerCase();
+  const extension = path.extname(filename).toLowerCase();
+  const isMp3 = extension === ".mp3" || (normalizedType ? AUDIO_MIME_MP3.has(normalizedType) : false);
+  const isCaf = extension === ".caf" || (normalizedType ? AUDIO_MIME_CAF.has(normalizedType) : false);
+  const isAudio = isMp3 || isCaf || Boolean(normalizedType?.startsWith("audio/"));
+  return { isAudio, isMp3, isCaf };
+}
 
 function resolveAccount(params: BlueBubblesAttachmentOpts) {
   const account = resolveBlueBubblesAccount({
@@ -68,65 +89,6 @@ export type SendBlueBubblesAttachmentResult = {
   messageId: string;
 };
 
-/**
- * Convert audio to Opus CAF format for iMessage voice messages.
- * iMessage voice memos use Opus codec at 48kHz in CAF container.
- */
-async function convertToVoiceFormat(
-  inputBuffer: Uint8Array,
-  inputFilename: string,
-): Promise<{ buffer: Uint8Array; filename: string; contentType: string }> {
-  const tempDir = await mkdtemp(join(tmpdir(), "bb-voice-"));
-  const inputPath = join(tempDir, inputFilename);
-  const outputPath = join(tempDir, "Audio Message.caf");
-
-  try {
-    await writeFile(inputPath, inputBuffer);
-
-    // Convert to Opus CAF (iMessage voice memo format)
-    await new Promise<void>((resolve, reject) => {
-      const ffmpeg = spawn("ffmpeg", [
-        "-y",
-        "-i", inputPath,
-        "-ar", "48000",
-        "-c:a", "libopus",
-        "-b:a", "32k",
-        "-f", "caf",
-        outputPath,
-      ]);
-
-      let stderr = "";
-      ffmpeg.stderr.on("data", (data) => {
-        stderr += data.toString();
-      });
-
-      ffmpeg.on("close", (code) => {
-        if (code === 0) {
-          resolve();
-        } else {
-          reject(new Error(`ffmpeg conversion failed (code ${code}): ${stderr.slice(-500)}`));
-        }
-      });
-
-      ffmpeg.on("error", (err) => {
-        reject(new Error(`ffmpeg spawn error: ${err.message}`));
-      });
-    });
-
-    const outputBuffer = await readFile(outputPath);
-    return {
-      buffer: new Uint8Array(outputBuffer),
-      filename: "Audio Message.caf",
-      contentType: "audio/x-caf",
-    };
-  } finally {
-    // Cleanup temp files
-    await unlink(inputPath).catch(() => {});
-    await unlink(outputPath).catch(() => {});
-    await unlink(tempDir).catch(() => {});
-  }
-}
-
 function resolveSendTarget(raw: string): BlueBubblesSendTarget {
   const parsed = parseBlueBubblesTarget(raw);
   if (parsed.kind === "handle") {
@@ -167,7 +129,7 @@ function extractMessageId(payload: unknown): string {
 /**
  * Send an attachment via BlueBubbles API.
  * Supports sending media files (images, videos, audio, documents) to a chat.
- * When asVoice is true, converts audio to iMessage voice memo format (Opus CAF).
+ * When asVoice is true, expects MP3/CAF audio and marks it as an iMessage voice memo.
  */
 export async function sendBlueBubblesAttachment(params: {
   to: string;
@@ -182,19 +144,29 @@ export async function sendBlueBubblesAttachment(params: {
 }): Promise<SendBlueBubblesAttachmentResult> {
   const { to, caption, replyToMessageGuid, replyToPartIndex, asVoice, opts = {} } = params;
   let { buffer, filename, contentType } = params;
+  const wantsVoice = asVoice === true;
+  const fallbackName = wantsVoice ? "Audio Message" : "attachment";
+  filename = sanitizeFilename(filename, fallbackName);
+  contentType = contentType?.trim() || undefined;
   const { baseUrl, password } = resolveAccount(opts);
 
-  // Convert to voice memo format if requested
-  const isAudioMessage = asVoice === true;
+  // Validate voice memo format when requested (BlueBubbles converts MP3 -> CAF when isAudioMessage).
+  const isAudioMessage = wantsVoice;
   if (isAudioMessage) {
-    try {
-      const converted = await convertToVoiceFormat(buffer, filename);
-      buffer = converted.buffer;
-      filename = converted.filename;
-      contentType = converted.contentType;
-    } catch (err) {
-      const msg = err instanceof Error ? err.message : String(err);
-      throw new Error(`Failed to convert audio to voice format: ${msg}`);
+    const voiceInfo = resolveVoiceInfo(filename, contentType);
+    if (!voiceInfo.isAudio) {
+      throw new Error("BlueBubbles voice messages require audio media (mp3 or caf).");
+    }
+    if (voiceInfo.isMp3) {
+      filename = ensureExtension(filename, ".mp3", fallbackName);
+      contentType = contentType ?? "audio/mpeg";
+    } else if (voiceInfo.isCaf) {
+      filename = ensureExtension(filename, ".caf", fallbackName);
+      contentType = contentType ?? "audio/x-caf";
+    } else {
+      throw new Error(
+        "BlueBubbles voice messages require mp3 or caf audio (convert before sending).",
+      );
     }
   }