OpenBMB/ChatDev · error · TypeError

register_bytes expects bytes or bytearray data

Error message

register_bytes expects bytes or bytearray data

What it means

AttachmentManager.register_bytes raises TypeError when the data argument is not bytes or bytearray. It is a strict type guard against passing str, memoryview, file objects, or None where a raw payload is required.

Source

Thrown at utils/attachments.py:157

            self._persistent_ids.discard(attachment_id)
        return record

    def register_bytes(
        self,
        data: bytes | bytearray,
        *,
        kind: MessageBlockType = MessageBlockType.FILE,
        mime_type: Optional[str] = None,
        display_name: Optional[str] = None,
        attachment_id: Optional[str] = None,
        description: Optional[str] = None,
        extra: Optional[Dict[str, Any]] = None,
        persist: bool = True,
    ) -> AttachmentRecord:
        """Register an in-memory payload as an attachment."""

        if not isinstance(data, (bytes, bytearray)):
            raise TypeError("register_bytes expects bytes or bytearray data")

        attachment_id = attachment_id or uuid.uuid4().hex
        filename = display_name or _default_filename_for_mime(mime_type)

        target_dir = self.root / attachment_id
        target_dir.mkdir(parents=True, exist_ok=True)
        target_path = target_dir / filename
        with target_path.open("wb") as handle:
            handle.write(bytes(data))

        return self.register_file(
            target_path,
            kind=kind,
            display_name=display_name or filename,
            mime_type=mime_type,
            attachment_id=attachment_id,
            copy_file=False,
            description=description,

View on GitHub (pinned to 4fb2db0ea9)

Solutions

1. Encode strings: data.encode('utf-8')
2. Convert buffers: bytes(memoryview_obj) or arr.tobytes()
3. Guard the call: isinstance(data, (bytes, bytearray)) with a fallback

Example fix

# before
rec = manager.register_bytes(text_payload, mime_type="text/plain")
# after
rec = manager.register_bytes(text_payload.encode("utf-8"), mime_type="text/plain")

if not isinstance(data, (bytes, bytearray)):
    data = data.encode('utf-8') if isinstance(data, str) else bytes(data)

def is_raw_bytes(d: object) -> bool:
    return isinstance(d, (bytes, bytearray))

try:
    rec = manager.register_bytes(data)
except TypeError:
    rec = manager.register_bytes(coerce_to_bytes(data))

When it happens

Trigger: Calling register_bytes with a str (e.g. forgetting .encode()), a memoryview, a generator, or None; passing an already-decoded JSON object.

Common situations: Reading text from an API response and passing it directly; numpy/array buffer passed without tobytes(); optional data parameter defaulting to None on a missing upload.

Related errors

• memory store {attachment.name} not found
• Failed to persist attachment '{attachment.name or attachment
• Workspace or node context missing for attachment persistence
• Attachment missing data for persistence
• Attachment source not found: {source}

AI-assisted analysis of OpenBMB/ChatDev@4fb2db0ea9 (2026-08-27). Data as JSON: /api/errors/8c2feaa8c65e83da. Report an issue: GitHub.