Handle media files (voice messages, audio, images, documents) on Telegram Channel

nanobot/agent/tools/message.py

@@ -52,6 +52,11 @@ class MessageTool(Tool):
                 "chat_id": {
                     "type": "string",
                     "description": "Optional: target chat/user ID"
+                },
+                "media": {
+                    "type": "array",
+                    "items": {"type": "string"},
+                    "description": "Optional: list of file paths to attach (images, audio, documents)"
                 }
             },
             "required": ["content"]
@@ -62,6 +67,7 @@ class MessageTool(Tool):
         content: str, 
         channel: str | None = None, 
         chat_id: str | None = None,
+        media: list[str] | None = None,
         **kwargs: Any
     ) -> str:
         channel = channel or self._default_channel
@@ -76,11 +82,13 @@ class MessageTool(Tool):
         msg = OutboundMessage(
             channel=channel,
             chat_id=chat_id,
-            content=content
+            content=content,
+            media=media or []
         )
         
         try:
             await self._send_callback(msg)
-            return f"Message sent to {channel}:{chat_id}"
+            media_info = f" with {len(media)} attachments" if media else ""
+            return f"Message sent to {channel}:{chat_id}{media_info}"
         except Exception as e:
             return f"Error sending message: {str(e)}"

nanobot/channels/telegram.py

@@ -198,6 +198,18 @@ class TelegramChannel(BaseChannel):
             await self._app.shutdown()
             self._app = None
     
+    def _get_media_type(self, path: str) -> str:
+        """Guess media type from file extension."""
+        path = path.lower()
+        if path.endswith(('.jpg', '.jpeg', '.png', '.gif', '.webp')):
+            return "photo"
+        elif path.endswith('.ogg'):
+            return "voice"
+        elif path.endswith(('.mp3', '.m4a', '.wav', '.aac')):
+            return "audio"
+        else:
+            return "document"
+
     async def send(self, msg: OutboundMessage) -> None:
         """Send a message through Telegram."""
         if not self._app:
@@ -212,16 +224,50 @@ class TelegramChannel(BaseChannel):
             logger.error(f"Invalid chat_id: {msg.chat_id}")
             return
 
-        for chunk in _split_message(msg.content):
+        # Handle media files
-            try:
+        if msg.media:
-                html = _markdown_to_telegram_html(chunk)
+            for media_path in msg.media:
-                await self._app.bot.send_message(chat_id=chat_id, text=html, parse_mode="HTML")
-            except Exception as e:
-                logger.warning(f"HTML parse failed, falling back to plain text: {e}")
                 try:
-                    await self._app.bot.send_message(chat_id=chat_id, text=chunk)
+                    media_type = self._get_media_type(media_path)
-                except Exception as e2:
+                    
-                    logger.error(f"Error sending Telegram message: {e2}")
+                    # Determine caption (only for first media or if explicitly set, 
+                    # but here we keep it simple: content is sent separately if media is present
+                    # to avoid length issues, unless we want to attach it to the first media)
+                    # For simplicity: send media first, then text if present.
+                    # Or: if single media, attach text as caption.
+                    
+                    # Let's attach content as caption to the last media if single, 
+                    # otherwise send text separately.
+                    
+                    with open(media_path, 'rb') as f:
+                        if media_type == "photo":
+                            await self._app.bot.send_photo(chat_id=chat_id, photo=f)
+                        elif media_type == "voice":
+                            await self._app.bot.send_voice(chat_id=chat_id, voice=f)
+                        elif media_type == "audio":
+                            await self._app.bot.send_audio(chat_id=chat_id, audio=f)
+                        else:
+                            await self._app.bot.send_document(chat_id=chat_id, document=f)
+                            
+                except Exception as e:
+                    logger.error(f"Failed to send media {media_path}: {e}")
+                    await self._app.bot.send_message(
+                        chat_id=chat_id, 
+                        text=f"[Failed to send file: {media_path}]"
+                    )
+
+        # Send text content if present
+        if msg.content and msg.content != "[empty message]":
+            for chunk in _split_message(msg.content):
+                try:
+                    html = _markdown_to_telegram_html(chunk)
+                    await self._app.bot.send_message(chat_id=chat_id, text=html, parse_mode="HTML")
+                except Exception as e:
+                    logger.warning(f"HTML parse failed, falling back to plain text: {e}")
+                    try:
+                        await self._app.bot.send_message(chat_id=chat_id, text=chunk)
+                    except Exception as e2:
+                        logger.error(f"Error sending Telegram message: {e2}")
     
     async def _on_start(self, update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
         """Handle /start command."""