feat(telegram): buffer audio blocks for [[audio_as_voice]] tag support

- Add [[audio_as_voice]] detection to splitMediaFromOutput()
- Pass audioAsVoice through onBlockReply callback chain
- Buffer audio blocks during streaming, flush at end with correct flag
- Non-audio media still streams immediately
- Fix: emit payloads with audioAsVoice flag even if text is empty

Co-authored-by: Manuel Hettich <17690367+ManuelHettich@users.noreply.github.com>

src/agents/pi-embedded-runner.ts

@@ -1044,6 +1044,7 @@ export async function runEmbeddedPiAgent(params: {
   onBlockReply?: (payload: {
     text?: string;
     mediaUrls?: string[];
+    audioAsVoice?: boolean;
   }) => void | Promise<void>;
   blockReplyBreak?: "text_end" | "message_end";
   blockReplyChunking?: BlockReplyChunking;
@@ -1640,6 +1641,7 @@ export async function runEmbeddedPiAgent(params: {
             text: string;
             media?: string[];
             isError?: boolean;
+            audioAsVoice?: boolean;
           }> = [];
 
           const errorText = lastAssistant
@@ -1656,10 +1658,10 @@ export async function runEmbeddedPiAgent(params: {
           if (inlineToolResults) {
             for (const { toolName, meta } of toolMetas) {
               const agg = formatToolAggregate(toolName, meta ? [meta] : []);
-              const { text: cleanedText, mediaUrls } =
+              const { text: cleanedText, mediaUrls, audioAsVoice } =
                 splitMediaFromOutput(agg);
               if (cleanedText)
-                replyItems.push({ text: cleanedText, media: mediaUrls });
+                replyItems.push({ text: cleanedText, media: mediaUrls, audioAsVoice });
             }
           }
 
@@ -1678,18 +1680,31 @@ export async function runEmbeddedPiAgent(params: {
               ? [fallbackAnswerText]
               : [];
           for (const text of answerTexts) {
-            const { text: cleanedText, mediaUrls } = splitMediaFromOutput(text);
-            if (!cleanedText && (!mediaUrls || mediaUrls.length === 0))
+            const { text: cleanedText, mediaUrls, audioAsVoice } =
+              splitMediaFromOutput(text);
+            if (
+              !cleanedText &&
+              (!mediaUrls || mediaUrls.length === 0) &&
+              !audioAsVoice
+            )
               continue;
-            replyItems.push({ text: cleanedText, media: mediaUrls });
+            replyItems.push({ text: cleanedText, media: mediaUrls, audioAsVoice });
           }
 
+          // Check if any replyItem has audioAsVoice tag - if so, apply to all media payloads
+          const hasAudioAsVoiceTag = replyItems.some(
+            (item) => item.audioAsVoice,
+          );
+
           const payloads = replyItems
             .map((item) => ({
               text: item.text?.trim() ? item.text.trim() : undefined,
               mediaUrls: item.media?.length ? item.media : undefined,
               mediaUrl: item.media?.[0],
               isError: item.isError,
+              // Apply audioAsVoice to media payloads if tag was found anywhere in response
+              audioAsVoice:
+                item.audioAsVoice || (hasAudioAsVoiceTag && item.media?.length),
             }))
             .filter(
               (p) =>

src/agents/pi-embedded-subscribe.ts

@@ -262,6 +262,7 @@ export function subscribeEmbeddedPiSession(params: {
   onBlockReply?: (payload: {
     text?: string;
     mediaUrls?: string[];
+    audioAsVoice?: boolean;
   }) => void | Promise<void>;
   blockReplyBreak?: "text_end" | "message_end";
   blockReplyChunking?: BlockReplyChunking;
@@ -436,11 +437,13 @@ export function subscribeEmbeddedPiSession(params: {
     lastBlockReplyText = chunk;
     assistantTexts.push(chunk);
     if (!params.onBlockReply) return;
-    const { text: cleanedText, mediaUrls } = splitMediaFromOutput(chunk);
-    if (!cleanedText && (!mediaUrls || mediaUrls.length === 0)) return;
+    const { text: cleanedText, mediaUrls, audioAsVoice } = splitMediaFromOutput(chunk);
+    // Skip empty payloads, but always emit if audioAsVoice is set (to propagate the flag)
+    if (!cleanedText && (!mediaUrls || mediaUrls.length === 0) && !audioAsVoice) return;
     void params.onBlockReply({
       text: cleanedText,
       mediaUrls: mediaUrls?.length ? mediaUrls : undefined,
+      audioAsVoice,
     });
   };
 
@@ -859,12 +862,18 @@ export function subscribeEmbeddedPiSession(params: {
                 );
               } else {
                 lastBlockReplyText = text;
-                const { text: cleanedText, mediaUrls } =
+                const { text: cleanedText, mediaUrls, audioAsVoice } =
                   splitMediaFromOutput(text);
-                if (cleanedText || (mediaUrls && mediaUrls.length > 0)) {
+                // Emit if there's content OR audioAsVoice flag (to propagate the flag)
+                if (
+                  cleanedText ||
+                  (mediaUrls && mediaUrls.length > 0) ||
+                  audioAsVoice
+                ) {
                   void onBlockReply({
                     text: cleanedText,
                     mediaUrls: mediaUrls?.length ? mediaUrls : undefined,
+                    audioAsVoice,
                   });
                 }
               }