fix(tts): generate audio when block streaming drops final reply

When block streaming succeeds, final replies are dropped but TTS was only
applied to final replies. Fix by accumulating block text during streaming
and generating TTS-only audio after streaming completes.

Also:
- Change truncate vs skip behavior when summary OFF (now truncates)
- Align TTS limits with Telegram max (4096 chars)
- Improve /tts command help messages with examples
- Add newline separator between accumulated blocks

src/auto-reply/reply/commands-tts.ts

@@ -25,11 +25,11 @@ type ParsedTtsCommand = {
 };
 
 function parseTtsCommand(normalized: string): ParsedTtsCommand | null {
-  // Accept `/tts` and `/tts <action> [args]` as a single control surface.
-  if (normalized === "/tts") return { action: "status", args: "" };
+  // Accept `/tts <action> [args]` - return null for `/tts` alone to trigger inline menu.
+  if (normalized === "/tts") return null;
   if (!normalized.startsWith("/tts ")) return null;
   const rest = normalized.slice(5).trim();
-  if (!rest) return { action: "status", args: "" };
+  if (!rest) return null;
   const [action, ...tail] = rest.split(/\s+/);
   return { action: action.toLowerCase(), args: tail.join(" ").trim() };
 }