diff --git a/cli.py b/cli.py
index 5a3079b99..d16954e4b 100755
--- a/cli.py
+++ b/cli.py
@@ -3825,7 +3825,8 @@ class HermesCLI:
         tts_status = " (TTS enabled)" if self._voice_tts else ""
         try:
             from hermes_cli.config import load_config
-            _ptt_key = load_config().get("voice", {}).get("push_to_talk_key", "c-b")
+            _raw_ptt = load_config().get("voice", {}).get("record_key", "ctrl+b")
+            _ptt_key = _raw_ptt.lower().replace("ctrl+", "c-").replace("alt+", "a-")
         except Exception:
             _ptt_key = "c-b"
         _ptt_display = _ptt_key.replace("c-", "Ctrl+").upper()
@@ -4818,11 +4819,13 @@ class HermesCLI:
             self._should_exit = True
             event.app.exit()
 
-        # Voice push-to-talk key: configurable via config.yaml (voice.push_to_talk_key)
+        # Voice push-to-talk key: configurable via config.yaml (voice.record_key)
         # Default: Ctrl+B (avoids conflict with Ctrl+R readline reverse-search)
+        # Config uses "ctrl+b" format; prompt_toolkit expects "c-b" format.
         try:
             from hermes_cli.config import load_config
-            _voice_key = load_config().get("voice", {}).get("push_to_talk_key", "c-b")
+            _raw_key = load_config().get("voice", {}).get("record_key", "ctrl+b")
+            _voice_key = _raw_key.lower().replace("ctrl+", "c-").replace("alt+", "a-")
         except Exception:
             _voice_key = "c-b"
 
diff --git a/tools/voice_mode.py b/tools/voice_mode.py
index 27de0fc55..2c3a168bd 100644
--- a/tools/voice_mode.py
+++ b/tools/voice_mode.py
@@ -291,13 +291,20 @@ class AudioRecorder:
                             if cb:
                                 threading.Thread(target=cb, daemon=True).start()
 
-            self._stream = sd.InputStream(
-                samplerate=SAMPLE_RATE,
-                channels=CHANNELS,
-                dtype=DTYPE,
-                callback=_callback,
-            )
-            self._stream.start()
+            try:
+                self._stream = sd.InputStream(
+                    samplerate=SAMPLE_RATE,
+                    channels=CHANNELS,
+                    dtype=DTYPE,
+                    callback=_callback,
+                )
+                self._stream.start()
+            except Exception as e:
+                self._stream = None
+                raise RuntimeError(
+                    f"Failed to open audio input stream: {e}. "
+                    "Check that a microphone is connected and accessible."
+                ) from e
             self._recording = True
             logger.info("Voice recording started (rate=%d, channels=%d)", SAMPLE_RATE, CHANNELS)