config file, executable, tray icon

agent-mAId/config.json

@@ -0,0 +1,8 @@
+{
+    "api_key": "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",
+    "kb_key": "ctrl",
+    "mouse_btn": "left",
+    "model": "distil-whisper-large-v3-en", 
+    "language":"en" // whisper-large-v3 or distil-whisper-large-v3-en
+  }
+  

agent-mAId/main.py

@@ -12,25 +12,39 @@ from pystray import MenuItem as item
 from PIL import Image
 import ctypes
 import time
+import json5
+
+
+# Load configuration from config.json
+def load_config():
+    config_path = os.path.join(os.path.dirname(__file__), 'config.json')
+    with open(config_path, 'r') as config_file:
+        return json5.load(config_file)
+    
+# Load the config
+config = load_config()
+# Extract API key and button from the config file
+API_KEY = config['api_key']
+KB_KEY = config['kb_key']
+MOUSE_BTN = config['mouse_btn']
+MODEL = config['model']
 
 # Constants
-API_KEY = "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"  # Make sure to use your actual API key
-BUTTON = 'ctrl'  # The keyboard button to listen for
 AUTO_START_PATH = os.path.expanduser(r"~\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\Startup")  # For autostart
 
 # Initialize the Groq client
 client = Groq(api_key=API_KEY)
 
 def record_audio(filename):
-    """Records audio when BUTTON or mouse button is pressed."""
+    """Records audio when key and mouse button is pressed."""
     audio = pyaudio.PyAudio()
-    stream = audio.open(format=pyaudio.paInt16, channels=1, rate=44100, input=True, frames_per_buffer=1024)
+    stream = audio.open(format=pyaudio.paInt16, channels=1, rate=16000, input=True, frames_per_buffer=1024)
     
     frames = []
     print("Recording...")
 
     # Record while button or mouse is pressed
-    while keyboard.is_pressed(BUTTON) and mouse.is_pressed(button='left'):
+    while keyboard.is_pressed(KB_KEY) and mouse.is_pressed(button=MOUSE_BTN):
         data = stream.read(1024)
         frames.append(data)
     
@@ -52,8 +66,9 @@ def transcribe_audio(filename):
     with open(filename, "rb") as file:
         transcription = client.audio.transcriptions.create(
             file=(filename, file.read()),  
-            model="distil-whisper-large-v3-en",  
+            model=MODEL, #"distil-whisper-large-v3-en",  
             prompt="Specify context or spelling",  
+            language=config['language'],
             response_format="json",  
             temperature=0.0  
         )
@@ -67,7 +82,7 @@ def simulate_keypress(text):
 def add_to_autostart():
     """Registers the app to autostart on login."""
     script_path = os.path.abspath(__file__)
-    shortcut_path = os.path.join(AUTO_START_PATH, "MyApp.lnk")
+    shortcut_path = os.path.join(AUTO_START_PATH, "mAId.lnk")
     
     # Use ctypes to create the shortcut (this is Windows specific)
     shell = ctypes.windll.shell32
@@ -79,27 +94,31 @@ def quit_app(icon):
     icon.stop()
     sys.exit()
 
+
 def setup_tray_icon():
     """Setup system tray icon and menu."""
-    icon_image = Image.new('RGB', (64, 64), color=(255, 0, 0))  # Red icon as an example
+    #icon_image = Image.new('RGB', (64, 64), color=(255, 0, 0))  # Red icon as an example
+    icon_image = Image.open('mic.webp')
 
     menu = (
         item('Register to Autostart', add_to_autostart),
         item('Exit', lambda: quit_app(icon))
     )
     
-    icon = pystray.Icon("my_app", icon_image, menu=pystray.Menu(*menu))
+    icon = pystray.Icon("mAId", icon_image, menu=pystray.Menu(*menu))
     icon.run()
 
+
+
 def main_loop():
     """Continuously listen for key or mouse press and transcribe audio."""
     filename = "output.wav"
     
     while True:
-        print("Waiting for button or mouse press...")
+        print("Waiting for key and mouse press...")
         
-        # Wait for button or mouse press
-        while not (keyboard.is_pressed(BUTTON) or mouse.is_pressed(button='left')):
+        # Wait for KB_KEY or mouse press
+        while not (keyboard.is_pressed(KB_KEY) and mouse.is_pressed(button=MOUSE_BTN)):
             time.sleep(0.1)  # Small sleep to avoid busy-waiting
         
         # Record audio

agent-mAId/readme.md

@@ -0,0 +1,3 @@
+# to install as an app:
+pip install pyinstaller
+pyinstaller --onefile main.py

config.json

@@ -1,4 +1,12 @@
+//C:\Users\popov\.continue\config.json
 {
+    "models": [    {
+        "title": "local ollama> yi-coder",
+        "provider": "ollama",
+        "model": "yi-coder:9b",
+        "apiBase": "http://localhost:11434"
+      }
+    ],
     "tabAutocompleteModel": {
         "title": "Tab Autocomplete Model",
         "provider": "ollama",
@@ -8,6 +16,7 @@
 }
 
 
+
 // original:   "tabAutocompleteModel": {
 //     "title": "Starcoder 3b",
 //     "provider": "ollama",