Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/Harsha200105/DesktopAssistant/llms.txt

Use this file to discover all available pages before exploring further.

Jarvis can send an email entirely by voice. Say "mail" or include "email" in your command, and Jarvis will prompt you to speak the recipient address and then dictate the message body. Under the hood, commands.py opens an SMTP connection to the server configured in config.ini, authenticates, and dispatches the message — all without touching a keyboard.

​
Gmail setup

1
Enable account access for Jarvis
2
Gmail blocks plain-username/password logins from third-party apps by default. Choose one of the following options based on whether you have two-factor authentication enabled:
3
  • Without 2FA — Log in to your Google account, go to myaccount.google.com/lesssecureapps, and turn on Allow less secure apps.
  • With 2FA (recommended) — Generate a 16-character App Password at myaccount.google.com/apppasswords. Select Mail as the app and Windows Computer as the device. Copy the generated password — you will use it in the next step instead of your real account password.
    • 4
    Fill in the [EMAIL] section of config.ini
    5
    Open src/config.ini and complete the [EMAIL] block:
    6
    [EMAIL]
server = smtp.gmail.com
port = 587
username = you@gmail.com
password = yourpassword
    7
    Replace you@gmail.com with your Gmail address and yourpassword with either your account password (less-secure-apps path) or the 16-character App Password (2FA path).
    8
    Verify your microphone is working
    9
    The command_mail function calls take_command() twice — once to capture the recipient and once to capture the message body. Both calls go through the speech recogniser, so a working microphone is required unless you are running with debug = True.
    10
    Set debug = True in config.ini before your first test. Jarvis will prompt you at the terminal for the recipient and message body, letting you confirm that the SMTP credentials are correct without needing to speak.

    ​
    Voice interaction flow

    Once setup is complete, triggering the email command works like this:
    1. Say a phrase that includes "mail" — for example “send a mail” or “email”.
    2. Jarvis replies “Who is the recipient?” — speak the recipient’s full email address.
    3. Jarvis replies “What should I say?” — dictate the message body.
    4. Jarvis sends the email and confirms “Email sent!”, or says “I am unable to send your message at this moment!” if an error occurred.

    ​
    How the SMTP code works

    The command_mail function in commands.py handles the entire flow: 
    def command_mail(take_command):
    speak("Who is the recipient? ")
    recipient = take_command()

    try:
        speak("What should I say? ")
        content = take_command()

        email = config['EMAIL']
        server = smtplib.SMTP(email['server'], email['port'])
        server.ehlo()
        server.starttls()
        server.login(email['username'], email['password'])
        server.sendmail(email['username'], recipient, content)
        server.close()
        speak("Email sent!")
    except Exception:
        speak("Sorry Sir!")
        speak("I am unable to send your message at this moment!")
    The connection sequence is:
    StepMethodPurpose
    1smtplib.SMTP(server, port)Opens a plain-text connection to port 587
    2server.ehlo()Identifies the client to the server (ESMTP handshake)
    3server.starttls()Upgrades the connection to TLS encryption
    4server.login(username, password)Authenticates with your credentials
    5server.sendmail(from, to, content)Sends the email (both From and To use the configured username)
    6server.close()Closes the connection cleanly
    The config['EMAIL'] dictionary is populated at module import time via: 
    config = configparser.ConfigParser()
config.read('config.ini')

    ​
    Security considerations

    Your email password is stored in plain text inside config.ini. Anyone who can read that file can access your email account.
    • Use a dedicated App Password (Gmail) rather than your main account password — you can revoke it at any time without changing your real password.
    • Do not commit config.ini to a public git repository. Add it to your .gitignore file: 
      echo "src/config.ini" >> .gitignore
    • Restrict the file permissions so only your user account can read it: 
      icacls src\config.ini /inheritance:r /grant:r "%USERNAME%:R"
    The recipient address is captured directly from Google’s speech recognition output. Addresses that contain unusual punctuation (dots, plus signs, subdomains) may not be recognised accurately. If the recognised text looks wrong in the terminal, use debug = True mode and type the address instead.

    Build docs developers (and LLMs) love

    Get started for freeTalk to us