talktollm 0.4.0

A Python utility for interacting with large language models (LLMs) via web automation

talktollm

A Python utility for interacting with large language models (LLMs) through a command-line interface. It leverages image recognition to automate interactions with LLM web interfaces, enabling seamless conversations and task execution.

Features

• Command-Line Interaction: Provides a simple and intuitive command-line interface for interacting with LLMs.
• Automated Image Recognition: Employs image recognition techniques (via optimisewait) to identify and interact with elements on the LLM interface. Includes fallback if optimisewait is not installed.
• Multi-LLM Support: Currently supports DeepSeek and Gemini.
• Automated Conversations: Facilitates automated conversations and task execution by simulating user interactions.
• Image Support: Allows sending images (base64 encoded) to the LLM.
• Robust Clipboard Handling: Includes configurable retry mechanisms (default 5 retries) for setting text/images to the clipboard and reading text from the clipboard to handle access errors and timing issues.
• Dynamic Image Path Management: Copies necessary recognition images to a temporary directory, ensuring they are accessible and up-to-date.
• Easy to use: Designed for simple setup and usage.

Core Functionality

The core function is talkto(llm, prompt, imagedata=None, debug=False, tabswitch=True, read_retries=5, read_delay=0.3).

Arguments:

• llm (str): The LLM name ('deepseek','gemini' or 'aistudio').
• prompt (str): The text prompt.
• imagedata (list[str] | None): Optional list of base64 encoded image strings (e.g., "data:image/png;base64,...").
• debug (bool): Enable detailed console output. Defaults to False.
• tabswitch (bool): Switch focus back to the previous window after closing the LLM tab. Defaults to True.
• read_retries (int): Number of attempts to read the final response from the clipboard. Defaults to 5.
• read_delay (float): Delay in seconds between clipboard read attempts. Defaults to 0.3.

Steps:

1. Validates the LLM name.
2. Sets up image paths for optimisewait using set_image_path.
3. Opens the LLM's website in a new browser tab.
4. Waits and clicks the message input area using optimiseWait('message', clicks=2).
5. If imagedata is provided:
   • Iterates through images.
   • Sets each image to the clipboard using set_clipboard_image (with retries).
   • Pastes the image (Ctrl+V).
   • Waits for potential upload (sleep(7)).
6. Sets the prompt text to the clipboard using set_clipboard (with retries).
7. Pastes the prompt (Ctrl+V).
8. Waits and clicks the 'run' button using optimiseWait('run').
9. Waits for the response generation, using optimiseWait('copy') as an indicator that the response is ready and the copy button is visible.
10. Waits briefly (sleep(0.5)) after optimiseWait('copy') clicks the copy button.
11. Closes the browser tab (Ctrl+W).
12. Switches focus back if tabswitch is True (Alt+Tab).
13. Attempts to read the LLM's response from the clipboard with retry logic (read_retries, read_delay).
14. Returns the retrieved text response, or an empty string if reading fails.

Helper Functions

Clipboard Handling:

• set_clipboard(text: str, retries: int = 5, delay: float = 0.2): Sets text to the clipboard, handling CF_UNICODETEXT. Retries on common access errors (winerror 5 or 1418).
• set_clipboard_image(image_data: str, retries: int = 5, delay: float = 0.2): Sets a base64 encoded image to the clipboard (CF_DIB format). Decodes, converts to BMP, and retries on common access errors.

Image Path Management:

• set_image_path(llm: str, debug: bool = False): Orchestrates copying images.
• copy_images_to_temp(llm: str, debug: bool = False): Copies necessary .png images for the specified llm from the package's images/<llm> directory to a temporary location (%TEMP%\\talktollm_images\\<llm>). Creates the temporary directory if needed and only copies if the source file is newer or the destination doesn't exist. Sets the optimisewait autopath. Includes error handling for missing package resources.

Installation

pip install talktollm

Note: Requires optimisewait for image recognition. Install separately if needed (pip install optimisewait).

Usage

Here are some examples of how to use talktollm.

Example 1: Simple Text Prompt

Send a basic text prompt to Gemini.

import talktollm

prompt_text = "Explain quantum entanglement in simple terms."
response = talktollm.talkto('gemini', prompt_text)
print("--- Simple Gemini Response ---")
print(response)

Example 2: Text Prompt with Debugging

Send a text prompt and enable debugging output to see more details about the process.

import talktollm

prompt_text = "What are the main features of Python 3.12?"
response = talktollm.talkto('deepseek', prompt_text, debug=True)
print("--- DeepSeek Debug Response ---")
print(response)

Example 3: Preparing Image Data

Load an image file, encode it in base64, and format it correctly for the imagedata argument.

import base64
import io
from PIL import Image

# Load your image (replace 'path/to/your/image.png' with the actual path)
try:
    with open("path/to/your/image.png", "rb") as image_file:
        # Encode to base64
        encoded_string = base64.b64encode(image_file.read()).decode('utf-8')
        # Format as a data URI
        image_data_uri = f"data:image/png;base64,{encoded_string}"
        print("Image prepared successfully!")
        # You can now pass [image_data_uri] to the imagedata parameter
except FileNotFoundError:
    print("Error: Image file not found. Please check the path.")
    image_data_uri = None
except Exception as e:
    print(f"Error processing image: {e}")
    image_data_uri = None

# This 'image_data_uri' variable holds the string needed for the next example

Example 4: Text and Image Prompt

Send a text prompt along with a prepared image to Gemini. (Assumes image_data_uri was successfully created in Example 3).

import talktollm

# Assuming image_data_uri is available from the previous example
if image_data_uri:
    prompt_text = "Describe the main subject of this image."
    response = talktollm.talkto(
        'gemini',
        prompt_text,
        imagedata=[image_data_uri], # Pass the image data as a list
        debug=True
    )
    print("--- Gemini Image Response ---")
    print(response)
else:
    print("Skipping image example because image data is not available.")

Dependencies

• pywin32: For Windows API access (clipboard).
• pyautogui: For GUI automation (keystrokes, potentially mouse if optimisewait fails).
• Pillow: For image processing (opening, converting for clipboard).
• optimisewait (Optional but Recommended): For robust image-based waiting and clicking.

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

License

MIT