can you make me a limbus macro that looks at the screen and picks actions based on what is displayed please

Creating a macro that interacts with the screen and makes decisions based on what it sees can be a complex task, especially since it involves image recognition and possibly interfacing with the game's API or memory. However, I can guide you through a simple conceptual approach using pseudocode, which you can adapt based on the tools or programming language you're using.

### Conceptual Approach

1. **Screen Capture**: You need a way to capture the contents of the game screen.
2. **Image Recognition**: Use a library to analyze the captured image to identify what is displayed.
3. **Action Decision**: Based on the identified content, decide what actions to take.
4. **Execution of Actions**: Simulate key presses or mouse clicks to perform the chosen actions.

### Example Pseudocode

Here's an example of how you might structure your macro:

```python
import time
import cv2  # OpenCV for image processing
import numpy as np
import pyautogui  # For simulating keyboard and mouse input

# Load reference images for what you want to recognize
action_image_1 = cv2.imread('action1.png')
action_image_2 = cv2.imread('action2.png')

def capture_screen():
    # Capture the screen and return an image
    screen = pyautogui.screenshot()
    return cv2.cvtColor(np.array(screen), cv2.COLOR_RGB2BGR)

def find_action(screen_image):
    # Check for action_image_1 in screen_image
    result1 = cv2.matchTemplate(screen_image, action_image_1, cv2.TM_CCOEFF_NORMED)
    result2 = cv2.matchTemplate(screen_image, action_image_2, cv2.TM_CCOEFF_NORMED)

    threshold = 0.8  # Adjust this threshold based on your needs

    if np.any(result1 >= threshold):
        return 'action1'  # Return identifier for action 1
    elif np.any(result2 >= threshold):
        return 'action2'  # Return identifier for action 2
    return None

def perform_action(action):
    if action == 'action1':
        pyautogui.press('a')  # Replace 'a' with the actual key you want to simulate
    elif action == 'action2':
        pyautogui.press('b')  # Replace 'b' with the actual key you want to simulate

while True:
    screen = capture_screen()
    action = find_action(screen)

    if action:
        perform_action(action)
    
    # Wait for a short time before the next iteration
    time.sleep(0.5)  # Adjust as needed
```

### Important Notes

1. **Dependencies**: You'll need to install the required libraries (`opencv-python`, `numpy`, and `pyautogui`).
2. **Images**: You'll need to provide images (`action1.png`, `action2.png`) that the program can use to recognize what is on the screen.
3. **Legality**: Check the game's terms of service to ensure that using a macro or automation tool does not violate any rules.
4. **Environment**: This pseudocode is for a generic Python environment and may need adjustments if you're working in a different context or using a different programming language.

This is a basic outline to help you get started. Depending on your specific use case and environment, further customizations and improvements might be necessary.