Merge branch 'main' of https://github.com/KillianLucas/open-interpreter

docs/guides/advanced-terminal-usage.mdx

@@ -4,8 +4,13 @@ title: Advanced Terminal Usage
 
 Magic commands can be used to control the interpreter's behavior in interactive mode:
 
-- `%verbose [true/false]`: Toggle verbose mode
-- `%reset`: Reset the current session
-- `%undo`: Remove the last message and its response
-- `%save_message [path]`: Save messages to a JSON file
-- `%load_message [path]`: Load messages from a JSON file
+- `%% [shell commands, like ls or cd]`: Run commands in Open Interpreter's shell instance
+- `%verbose [true/false]`: Toggle verbose mode. Without arguments or with 'true', it enters verbose mode. With 'false', it exits verbose mode.
+- `%reset`: Reset the current session.
+- `%undo`: Remove previous messages and its response from the message history.
+- `%save_message [path]`: Saves messages to a specified JSON path. If no path is provided, it defaults to 'messages.json'.
+- `%load_message [path]`: Loads messages from a specified JSON path. If no path is provided, it defaults to 'messages.json'.
+- `%tokens [prompt]`: EXPERIMENTAL: Calculate the tokens used by the next request based on the current conversation's messages and estimate the cost of that request; optionally provide a prompt to also calulate the tokens used by that prompt and the total amount of tokens that will be sent with the next request.
+- `%info`: Show system and interpreter information.
+- `%help`: Show this help message.
+- `%jupyter`: Export the current session to a Jupyter notebook file (.ipynb) to the Downloads folder.

docs/usage/terminal/magic-commands.mdx

@@ -4,8 +4,12 @@ title: Magic Commands
 
 Magic commands can be used to control the interpreter's behavior in interactive mode:
 
-- `%verbose [true/false]`: Toggle verbose mode
+- `%% [commands]`: Run commands in system shell
+- `%verbose [true/false]`: Toggle verbose mode. Without arguments or with 'true', it enters verbose mode. With 'false', it exits verbose mode.
 - `%reset`: Reset the current session
-- `%undo`: Remove the last message and its response
-- `%save_message [path]`: Save messages to a JSON file
-- `%load_message [path]`: Load messages from a JSON file
+- `%undo`: Remove previous messages and its response from the message history.
+- `%save_message [path]`: Saves messages to a specified JSON path. If no path is provided, it defaults to 'messages.json'.
+- `%load_message [path]`: Loads messages from a specified JSON path. If no path is provided, it defaults to 'messages.json'.
+- `%tokens [prompt]`: EXPERIMENTAL: Calculate the tokens used by the next request based on the current conversation's messages and estimate the cost of that request; optionally provide a prompt to also calulate the tokens used by that prompt and the total amount of tokens that will be sent with the next request.
+- `%info`: Show system and interpreter information.
+- `%help`: Show this help message.

interpreter/core/computer/browser/browser.py

@@ -6,6 +6,9 @@ def __init__(self, computer):
         self.computer = computer
 
     def search(self, query):
+        """
+        Searches the web for the specified query and returns the results.
+        """
         response = requests.get(
             f'{self.computer.api_base.strip("/")}/browser/search', params={"q": query}
         )

interpreter/core/computer/clipboard/clipboard.py

@@ -17,13 +17,22 @@ def __init__(self, computer):
             self.modifier_key = "command"
 
     def view(self):
+        """
+        Returns the current content of on the clipboard.
+        """
         return pyperclip.paste()
 
     def copy(self, text=None):
+        """
+        Copies the given text to the clipboard.
+        """
         if text is not None:
             pyperclip.copy(text)
         else:
             self.computer.keyboard.hotkey(self.modifier_key, "c")
 
     def paste(self):
+        """
+        Pastes the current content of the clipboard.
+        """
         self.computer.keyboard.hotkey(self.modifier_key, "v")

interpreter/core/computer/computer.py

@@ -49,8 +49,8 @@ def run(self, *args, **kwargs):
 
     def exec(self, code):
         """
-        It has hallucinated this.
         Shortcut for computer.terminal.run("shell", code)
+        It has hallucinated this.
         """
         return self.terminal.run("shell", code)
 

interpreter/core/computer/display/display.py

@@ -33,9 +33,15 @@ def __init__(self, computer):
             pass
 
     def size(self):
+        """
+        Returns the current screen size as a tuple (width, height).
+        """
         return pyautogui.size()
 
     def center(self):
+        """
+        Calculates and returns the center point of the screen as a tuple (x, y).
+        """
         return self.width // 2, self.height // 2
 
     def view(self, show=True, quadrant=None):
@@ -48,6 +54,9 @@ def view(self, show=True, quadrant=None):
     #     return get_active_window()
 
     def screenshot(self, show=True, quadrant=None, active_app_only=False):
+        """
+       Shows you what's on the screen by taking a screenshot of the entire screen or a specified quadrant. Returns a `pil_image` `in case you need it (rarely). **You almost always want to do this first!** 
+        """
         time.sleep(2)
         if not self.computer.emit_images:
             text = self.get_text_as_list_of_lists()
@@ -110,7 +119,9 @@ def screenshot(self, show=True, quadrant=None, active_app_only=False):
         return screenshot
 
     def find_text(self, text, screenshot=None):
-        # Take a screenshot
+        """
+        Searches for specified text within a screenshot or the current screen if no screenshot is provided.
+        """
         if screenshot == None:
             screenshot = self.screenshot(show=False)
 
@@ -140,7 +151,9 @@ def find_text(self, text, screenshot=None):
         ]  # Have it deliver the text properly soon.
 
     def get_text_as_list_of_lists(self, screenshot=None):
-        # Take a screenshot
+        """
+        Extracts and returns text from a screenshot or the current screen as a list of lists, each representing a line of text.
+        """
         if screenshot == None:
             screenshot = self.screenshot(show=False)
 
@@ -172,6 +185,9 @@ def get_text_as_list_of_lists(self, screenshot=None):
 
     # locate text should be moved here as well!
     def find_icon(self, query, screenshot=None):
+        """
+        Locates an icon on the screen and returns its coordinates.
+        """
         message = format_to_recipient(
             "Locating this icon will take ~30 seconds. We're working on speeding this up.",
             recipient="user",
@@ -201,5 +217,5 @@ def find_icon(self, query, screenshot=None):
         except Exception as e:
             raise Exception(
                 str(e)
-                + "\n\nIcon locating API not avaliable, or we were unable to find the icon. Please try another method to find this icon."
+                + "\n\nIcon locating API not available, or we were unable to find the icon. Please try another method to find this icon."
             )

interpreter/core/computer/files/files.py

@@ -9,11 +9,14 @@ def __init__(self, computer):
 
     def search(self, *args, **kwargs):
         """
-        AI Filesystem Search
+        Search the filesystem for the given query.
         """
         return search(*args, **kwargs)
 
     def edit(self, path, original_text, replacement_text):
+        """
+        Edits a file on the filesystem, replacing the original text with the replacement text.
+        """
         with open(path, "r") as file:
             filedata = file.read()
 
@@ -32,6 +35,9 @@ def edit(self, path, original_text, replacement_text):
 
 
 def get_close_matches_in_text(original_text, filedata, n=3):
+    """
+    Returns the closest matches to the original text in the content of the file.
+    """
     words = filedata.split()
     original_words = original_text.split()
     len_original = len(original_words)

interpreter/core/computer/keyboard/keyboard.py

@@ -17,11 +17,7 @@ def __init__(self, computer):
 
     def write(self, text, interval=None, **kwargs):
         """
-        Type out a string of characters.
-
-        Args:
-            text (str): The string to be typed out.
-            interval (int or float, optional): The delay between pressing each character key. Defaults to 0.1.
+        Type out a string of characters. 
         """
         time.sleep(0.15)
 
@@ -68,11 +64,6 @@ def press(self, *args, presses=1, interval=0.1):
 
         If keys is a string, it is treated as a single key and is pressed the number of times specified by presses.
         If keys is a list, each key in the list is pressed once.
-
-        Args:
-            keys (str or list): The key(s) to be pressed.
-            presses (int, optional): The number of times to press the key. Defaults to 1.
-            interval (float, optional): The delay between each key press. Defaults to 0.1.
         """
         time.sleep(0.15)
         pyautogui.press(keys, presses=presses, interval=interval)
@@ -81,9 +72,6 @@ def press(self, *args, presses=1, interval=0.1):
     def hotkey(self, *args, interval=0.1):
         """
         Press a sequence of keys in the order they are provided, and then release them in reverse order.
-
-        Args:
-            *args: The keys to be pressed.
         """
         time.sleep(0.15)
         modifiers = ["command", "option", "alt", "ctrl", "shift"]
@@ -121,21 +109,15 @@ def hotkey(self, *args, interval=0.1):
 
     def down(self, key):
         """
-        Simulate the pressing down of a key.
-
-        Args:
-            key (str): The key to be pressed down.
+        Press down a key.
         """
         time.sleep(0.15)
         pyautogui.keyDown(key)
         time.sleep(0.15)
 
     def up(self, key):
         """
-        Simulate the releasing of a key.
-
-        Args:
-            key (str): The key to be released.
+        Release a key.
         """
         time.sleep(0.15)
         pyautogui.keyUp(key)

interpreter/core/computer/mouse/mouse.py

@@ -19,6 +19,9 @@ def __init__(self, computer):
         self.computer = computer
 
     def scroll(self, clicks):
+        """
+        Scrolls the mouse wheel up or down the specified number of clicks.
+        """
         pyautogui.scroll(clicks)
 
     def position(self):
@@ -36,6 +39,10 @@ def position(self):
             )
 
     def move(self, *args, x=None, y=None, icon=None, text=None, screenshot=None):
+        """
+        Moves the mouse to specified coordinates, an icon, or text.
+        """
+        screenshot = None
         if len(args) > 1:
             raise ValueError(
                 "Too many positional arguments provided. To move/click specific coordinates, use kwargs (x=x, y=y).\n\nPlease take a screenshot with computer.display.view() to find text/icons to click, then use computer.mouse.click(text) or computer.mouse.click(icon=description_of_icon) if at all possible. This is **significantly** more accurate than using coordinates. Specifying (x=x, y=y) is highly likely to fail. Specifying ('text to click') is highly likely to succeed."
@@ -216,29 +223,47 @@ def move(self, *args, x=None, y=None, icon=None, text=None, screenshot=None):
         smooth_move_to(x, y)
 
     def click(self, *args, button="left", clicks=1, interval=0.1, **kwargs):
+        """
+        Clicks the mouse at the specified coordinates, icon, or text.
+        """
         if args or kwargs:
             self.move(*args, **kwargs)
         pyautogui.click(button=button, clicks=clicks, interval=interval)
 
     def double_click(self, *args, button="left", interval=0.1, **kwargs):
+        """
+        Double-clicks the mouse at the specified coordinates, icon, or text.
+        """
         if args or kwargs:
             self.move(*args, **kwargs)
         pyautogui.doubleClick(button=button, interval=interval)
 
     def triple_click(self, *args, button="left", interval=0.1, **kwargs):
+        """
+        Triple-clicks the mouse at the specified coordinates, icon, or text.
+        """
         if args or kwargs:
             self.move(*args, **kwargs)
         pyautogui.tripleClick(button=button, interval=interval)
 
     def right_click(self, *args, **kwargs):
+        """
+        Right-clicks the mouse at the specified coordinates, icon, or text.
+        """
         if args or kwargs:
             self.move(*args, **kwargs)
         pyautogui.rightClick()
 
     def down(self):
+        """
+        Presses the mouse button down.
+        """
         pyautogui.mouseDown()
 
     def up(self):
+        """
+        Releases the mouse button.
+        """
         pyautogui.mouseUp()
 
 

interpreter/core/computer/os/os.py

@@ -7,6 +7,9 @@ def __init__(self, computer):
         self.computer = computer
 
     def get_selected_text(self):
+        """
+        Returns the currently selected text.
+        """
         # Store the current clipboard content
         current_clipboard = self.computer.clipboard.view()
         # Copy the selected text to clipboard
@@ -18,6 +21,9 @@ def get_selected_text(self):
         return selected_text
 
     def notify(self, text):
+        """
+        Displays a notification on the computer.
+        """
         try:
             title = "Open Interpreter"
 

interpreter/core/llm/llm.py

@@ -172,10 +172,11 @@ def run(self, messages):
         }
 
         # Optional inputs
-        if self.api_base:
-            params["api_base"] = self.api_base
         if self.api_key:
             params["api_key"] = self.api_key
+        if self.api_base:
+            params["api_base"] = self.api_base
+            params["custom_llm_provider"] = "openai"
         if self.api_version:
             params["api_version"] = self.api_version
         if self.max_tokens: