OpenInterpreter · KillianLucas · Jun 1, 2024 · May 3, 2024 · May 17, 2024 · May 18, 2024
diff --git a/docs/CONTRIBUTING.md b/docs/CONTRIBUTING.md
@@ -14,7 +14,7 @@ If you encounter a bug or have a feature in mind, don't hesitate to [open a new
 
 ## Philosophy
 
-This is a minimalist, **tighly scoped** project that places a premium on simplicity. We're skeptical of new extensions, integrations, and extra features. We would rather not extend the system if it adds nonessential complexity.
+This is a minimalist, **tightly scoped** project that places a premium on simplicity. We're skeptical of new extensions, integrations, and extra features. We would rather not extend the system if it adds nonessential complexity.
 
 # Contribution Guidelines
 

diff --git a/docs/ROADMAP.md b/docs/ROADMAP.md
@@ -48,14 +48,14 @@
 
 # What's in our scope?
 
-Open Interpreter contains two projects which support eachother, whose scopes are as follows:
+Open Interpreter contains two projects which support each other, whose scopes are as follows:
 
 1. `core`, which is dedicated to figuring out how to get LLMs to safely control a computer. Right now, this means creating a real-time code execution environment that language models can operate.
 2. `terminal_interface`, a text-only way for users to direct the code-running LLM running inside `core`. This includes functions for connecting the `core` to various local and hosted LLMs (which the `core` itself should not know about).
 
 # What's not in our scope?
 
-Our guiding philosphy is minimalism, so we have also decided to explicitly consider the following as **out of scope**:
+Our guiding philosophy is minimalism, so we have also decided to explicitly consider the following as **out of scope**:
 
 1. Additional functions in `core` beyond running code.
 2. More complex interactions with the LLM in `terminal_interface` beyond text (but file paths to more complex inputs, like images or video, can be included in that text).

diff --git a/docs/getting-started/setup.mdx b/docs/getting-started/setup.mdx
@@ -4,7 +4,7 @@ title: Setup
 
 ## Experimental one-line installers
 
-To try our experimental installers, open your Terminal with admin priviledges [(click here to learn how)](https://chat.openai.com/share/66672c0f-0935-4c16-ac96-75c1afe14fe3), then paste the following commands:
+To try our experimental installers, open your Terminal with admin privileges [(click here to learn how)](https://chat.openai.com/share/66672c0f-0935-4c16-ac96-75c1afe14fe3), then paste the following commands:
 
 <CodeGroup>
 
@@ -57,7 +57,7 @@ from interpreter import interpreter
 interpreter.chat()
 ```
 
-You can also pass messages to `interpreter` programatically:
+You can also pass messages to `interpreter` programmatically:
 
 ```python
 interpreter.chat("Get the last 5 BBC news headlines.")

diff --git a/docs/guides/advanced-terminal-usage.mdx b/docs/guides/advanced-terminal-usage.mdx
@@ -10,7 +10,7 @@ Magic commands can be used to control the interpreter's behavior in interactive
 - `%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.
+- `%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 calculate 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.
diff --git a/docs/guides/demos.mdx b/docs/guides/demos.mdx
@@ -42,7 +42,7 @@ OS mode using Logic Pro X to record a piano song and play it back:
 
 #### Generating images in Everart.ai
 
-Open Interpreter descibing pictures it wants to make, then creating them using OS mode:
+Open Interpreter describing pictures it wants to make, then creating them using OS mode:
 
 <iframe src="data:text/html;charset=utf-8,%0A%20%20%20%20%20%20%20%20%20%20%20%20%3Cblockquote%20class%3D%22twitter-tweet%22%20data-media-max-width%3D%22560%22%3E%0A%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%3Cp%20lang%3D%22en%22%20dir%3D%22ltr%22%3EThis%20is%20wild.%20I%20gave%20OS%20control%20to%20GPT-4%20via%20the%20latest%20update%20of%20Open%20Interpreter%20and%20now%20it%27s%20generating%20pictures%20it%20wants%20to%20see%20in%20%40everartai%20%F0%9F%A4%AF%3Cbr%3E%3Cbr%3E%0A%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20GPT%20is%20controlling%20the%20mouse%20and%20adding%20text%20in%20the%20fields%2C%20I%20am%20not%20doing%20anything.%20%0A%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%3Ca%20href%3D%22https%3A//t.co/hGgML9epEc%22%3Epic.twitter.com/hGgML9epEc%3C/a%3E%0A%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%3C/p%3E%26mdash%3B%20Pietro%20Schirano%20%28%40skirano%29%20%0A%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%3Ca%20href%3D%22https%3A//twitter.com/skirano/status/1747670816437735836%3Fref_src%3Dtwsrc%255Etfw%22%3EJanuary%2017%2C%202024%3C/a%3E%0A%20%20%20%20%20%20%20%20%20%20%20%20%3C/blockquote%3E%20%0A%20%20%20%20%20%20%20%20%20%20%20%20%3Cscript%20async%20src%3D%22https%3A//platform.twitter.com/widgets.js%22%20charset%3D%22utf-8%22%3E%3C/script%3E%0A%20%20%20%20%20%20%20%20" width="100%" height="500"></iframe>
 

diff --git a/docs/guides/multiple-instances.mdx b/docs/guides/multiple-instances.mdx
@@ -8,7 +8,7 @@ To create multiple instances, use the base class, `OpenInterpreter`:
 from interpreter import OpenInterpreter
 
 agent_1 = OpenInterpreter()
-agent_1.system_message = "This is a seperate instance."
+agent_1.system_message = "This is a separate instance."
 
 agent_2 = OpenInterpreter()
 agent_2.system_message = "This is yet another instance."

diff --git a/docs/language-models/hosted-models/aws-sagemaker.mdx b/docs/language-models/hosted-models/aws-sagemaker.mdx
@@ -65,6 +65,6 @@ Set the following environment variables [(click here to learn how)](https://chat
 
 | Environment Variable    | Description                                     | Where to Find                                                                       |
 | ----------------------- | ----------------------------------------------- | ----------------------------------------------------------------------------------- |
-| `AWS_ACCESS_KEY_ID`     | The API access key for your AWS account.        | [AWS Account Overview -> Security Credintials](https://console.aws.amazon.com/)     |
-| `AWS_SECRET_ACCESS_KEY` | The API secret access key for your AWS account. | [AWS Account Overview -> Security Credintials](https://console.aws.amazon.com/)     |
+| `AWS_ACCESS_KEY_ID`     | The API access key for your AWS account.        | [AWS Account Overview -> Security Credentials](https://console.aws.amazon.com/)     |
+| `AWS_SECRET_ACCESS_KEY` | The API secret access key for your AWS account. | [AWS Account Overview -> Security Credentials](https://console.aws.amazon.com/)     |
 | `AWS_REGION_NAME`       | The AWS region you want to use                  | [AWS Account Overview -> Navigation bar -> Region](https://console.aws.amazon.com/) |
diff --git a/docs/safety/introduction.mdx b/docs/safety/introduction.mdx
@@ -10,7 +10,7 @@ Safety is a top priority for us at Open Interpreter. Running LLM generated code
 
 - Requiring confirmation with the user before the code is actually run. This is a simple measure that can prevent a lot of accidents. It exists as another layer of protection, but can be disabled with the `--auto-run` flag if you wish.
 
-- Sandboxing code execution. Open Interpreter can be run in a sandboxed envirnoment using [Docker](/integrations/docker). This is a great way to run code without worrying about it affecting your system. Docker support is currently experimental, but we are working on making it a core feature of Open Interpreter. Another option for sandboxing is [E2B](https://e2b.dev/), which overrides the default python language with a sandboxed, hosted version of python through E2B. Follow [this guide](/integrations/e2b) to set it up.
+- Sandboxing code execution. Open Interpreter can be run in a sandboxed environment using [Docker](/integrations/docker). This is a great way to run code without worrying about it affecting your system. Docker support is currently experimental, but we are working on making it a core feature of Open Interpreter. Another option for sandboxing is [E2B](https://e2b.dev/), which overrides the default python language with a sandboxed, hosted version of python through E2B. Follow [this guide](/integrations/e2b) to set it up.
 
 ## Notice
 

diff --git a/docs/usage/python/multiple-instances.mdx b/docs/usage/python/multiple-instances.mdx
@@ -4,13 +4,13 @@ To create multiple instances, use the base class, `OpenInterpreter`:
 from interpreter import OpenInterpreter
 
 agent_1 = OpenInterpreter()
-agent_1.system_message = "This is a seperate instance."
+agent_1.system_message = "This is a separate instance."
 
 agent_2 = OpenInterpreter()
 agent_2.system_message = "This is yet another instance."
 ```
 
-For fun, you could make these instances talk to eachother:
+For fun, you could make these instances talk to each other:
 
 ```python
 def swap_roles(messages):

diff --git a/docs/usage/terminal/magic-commands.mdx b/docs/usage/terminal/magic-commands.mdx
@@ -10,6 +10,6 @@ Magic commands can be used to control the interpreter's behavior in interactive
 - `%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.
+- `%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 calculate 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.
diff --git a/interpreter/core/computer/calendar/calendar.py b/interpreter/core/computer/calendar/calendar.py
@@ -23,7 +23,7 @@
 class Calendar:
     def __init__(self, computer):
         self.computer = computer
-        # In the future, we might consider a way to use a different calender app. For now its Calendar
+        # In the future, we might consider a way to use a different calendar app. For now its Calendar
         self.calendar_app = "Calendar"
 
     def get_events(self, start_date=datetime.date.today(), end_date=None):

diff --git a/interpreter/core/computer/display/display.py b/interpreter/core/computer/display/display.py
@@ -65,7 +65,7 @@ def center(self):
 
     def info(self):
         """
-        Returns a list of all connected monitor/displays and thir information
+        Returns a list of all connected monitor/displays and their information
         """
         return get_displays()
 
@@ -145,7 +145,7 @@ def screenshot(
                     screen=screen, combine_screens=combine_screens
                 )  #  this function uses pyautogui.screenshot which works fine for all OS (mac, linux and windows)
                 message = format_to_recipient(
-                    "Taking a screenshot of the entire screen. This is not recommended. You (the language model assistant) will recieve it with low resolution.\n\nTo maximize performance, use computer.display.view(active_app_only=True). This will produce an ultra high quality image of the active application.",
+                    "Taking a screenshot of the entire screen. This is not recommended. You (the language model assistant) will receive it with low resolution.\n\nTo maximize performance, use computer.display.view(active_app_only=True). This will produce an ultra high quality image of the active application.",
                     "assistant",
                 )
                 print(message)

diff --git a/interpreter/core/computer/mouse/mouse.py b/interpreter/core/computer/mouse/mouse.py
@@ -136,7 +136,7 @@ def move(self, *args, x=None, y=None, icon=None, text=None, screenshot=None):
         elif x is not None and y is not None:
             print(
                 format_to_recipient(
-                    "Unless you have just recieved these EXACT coordinates from a computer.mouse.move or computer.mouse.click command, PLEASE take a screenshot with computer.display.view() to find TEXT OR 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.",
+                    "Unless you have just received these EXACT coordinates from a computer.mouse.move or computer.mouse.click command, PLEASE take a screenshot with computer.display.view() to find TEXT OR 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.",
                     "assistant",
                 )
             )

diff --git a/interpreter/core/computer/terminal/languages/jupyter_language.py b/interpreter/core/computer/terminal/languages/jupyter_language.py
@@ -119,7 +119,7 @@ def iopub_message_listener():
 
                 if DEBUG_MODE:
                     print("-----------" * 10)
-                    print("Message recieved:", msg["content"])
+                    print("Message received:", msg["content"])
                     print("-----------" * 10)
 
                 if (

diff --git a/interpreter/core/computer/terminal/languages/python.py b/interpreter/core/computer/terminal/languages/python.py
@@ -2,7 +2,7 @@
 
 from .jupyter_language import JupyterLanguage
 
-# Supresses a weird debugging error
+# Suppresses a weird debugging error
 os.environ["PYDEVD_DISABLE_FILE_VALIDATION"] = "1"
 # turn off colors in "terminal"
 os.environ["ANSI_COLORS_DISABLED"] = "1"

diff --git a/interpreter/core/core.py b/interpreter/core/core.py
@@ -364,7 +364,7 @@ def is_active_line_chunk(chunk):
 
                 last_flag_base = {"role": chunk["role"], "type": chunk["type"]}
 
-                # Don't add format to type: "console" flags, to accomodate active_line AND output formats
+                # Don't add format to type: "console" flags, to accommodate active_line AND output formats
                 if "format" in chunk and chunk["type"] != "console":
                     last_flag_base["format"] = chunk["format"]
 

diff --git a/interpreter/core/respond.py b/interpreter/core/respond.py
@@ -59,7 +59,7 @@ def respond(interpreter):
                     "content": force_task_completion_message,
                 }
             )
-            # Yield two newlines to seperate the LLMs reply from previous messages.
+            # Yield two newlines to separate the LLMs reply from previous messages.
             yield {"role": "assistant", "type": "message", "content": "\n\n"}
             insert_force_task_completion_message = False
 

diff --git a/interpreter/core/utils/scan_code.py b/interpreter/core/utils/scan_code.py
@@ -30,7 +30,7 @@ def scan_code(code, language, interpreter):
     # Run semgrep
     try:
         # HACK: we need to give the subprocess shell access so that the semgrep from our pyproject.toml is available
-        # the global namespace might have semgrep from guarddog installed, but guarddog is currenlty
+        # the global namespace might have semgrep from guarddog installed, but guarddog is currently
         # pinned to an old semgrep version that has issues with reading the semgrep registry
         # while scanning a single file like the temporary one we generate
         # if guarddog solves [#249](https://github.com/DataDog/guarddog/issues/249) we can change this approach a bit

diff --git a/interpreter/terminal_interface/magic_commands.py b/interpreter/terminal_interface/magic_commands.py
@@ -54,7 +54,7 @@ def handle_help(self, arguments):
         "%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",
+        "%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 calculate the tokens used by that prompt and the total amount of tokens that will be sent with the next request",
         "%help": "Show this help message.",
         "%info": "Show system and interpreter information",
         "%jupyter": "Export the conversation to a Jupyter notebook file",

diff --git a/interpreter/terminal_interface/profiles/defaults/01.py b/interpreter/terminal_interface/profiles/defaults/01.py
@@ -90,7 +90,7 @@
 You may use the `computer` module to control the user's keyboard and mouse, if the task **requires** it:
 
 ```python
-computer.display.info() # Returns a list of connected monitors/Displays and their info (x and y cordinates, width, height, width_mm, height_mm, name). Use this to verify the monitors connected before using computer.display.view() when neccessary
+computer.display.info() # Returns a list of connected monitors/Displays and their info (x and y coordinates, width, height, width_mm, height_mm, name). Use this to verify the monitors connected before using computer.display.view() when necessary
 computer.display.view() # Shows you what's on the screen (primary display by default), returns a `pil_image` `in case you need it (rarely). To get a specific display, use the parameter screen=DISPLAY_NUMBER (0 for primary monitor 1 and above for secondary monitors). **You almost always want to do this first!**
 computer.keyboard.hotkey(" ", "command") # Opens spotlight
 computer.keyboard.write("hello")

diff --git a/interpreter/terminal_interface/profiles/defaults/os.py b/interpreter/terminal_interface/profiles/defaults/os.py
@@ -36,7 +36,7 @@
 ```python
 computer.browser.search(query) # Silently searches Google for the query, returns result. The user's browser is unaffected. (does not open a browser!)
 
-computer.display.info() # Returns a list of connected monitors/Displays and their info (x and y cordinates, width, height, width_mm, height_mm, name). Use this to verify the monitors connected before using computer.display.view() when neccessary
+computer.display.info() # Returns a list of connected monitors/Displays and their info (x and y coordinates, width, height, width_mm, height_mm, name). Use this to verify the monitors connected before using computer.display.view() when necessary
 computer.display.view() # Shows you what's on the screen (primary display by default), returns a `pil_image` `in case you need it (rarely). To get a specific display, use the parameter screen=DISPLAY_NUMBER (0 for primary monitor 1 and above for secondary monitors). **You almost always want to do this first!**
 
 computer.keyboard.hotkey(" ", "command") # Opens spotlight (very useful)

diff --git a/interpreter/terminal_interface/start_terminal_interface.py b/interpreter/terminal_interface/start_terminal_interface.py
@@ -75,7 +75,7 @@ def start_terminal_interface(interpreter):
         {
             "name": "llm_supports_vision",
             "nickname": "lsv",
-            "help_text": "inform OI that your model supports vision, and can recieve vision inputs",
+            "help_text": "inform OI that your model supports vision, and can receive vision inputs",
             "type": bool,
             "action": argparse.BooleanOptionalAction,
             "attribute": {"object": interpreter.llm, "attr_name": "supports_vision"},

diff --git a/interpreter/terminal_interface/terminal_interface.py b/interpreter/terminal_interface/terminal_interface.py
@@ -45,7 +45,7 @@
 
 
 def terminal_interface(interpreter, message):
-    # Auto run and offline (this.. this isnt right) don't display messages.
+    # Auto run and offline (this.. this isn't right) don't display messages.
     # Probably worth abstracting this to something like "debug_cli" at some point.
     if not interpreter.auto_run and not interpreter.offline:
         interpreter_intro_message = [

diff --git a/interpreter/terminal_interface/utils/display_output.py b/interpreter/terminal_interface/utils/display_output.py
@@ -30,7 +30,7 @@ def display_output(output):
 
     # Return a message for the LLM.
     # We should make this specific to what happened in the future,
-    # like saying WHAT temporary file we made, ect. Keep the LLM informed.
+    # like saying WHAT temporary file we made, etc. Keep the LLM informed.
     return "Displayed on the user's machine."