github/AutoGPT
1
1
Fork 0
mirror of https://github.com/Significant-Gravitas/AutoGPT.git synced 2026-02-05 04:15:08 -05:00
Code Issues Packages Projects Releases Wiki Activity

Convert to python module named autogpt.

Browse Source 
Also fixed the Dockerfile.
Converting to module makes development easier.
Fixes coverage script in CI and test imports.
This commit is contained in:
Dino Hensen
2023-04-14 18:28:58 +02:00
committed by Merwane Hamadi
parent a17a850b25
commit d64f866bfa
45 changed files with 352 additions and 90 deletions
Download Patch File Download Diff File Expand all files Collapse all files

0
scripts/__init__.py
View File

434
scripts/main.py → scripts/agent.py
View File

@@ -1,331 +1,15 @@
import json
import random
import commands as cmd
import utils
from memory import get_memory, get_supported_memory_backends
import chat
from colorama import Fore, Style
from spinner import Spinner
import time
import speak
from config import Config
from json_parser import fix_and_parse_json
from ai_config import AIConfig
import json
import traceback
import yaml
import argparse
from tkinter.ttk import Style
from colorama import Fore
import chat
from config import Config
from logger import logger
import logging
from prompt import get_prompt
cfg = Config()
def check_openai_api_key():
"""Check if the OpenAI API key is set in config.py or as an environment variable."""
if not cfg.openai_api_key:
print(
Fore.RED +
"Please set your OpenAI API key in .env or as an environment variable."
)
print("You can get your key from https://beta.openai.com/account/api-keys")
exit(1)
def attempt_to_fix_json_by_finding_outermost_brackets(json_string):
if cfg.speak_mode and cfg.debug_mode:
speak.say_text("I have received an invalid JSON response from the OpenAI API. Trying to fix it now.")
logger.typewriter_log("Attempting to fix JSON by finding outermost brackets\n")
try:
# Use regex to search for JSON objects
import regex
json_pattern = regex.compile(r"\{(?:[^{}]|(?R))*\}")
json_match = json_pattern.search(json_string)
if json_match:
# Extract the valid JSON object from the string
json_string = json_match.group(0)
logger.typewriter_log(title="Apparently json was fixed.", title_color=Fore.GREEN)
if cfg.speak_mode and cfg.debug_mode:
speak.say_text("Apparently json was fixed.")
else:
raise ValueError("No valid JSON object found")
except (json.JSONDecodeError, ValueError) as e:
if cfg.speak_mode:
speak.say_text("Didn't work. I will have to ignore this response then.")
logger.error("Error: Invalid JSON, setting it to empty JSON now.\n")
json_string = {}
return json_string
def print_assistant_thoughts(assistant_reply):
"""Prints the assistant's thoughts to the console"""
global ai_name
global cfg
try:
try:
# Parse and print Assistant response
assistant_reply_json = fix_and_parse_json(assistant_reply)
except json.JSONDecodeError as e:
logger.error("Error: Invalid JSON in assistant thoughts\n", assistant_reply)
assistant_reply_json = attempt_to_fix_json_by_finding_outermost_brackets(assistant_reply)
assistant_reply_json = fix_and_parse_json(assistant_reply_json)
# Check if assistant_reply_json is a string and attempt to parse it into a JSON object
if isinstance(assistant_reply_json, str):
try:
assistant_reply_json = json.loads(assistant_reply_json)
except json.JSONDecodeError as e:
logger.error("Error: Invalid JSON\n", assistant_reply)
assistant_reply_json = attempt_to_fix_json_by_finding_outermost_brackets(assistant_reply_json)
assistant_thoughts_reasoning = None
assistant_thoughts_plan = None
assistant_thoughts_speak = None
assistant_thoughts_criticism = None
assistant_thoughts = assistant_reply_json.get("thoughts", {})
assistant_thoughts_text = assistant_thoughts.get("text")
if assistant_thoughts:
assistant_thoughts_reasoning = assistant_thoughts.get("reasoning")
assistant_thoughts_plan = assistant_thoughts.get("plan")
assistant_thoughts_criticism = assistant_thoughts.get("criticism")
assistant_thoughts_speak = assistant_thoughts.get("speak")
logger.typewriter_log(f"{ai_name.upper()} THOUGHTS:", Fore.YELLOW, assistant_thoughts_text)
logger.typewriter_log("REASONING:", Fore.YELLOW, assistant_thoughts_reasoning)
if assistant_thoughts_plan:
logger.typewriter_log("PLAN:", Fore.YELLOW, "")
# If it's a list, join it into a string
if isinstance(assistant_thoughts_plan, list):
assistant_thoughts_plan = "\n".join(assistant_thoughts_plan)
elif isinstance(assistant_thoughts_plan, dict):
assistant_thoughts_plan = str(assistant_thoughts_plan)
# Split the input_string using the newline character and dashes
lines = assistant_thoughts_plan.split('\n')
for line in lines:
line = line.lstrip("- ")
logger.typewriter_log("- ", Fore.GREEN, line.strip())
logger.typewriter_log("CRITICISM:", Fore.YELLOW, assistant_thoughts_criticism)
# Speak the assistant's thoughts
if cfg.speak_mode and assistant_thoughts_speak:
speak.say_text(assistant_thoughts_speak)
return assistant_reply_json
except json.decoder.JSONDecodeError as e:
logger.error("Error: Invalid JSON\n", assistant_reply)
if cfg.speak_mode:
speak.say_text("I have received an invalid JSON response from the OpenAI API. I cannot ignore this response.")
# All other errors, return "Error: + error message"
except Exception as e:
call_stack = traceback.format_exc()
logger.error("Error: \n", call_stack)
def construct_prompt():
"""Construct the prompt for the AI to respond to"""
config = AIConfig.load(cfg.ai_settings_file)
if cfg.skip_reprompt and config.ai_name:
logger.typewriter_log("Name :", Fore.GREEN, config.ai_name)
logger.typewriter_log("Role :", Fore.GREEN, config.ai_role)
logger.typewriter_log("Goals:", Fore.GREEN, config.ai_goals)
elif config.ai_name:
logger.typewriter_log(
f"Welcome back! ",
Fore.GREEN,
f"Would you like me to return to being {config.ai_name}?",
speak_text=True)
should_continue = utils.clean_input(f"""Continue with the last settings?
Name: {config.ai_name}
Role: {config.ai_role}
Goals: {config.ai_goals}
Continue (y/n): """)
if should_continue.lower() == "n":
config = AIConfig()
if not config.ai_name:
config = prompt_user()
config.save()
# Get rid of this global:
global ai_name
ai_name = config.ai_name
full_prompt = config.construct_full_prompt()
return full_prompt
def prompt_user():
"""Prompt the user for input"""
ai_name = ""
# Construct the prompt
logger.typewriter_log(
"Welcome to Auto-GPT! ",
Fore.GREEN,
"Enter the name of your AI and its role below. Entering nothing will load defaults.",
speak_text=True)
# Get AI Name from User
logger.typewriter_log(
"Name your AI: ",
Fore.GREEN,
"For example, 'Entrepreneur-GPT'")
ai_name = utils.clean_input("AI Name: ")
if ai_name == "":
ai_name = "Entrepreneur-GPT"
logger.typewriter_log(
f"{ai_name} here!",
Fore.LIGHTBLUE_EX,
"I am at your service.",
speak_text=True)
# Get AI Role from User
logger.typewriter_log(
"Describe your AI's role: ",
Fore.GREEN,
"For example, 'an AI designed to autonomously develop and run businesses with the sole goal of increasing your net worth.'")
ai_role = utils.clean_input(f"{ai_name} is: ")
if ai_role == "":
ai_role = "an AI designed to autonomously develop and run businesses with the sole goal of increasing your net worth."
# Enter up to 5 goals for the AI
logger.typewriter_log(
"Enter up to 5 goals for your AI: ",
Fore.GREEN,
"For example: \nIncrease net worth, Grow Twitter Account, Develop and manage multiple businesses autonomously'")
print("Enter nothing to load defaults, enter nothing when finished.", flush=True)
ai_goals = []
for i in range(5):
ai_goal = utils.clean_input(f"{Fore.LIGHTBLUE_EX}Goal{Style.RESET_ALL} {i+1}: ")
if ai_goal == "":
break
ai_goals.append(ai_goal)
if len(ai_goals) == 0:
ai_goals = ["Increase net worth", "Grow Twitter Account",
"Develop and manage multiple businesses autonomously"]
config = AIConfig(ai_name, ai_role, ai_goals)
return config
def parse_arguments():
"""Parses the arguments passed to the script"""
global cfg
cfg.set_debug_mode(False)
cfg.set_continuous_mode(False)
cfg.set_speak_mode(False)
parser = argparse.ArgumentParser(description='Process arguments.')
parser.add_argument('--continuous', '-c', action='store_true', help='Enable Continuous Mode')
parser.add_argument('--continuous-limit', '-l', type=int, dest="continuous_limit", help='Defines the number of times to run in continuous mode')
parser.add_argument('--speak', action='store_true', help='Enable Speak Mode')
parser.add_argument('--debug', action='store_true', help='Enable Debug Mode')
parser.add_argument('--gpt3only', action='store_true', help='Enable GPT3.5 Only Mode')
parser.add_argument('--gpt4only', action='store_true', help='Enable GPT4 Only Mode')
parser.add_argument('--use-memory', '-m', dest="memory_type", help='Defines which Memory backend to use')
parser.add_argument('--skip-reprompt', '-y', dest='skip_reprompt', action='store_true', help='Skips the re-prompting messages at the beginning of the script')
parser.add_argument('--ai-settings', '-C', dest='ai_settings_file', help="Specifies which ai_settings.yaml file to use, will also automatically skip the re-prompt.")
args = parser.parse_args()
if args.debug:
logger.typewriter_log("Debug Mode: ", Fore.GREEN, "ENABLED")
cfg.set_debug_mode(True)
if args.continuous:
logger.typewriter_log("Continuous Mode: ", Fore.RED, "ENABLED")
logger.typewriter_log(
"WARNING: ",
Fore.RED,
"Continuous mode is not recommended. It is potentially dangerous and may cause your AI to run forever or carry out actions you would not usually authorise. Use at your own risk.")
cfg.set_continuous_mode(True)
if args.continuous_limit:
logger.typewriter_log(
"Continuous Limit: ",
Fore.GREEN,
f"{args.continuous_limit}")
cfg.set_continuous_limit(args.continuous_limit)
# Check if continuous limit is used without continuous mode
if args.continuous_limit and not args.continuous:
parser.error("--continuous-limit can only be used with --continuous")
if args.speak:
logger.typewriter_log("Speak Mode: ", Fore.GREEN, "ENABLED")
cfg.set_speak_mode(True)
if args.gpt3only:
logger.typewriter_log("GPT3.5 Only Mode: ", Fore.GREEN, "ENABLED")
cfg.set_smart_llm_model(cfg.fast_llm_model)
if args.gpt4only:
logger.typewriter_log("GPT4 Only Mode: ", Fore.GREEN, "ENABLED")
cfg.set_fast_llm_model(cfg.smart_llm_model)
if args.memory_type:
supported_memory = get_supported_memory_backends()
chosen = args.memory_type
if not chosen in supported_memory:
logger.typewriter_log("ONLY THE FOLLOWING MEMORY BACKENDS ARE SUPPORTED: ", Fore.RED, f'{supported_memory}')
logger.typewriter_log(f"Defaulting to: ", Fore.YELLOW, cfg.memory_backend)
else:
cfg.memory_backend = chosen
if args.skip_reprompt:
logger.typewriter_log("Skip Re-prompt: ", Fore.GREEN, "ENABLED")
cfg.skip_reprompt = True
if args.ai_settings_file:
file = args.ai_settings_file
# Validate file
(validated, message) = utils.validate_yaml_file(file)
if not validated:
logger.typewriter_log("FAILED FILE VALIDATION", Fore.RED, message)
logger.double_check()
exit(1)
logger.typewriter_log("Using AI Settings File:", Fore.GREEN, file)
cfg.ai_settings_file = file
cfg.skip_reprompt = True
def main():
global ai_name, memory
# TODO: fill in llm values here
check_openai_api_key()
parse_arguments()
logger.set_level(logging.DEBUG if cfg.debug_mode else logging.INFO)
ai_name = ""
prompt = construct_prompt()
# print(prompt)
# Initialize variables
full_message_history = []
result = None
next_action_count = 0
# Make a constant:
user_input = "Determine which next command to use, and respond using the format specified above:"
# Initialize memory and make sure it is empty.
# this is particularly important for indexing and referencing pinecone memory
memory = get_memory(cfg, init=True)
print('Using memory of type: ' + memory.__class__.__name__)
agent = Agent(
ai_name=ai_name,
memory=memory,
full_message_history=full_message_history,
next_action_count=next_action_count,
prompt=prompt,
user_input=user_input
)
agent.start_interaction_loop()
import speak
from spinner import Spinner
class Agent:
@@ -356,6 +40,7 @@ class Agent:
def start_interaction_loop(self):
# Interaction Loop
cfg = Config()
loop_count = 0
while True:
# Discontinue if continuous limit is reached
@@ -461,5 +146,100 @@ class Agent:
logger.typewriter_log("SYSTEM: ", Fore.YELLOW, "Unable to execute command")
if __name__ == "__main__":
main()
def attempt_to_fix_json_by_finding_outermost_brackets(json_string):
cfg = Config()
if cfg.speak_mode and cfg.debug_mode:
speak.say_text("I have received an invalid JSON response from the OpenAI API. Trying to fix it now.")
logger.typewriter_log("Attempting to fix JSON by finding outermost brackets\n")
try:
# Use regex to search for JSON objects
import regex
json_pattern = regex.compile(r"\{(?:[^{}]|(?R))*\}")
json_match = json_pattern.search(json_string)
if json_match:
# Extract the valid JSON object from the string
json_string = json_match.group(0)
logger.typewriter_log(title="Apparently json was fixed.", title_color=Fore.GREEN)
if cfg.speak_mode and cfg.debug_mode:
speak.say_text("Apparently json was fixed.")
else:
raise ValueError("No valid JSON object found")
except (json.JSONDecodeError, ValueError) as e:
if cfg.speak_mode:
speak.say_text("Didn't work. I will have to ignore this response then.")
logger.error("Error: Invalid JSON, setting it to empty JSON now.\n")
json_string = {}
return json_string
def print_assistant_thoughts(assistant_reply):
"""Prints the assistant's thoughts to the console"""
global ai_name
global cfg
cfg = Config()
try:
try:
# Parse and print Assistant response
assistant_reply_json = fix_and_parse_json(assistant_reply)
except json.JSONDecodeError as e:
logger.error("Error: Invalid JSON in assistant thoughts\n", assistant_reply)
assistant_reply_json = attempt_to_fix_json_by_finding_outermost_brackets(assistant_reply)
assistant_reply_json = fix_and_parse_json(assistant_reply_json)
# Check if assistant_reply_json is a string and attempt to parse it into a JSON object
if isinstance(assistant_reply_json, str):
try:
assistant_reply_json = json.loads(assistant_reply_json)
except json.JSONDecodeError as e:
logger.error("Error: Invalid JSON\n", assistant_reply)
assistant_reply_json = attempt_to_fix_json_by_finding_outermost_brackets(assistant_reply_json)
assistant_thoughts_reasoning = None
assistant_thoughts_plan = None
assistant_thoughts_speak = None
assistant_thoughts_criticism = None
assistant_thoughts = assistant_reply_json.get("thoughts", {})
assistant_thoughts_text = assistant_thoughts.get("text")
if assistant_thoughts:
assistant_thoughts_reasoning = assistant_thoughts.get("reasoning")
assistant_thoughts_plan = assistant_thoughts.get("plan")
assistant_thoughts_criticism = assistant_thoughts.get("criticism")
assistant_thoughts_speak = assistant_thoughts.get("speak")
logger.typewriter_log(f"{ai_name.upper()} THOUGHTS:", Fore.YELLOW, assistant_thoughts_text)
logger.typewriter_log("REASONING:", Fore.YELLOW, assistant_thoughts_reasoning)
if assistant_thoughts_plan:
logger.typewriter_log("PLAN:", Fore.YELLOW, "")
# If it's a list, join it into a string
if isinstance(assistant_thoughts_plan, list):
assistant_thoughts_plan = "\n".join(assistant_thoughts_plan)
elif isinstance(assistant_thoughts_plan, dict):
assistant_thoughts_plan = str(assistant_thoughts_plan)
# Split the input_string using the newline character and dashes
lines = assistant_thoughts_plan.split('\n')
for line in lines:
line = line.lstrip("- ")
logger.typewriter_log("- ", Fore.GREEN, line.strip())
logger.typewriter_log("CRITICISM:", Fore.YELLOW, assistant_thoughts_criticism)
# Speak the assistant's thoughts
if cfg.speak_mode and assistant_thoughts_speak:
speak.say_text(assistant_thoughts_speak)
return assistant_reply_json
except json.decoder.JSONDecodeError as e:
logger.error("Error: Invalid JSON\n", assistant_reply)
if cfg.speak_mode:
speak.say_text("I have received an invalid JSON response from the OpenAI API. I cannot ignore this response.")
# All other errors, return "Error: + error message"
except Exception as e:
call_stack = traceback.format_exc()
logger.error("Error: \n", call_stack)

73
scripts/agent_manager.py
View File

@@ -1,73 +0,0 @@
from llm_utils import create_chat_completion
next_key = 0
agents = {} # key, (task, full_message_history, model)
# Create new GPT agent
# TODO: Centralise use of create_chat_completion() to globally enforce token limit
def create_agent(task, prompt, model):
"""Create a new agent and return its key"""
global next_key
global agents
messages = [{"role": "user", "content": prompt}, ]
# Start GPT instance
agent_reply = create_chat_completion(
model=model,
messages=messages,
)
# Update full message history
messages.append({"role": "assistant", "content": agent_reply})
key = next_key
# This is done instead of len(agents) to make keys unique even if agents
# are deleted
next_key += 1
agents[key] = (task, messages, model)
return key, agent_reply
def message_agent(key, message):
"""Send a message to an agent and return its response"""
global agents
task, messages, model = agents[int(key)]
# Add user message to message history before sending to agent
messages.append({"role": "user", "content": message})
# Start GPT instance
agent_reply = create_chat_completion(
model=model,
messages=messages,
)
# Update full message history
messages.append({"role": "assistant", "content": agent_reply})
return agent_reply
def list_agents():
"""Return a list of all agents"""
global agents
# Return a list of agent keys and their tasks
return [(key, task) for key, (task, _, _) in agents.items()]
def delete_agent(key):
"""Delete an agent and return True if successful, False otherwise"""
global agents
try:
del agents[int(key)]
return True
except KeyError:
return False

95
scripts/ai_config.py
View File

@@ -1,95 +0,0 @@
import yaml
import os
from prompt import get_prompt
class AIConfig:
"""
A class object that contains the configuration information for the AI
Attributes:
ai_name (str): The name of the AI.
ai_role (str): The description of the AI's role.
ai_goals (list): The list of objectives the AI is supposed to complete.
"""
def __init__(self, ai_name: str="", ai_role: str="", ai_goals: list=[]) -> None:
"""
Initialize a class instance
Parameters:
ai_name (str): The name of the AI.
ai_role (str): The description of the AI's role.
ai_goals (list): The list of objectives the AI is supposed to complete.
Returns:
None
"""
self.ai_name = ai_name
self.ai_role = ai_role
self.ai_goals = ai_goals
# Soon this will go in a folder where it remembers more stuff about the run(s)
SAVE_FILE = os.path.join(os.path.dirname(__file__), '..', 'ai_settings.yaml')
@classmethod
def load(cls: object, config_file: str=SAVE_FILE) -> object:
"""
Returns class object with parameters (ai_name, ai_role, ai_goals) loaded from yaml file if yaml file exists,
else returns class with no parameters.
Parameters:
cls (class object): An AIConfig Class object.
config_file (int): The path to the config yaml file. DEFAULT: "../ai_settings.yaml"
Returns:
cls (object): An instance of given cls object
"""
try:
with open(config_file, encoding='utf-8') as file:
config_params = yaml.load(file, Loader=yaml.FullLoader)
except FileNotFoundError:
config_params = {}
ai_name = config_params.get("ai_name", "")
ai_role = config_params.get("ai_role", "")
ai_goals = config_params.get("ai_goals", [])
return cls(ai_name, ai_role, ai_goals)
def save(self, config_file: str=SAVE_FILE) -> None:
"""
Saves the class parameters to the specified file yaml file path as a yaml file.
Parameters:
config_file(str): The path to the config yaml file. DEFAULT: "../ai_settings.yaml"
Returns:
None
"""
config = {"ai_name": self.ai_name, "ai_role": self.ai_role, "ai_goals": self.ai_goals}
with open(config_file, "w", encoding='utf-8') as file:
yaml.dump(config, file, allow_unicode=True)
def construct_full_prompt(self) -> str:
"""
Returns a prompt to the user with the class information in an organized fashion.
Parameters:
None
Returns:
full_prompt (str): A string containing the initial prompt for the user including the ai_name, ai_role and ai_goals.
"""
prompt_start = """Your decisions must always be made independently without seeking user assistance. Play to your strengths as an LLM and pursue simple strategies with no legal complications."""
# Construct full prompt
full_prompt = f"You are {self.ai_name}, {self.ai_role}\n{prompt_start}\n\nGOALS:\n\n"
for i, goal in enumerate(self.ai_goals):
full_prompt += f"{i+1}. {goal}\n"
full_prompt += f"\n\n{get_prompt()}"
return full_prompt

66
scripts/ai_functions.py
View File

@@ -1,66 +0,0 @@
from typing import List
import json
from config import Config
from call_ai_function import call_ai_function
cfg = Config()
def evaluate_code(code: str) -> List[str]:
"""
A function that takes in a string and returns a response from create chat completion api call.
Parameters:
code (str): Code to be evaluated.
Returns:
A result string from create chat completion. A list of suggestions to improve the code.
"""
function_string = "def analyze_code(code: str) -> List[str]:"
args = [code]
description_string = """Analyzes the given code and returns a list of suggestions for improvements."""
result_string = call_ai_function(function_string, args, description_string)
return result_string
def improve_code(suggestions: List[str], code: str) -> str:
"""
A function that takes in code and suggestions and returns a response from create chat completion api call.
Parameters:
suggestions (List): A list of suggestions around what needs to be improved.
code (str): Code to be improved.
Returns:
A result string from create chat completion. Improved code in response.
"""
function_string = (
"def generate_improved_code(suggestions: List[str], code: str) -> str:"
)
args = [json.dumps(suggestions), code]
description_string = """Improves the provided code based on the suggestions provided, making no other changes."""
result_string = call_ai_function(function_string, args, description_string)
return result_string
def write_tests(code: str, focus: List[str]) -> str:
"""
A function that takes in code and focus topics and returns a response from create chat completion api call.
Parameters:
focus (List): A list of suggestions around what needs to be improved.
code (str): Code for test cases to be generated against.
Returns:
A result string from create chat completion. Test cases for the submitted code in response.
"""
function_string = (
"def create_test_cases(code: str, focus: Optional[str] = None) -> str:"
)
args = [code, json.dumps(focus)]
description_string = """Generates test cases for the existing code, focusing on specific areas if required."""
result_string = call_ai_function(function_string, args, description_string)
return result_string

187
scripts/browse.py
View File

@@ -1,187 +0,0 @@
import requests
from bs4 import BeautifulSoup
from memory import get_memory
from config import Config
from llm_utils import create_chat_completion
from urllib.parse import urlparse, urljoin
cfg = Config()
memory = get_memory(cfg)
session = requests.Session()
session.headers.update({'User-Agent': cfg.user_agent})
# Function to check if the URL is valid
def is_valid_url(url):
try:
result = urlparse(url)
return all([result.scheme, result.netloc])
except ValueError:
return False
# Function to sanitize the URL
def sanitize_url(url):
return urljoin(url, urlparse(url).path)
# Define and check for local file address prefixes
def check_local_file_access(url):
local_prefixes = ['file:///', 'file://localhost', 'http://localhost', 'https://localhost']
return any(url.startswith(prefix) for prefix in local_prefixes)
def get_response(url, timeout=10):
try:
# Restrict access to local files
if check_local_file_access(url):
raise ValueError('Access to local files is restricted')
# Most basic check if the URL is valid:
if not url.startswith('http://') and not url.startswith('https://'):
raise ValueError('Invalid URL format')
sanitized_url = sanitize_url(url)
response = session.get(sanitized_url, timeout=timeout)
# Check if the response contains an HTTP error
if response.status_code >= 400:
return None, "Error: HTTP " + str(response.status_code) + " error"
return response, None
except ValueError as ve:
# Handle invalid URL format
return None, "Error: " + str(ve)
except requests.exceptions.RequestException as re:
# Handle exceptions related to the HTTP request (e.g., connection errors, timeouts, etc.)
return None, "Error: " + str(re)
def scrape_text(url):
"""Scrape text from a webpage"""
response, error_message = get_response(url)
if error_message:
return error_message
soup = BeautifulSoup(response.text, "html.parser")
for script in soup(["script", "style"]):
script.extract()
text = soup.get_text()
lines = (line.strip() for line in text.splitlines())
chunks = (phrase.strip() for line in lines for phrase in line.split(" "))
text = '\n'.join(chunk for chunk in chunks if chunk)
return text
def extract_hyperlinks(soup):
"""Extract hyperlinks from a BeautifulSoup object"""
hyperlinks = []
for link in soup.find_all('a', href=True):
hyperlinks.append((link.text, link['href']))
return hyperlinks
def format_hyperlinks(hyperlinks):
"""Format hyperlinks into a list of strings"""
formatted_links = []
for link_text, link_url in hyperlinks:
formatted_links.append(f"{link_text} ({link_url})")
return formatted_links
def scrape_links(url):
"""Scrape links from a webpage"""
response, error_message = get_response(url)
if error_message:
return error_message
soup = BeautifulSoup(response.text, "html.parser")
for script in soup(["script", "style"]):
script.extract()
hyperlinks = extract_hyperlinks(soup)
return format_hyperlinks(hyperlinks)
def split_text(text, max_length=cfg.browse_chunk_max_length):
"""Split text into chunks of a maximum length"""
paragraphs = text.split("\n")
current_length = 0
current_chunk = []
for paragraph in paragraphs:
if current_length + len(paragraph) + 1 <= max_length:
current_chunk.append(paragraph)
current_length += len(paragraph) + 1
else:
yield "\n".join(current_chunk)
current_chunk = [paragraph]
current_length = len(paragraph) + 1
if current_chunk:
yield "\n".join(current_chunk)
def create_message(chunk, question):
"""Create a message for the user to summarize a chunk of text"""
return {
"role": "user",
"content": f"\"\"\"{chunk}\"\"\" Using the above text, please answer the following question: \"{question}\" -- if the question cannot be answered using the text, please summarize the text."
}
def summarize_text(url, text, question):
"""Summarize text using the LLM model"""
if not text:
return "Error: No text to summarize"
text_length = len(text)
print(f"Text length: {text_length} characters")
summaries = []
chunks = list(split_text(text))
for i, chunk in enumerate(chunks):
print(f"Adding chunk {i + 1} / {len(chunks)} to memory")
memory_to_add = f"Source: {url}\n" \
f"Raw content part#{i + 1}: {chunk}"
memory.add(memory_to_add)
print(f"Summarizing chunk {i + 1} / {len(chunks)}")
messages = [create_message(chunk, question)]
summary = create_chat_completion(
model=cfg.fast_llm_model,
messages=messages,
max_tokens=cfg.browse_summary_max_token,
)
summaries.append(summary)
print(f"Added chunk {i + 1} summary to memory")
memory_to_add = f"Source: {url}\n" \
f"Content summary part#{i + 1}: {summary}"
memory.add(memory_to_add)
print(f"Summarized {len(chunks)} chunks.")
combined_summary = "\n".join(summaries)
messages = [create_message(combined_summary, question)]
final_summary = create_chat_completion(
model=cfg.fast_llm_model,
messages=messages,
max_tokens=cfg.browse_summary_max_token,
)
return final_summary

30
scripts/call_ai_function.py
View File

@@ -1,30 +0,0 @@
from config import Config
cfg = Config()
from llm_utils import create_chat_completion
# This is a magic function that can do anything with no-code. See
# https://github.com/Torantulino/AI-Functions for more info.
def call_ai_function(function, args, description, model=None):
"""Call an AI function"""
if model is None:
model = cfg.smart_llm_model
# For each arg, if any are None, convert to "None":
args = [str(arg) if arg is not None else "None" for arg in args]
# parse args to comma separated string
args = ", ".join(args)
messages = [
{
"role": "system",
"content": f"You are now the following python function: ```# {description}\n{function}```\n\nOnly respond with your `return` value.",
},
{"role": "user", "content": args},
]
response = create_chat_completion(
model=model, messages=messages, temperature=0
)
return response

144
scripts/chat.py
View File

@@ -1,144 +0,0 @@
import time
import openai
from dotenv import load_dotenv
from config import Config
import token_counter
from llm_utils import create_chat_completion
from logger import logger
import logging
cfg = Config()
def create_chat_message(role, content):
"""
Create a chat message with the given role and content.
Args:
role (str): The role of the message sender, e.g., "system", "user", or "assistant".
content (str): The content of the message.
Returns:
dict: A dictionary containing the role and content of the message.
"""
return {"role": role, "content": content}
def generate_context(prompt, relevant_memory, full_message_history, model):
current_context = [
create_chat_message(
"system", prompt),
create_chat_message(
"system", f"The current time and date is {time.strftime('%c')}"),
create_chat_message(
"system", f"This reminds you of these events from your past:\n{relevant_memory}\n\n")]
# Add messages from the full message history until we reach the token limit
next_message_to_add_index = len(full_message_history) - 1
insertion_index = len(current_context)
# Count the currently used tokens
current_tokens_used = token_counter.count_message_tokens(current_context, model)
return next_message_to_add_index, current_tokens_used, insertion_index, current_context
# TODO: Change debug from hardcode to argument
def chat_with_ai(
prompt,
user_input,
full_message_history,
permanent_memory,
token_limit):
"""Interact with the OpenAI API, sending the prompt, user input, message history, and permanent memory."""
while True:
try:
"""
Interact with the OpenAI API, sending the prompt, user input, message history, and permanent memory.
Args:
prompt (str): The prompt explaining the rules to the AI.
user_input (str): The input from the user.
full_message_history (list): The list of all messages sent between the user and the AI.
permanent_memory (Obj): The memory object containing the permanent memory.
token_limit (int): The maximum number of tokens allowed in the API call.
Returns:
str: The AI's response.
"""
model = cfg.fast_llm_model # TODO: Change model from hardcode to argument
# Reserve 1000 tokens for the response
logger.debug(f"Token limit: {token_limit}")
send_token_limit = token_limit - 1000
relevant_memory = '' if len(full_message_history) ==0 else permanent_memory.get_relevant(str(full_message_history[-9:]), 10)
logger.debug(f'Memory Stats: {permanent_memory.get_stats()}')
next_message_to_add_index, current_tokens_used, insertion_index, current_context = generate_context(
prompt, relevant_memory, full_message_history, model)
while current_tokens_used > 2500:
# remove memories until we are under 2500 tokens
relevant_memory = relevant_memory[1:]
next_message_to_add_index, current_tokens_used, insertion_index, current_context = generate_context(
prompt, relevant_memory, full_message_history, model)
current_tokens_used += token_counter.count_message_tokens([create_chat_message("user", user_input)], model) # Account for user input (appended later)
while next_message_to_add_index >= 0:
# print (f"CURRENT TOKENS USED: {current_tokens_used}")
message_to_add = full_message_history[next_message_to_add_index]
tokens_to_add = token_counter.count_message_tokens([message_to_add], model)
if current_tokens_used + tokens_to_add > send_token_limit:
break
# Add the most recent message to the start of the current context, after the two system prompts.
current_context.insert(insertion_index, full_message_history[next_message_to_add_index])
# Count the currently used tokens
current_tokens_used += tokens_to_add
# Move to the next most recent message in the full message history
next_message_to_add_index -= 1
# Append user input, the length of this is accounted for above
current_context.extend([create_chat_message("user", user_input)])
# Calculate remaining tokens
tokens_remaining = token_limit - current_tokens_used
# assert tokens_remaining >= 0, "Tokens remaining is negative. This should never happen, please submit a bug report at https://www.github.com/Torantulino/Auto-GPT"
# Debug print the current context
logger.debug(f"Token limit: {token_limit}")
logger.debug(f"Send Token Count: {current_tokens_used}")
logger.debug(f"Tokens remaining for response: {tokens_remaining}")
logger.debug("------------ CONTEXT SENT TO AI ---------------")
for message in current_context:
# Skip printing the prompt
if message["role"] == "system" and message["content"] == prompt:
continue
logger.debug(f"{message['role'].capitalize()}: {message['content']}")
logger.debug("")
logger.debug("----------- END OF CONTEXT ----------------")
# TODO: use a model defined elsewhere, so that model can contain temperature and other settings we care about
assistant_reply = create_chat_completion(
model=model,
messages=current_context,
max_tokens=tokens_remaining,
)
# Update full message history
full_message_history.append(
create_chat_message(
"user", user_input))
full_message_history.append(
create_chat_message(
"assistant", assistant_reply))
return assistant_reply
except openai.error.RateLimitError:
# TODO: When we switch to langchain, this is built in
print("Error: ", "API Rate Limit Reached. Waiting 10 seconds...")
time.sleep(10)

309
scripts/commands.py
View File

@@ -1,309 +0,0 @@
import browse
import json
from memory import get_memory
import datetime
import agent_manager as agents
import speak
from config import Config
import ai_functions as ai
from file_operations import read_file, write_to_file, append_to_file, delete_file, search_files
from execute_code import execute_python_file, execute_shell
from json_parser import fix_and_parse_json
from image_gen import generate_image
from duckduckgo_search import ddg
from googleapiclient.discovery import build
from googleapiclient.errors import HttpError
cfg = Config()
def is_valid_int(value):
try:
int(value)
return True
except ValueError:
return False
def get_command(response):
"""Parse the response and return the command name and arguments"""
try:
response_json = fix_and_parse_json(response)
if "command" not in response_json:
return "Error:" , "Missing 'command' object in JSON"
command = response_json["command"]
if "name" not in command:
return "Error:", "Missing 'name' field in 'command' object"
command_name = command["name"]
# Use an empty dictionary if 'args' field is not present in 'command' object
arguments = command.get("args", {})
return command_name, arguments
except json.decoder.JSONDecodeError:
return "Error:", "Invalid JSON"
# All other errors, return "Error: + error message"
except Exception as e:
return "Error:", str(e)
def execute_command(command_name, arguments):
"""Execute the command and return the result"""
memory = get_memory(cfg)
try:
if command_name == "google":
# Check if the Google API key is set and use the official search method
# If the API key is not set or has only whitespaces, use the unofficial search method
if cfg.google_api_key and (cfg.google_api_key.strip() if cfg.google_api_key else None):
return google_official_search(arguments["input"])
else:
return google_search(arguments["input"])
elif command_name == "memory_add":
return memory.add(arguments["string"])
elif command_name == "start_agent":
return start_agent(
arguments["name"],
arguments["task"],
arguments["prompt"])
elif command_name == "message_agent":
return message_agent(arguments["key"], arguments["message"])
elif command_name == "list_agents":
return list_agents()
elif command_name == "delete_agent":
return delete_agent(arguments["key"])
elif command_name == "get_text_summary":
return get_text_summary(arguments["url"], arguments["question"])
elif command_name == "get_hyperlinks":
return get_hyperlinks(arguments["url"])
elif command_name == "read_file":
return read_file(arguments["file"])
elif command_name == "write_to_file":
return write_to_file(arguments["file"], arguments["text"])
elif command_name == "append_to_file":
return append_to_file(arguments["file"], arguments["text"])
elif command_name == "delete_file":
return delete_file(arguments["file"])
elif command_name == "search_files":
return search_files(arguments["directory"])
elif command_name == "browse_website":
return browse_website(arguments["url"], arguments["question"])
# TODO: Change these to take in a file rather than pasted code, if
# non-file is given, return instructions "Input should be a python
# filepath, write your code to file and try again"
elif command_name == "evaluate_code":
return ai.evaluate_code(arguments["code"])
elif command_name == "improve_code":
return ai.improve_code(arguments["suggestions"], arguments["code"])
elif command_name == "write_tests":
return ai.write_tests(arguments["code"], arguments.get("focus"))
elif command_name == "execute_python_file": # Add this command
return execute_python_file(arguments["file"])
elif command_name == "execute_shell":
if cfg.execute_local_commands:
return execute_shell(arguments["command_line"])
else:
return "You are not allowed to run local shell commands. To execute shell commands, EXECUTE_LOCAL_COMMANDS must be set to 'True' in your config. Do not attempt to bypass the restriction."
elif command_name == "generate_image":
return generate_image(arguments["prompt"])
elif command_name == "do_nothing":
return "No action performed."
elif command_name == "task_complete":
shutdown()
else:
return f"Unknown command '{command_name}'. Please refer to the 'COMMANDS' list for available commands and only respond in the specified JSON format."
# All errors, return "Error: + error message"
except Exception as e:
return "Error: " + str(e)
def get_datetime():
"""Return the current date and time"""
return "Current date and time: " + \
datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")
def google_search(query, num_results=8):
"""Return the results of a google search"""
search_results = []
for j in ddg(query, max_results=num_results):
search_results.append(j)
return json.dumps(search_results, ensure_ascii=False, indent=4)
def google_official_search(query, num_results=8):
"""Return the results of a google search using the official Google API"""
from googleapiclient.discovery import build
from googleapiclient.errors import HttpError
import json
try:
# Get the Google API key and Custom Search Engine ID from the config file
api_key = cfg.google_api_key
custom_search_engine_id = cfg.custom_search_engine_id
# Initialize the Custom Search API service
service = build("customsearch", "v1", developerKey=api_key)
# Send the search query and retrieve the results
result = service.cse().list(q=query, cx=custom_search_engine_id, num=num_results).execute()
# Extract the search result items from the response
search_results = result.get("items", [])
# Create a list of only the URLs from the search results
search_results_links = [item["link"] for item in search_results]
except HttpError as e:
# Handle errors in the API call
error_details = json.loads(e.content.decode())
# Check if the error is related to an invalid or missing API key
if error_details.get("error", {}).get("code") == 403 and "invalid API key" in error_details.get("error", {}).get("message", ""):
return "Error: The provided Google API key is invalid or missing."
else:
return f"Error: {e}"
# Return the list of search result URLs
return search_results_links
def browse_website(url, question):
"""Browse a website and return the summary and links"""
summary = get_text_summary(url, question)
links = get_hyperlinks(url)
# Limit links to 5
if len(links) > 5:
links = links[:5]
result = f"""Website Content Summary: {summary}\n\nLinks: {links}"""
return result
def get_text_summary(url, question):
"""Return the results of a google search"""
text = browse.scrape_text(url)
summary = browse.summarize_text(url, text, question)
return """ "Result" : """ + summary
def get_hyperlinks(url):
"""Return the results of a google search"""
link_list = browse.scrape_links(url)
return link_list
def commit_memory(string):
"""Commit a string to memory"""
_text = f"""Committing memory with string "{string}" """
mem.permanent_memory.append(string)
return _text
def delete_memory(key):
"""Delete a memory with a given key"""
if key >= 0 and key < len(mem.permanent_memory):
_text = "Deleting memory with key " + str(key)
del mem.permanent_memory[key]
print(_text)
return _text
else:
print("Invalid key, cannot delete memory.")
return None
def overwrite_memory(key, string):
"""Overwrite a memory with a given key and string"""
# Check if the key is a valid integer
if is_valid_int(key):
key_int = int(key)
# Check if the integer key is within the range of the permanent_memory list
if 0 <= key_int < len(mem.permanent_memory):
_text = "Overwriting memory with key " + str(key) + " and string " + string
# Overwrite the memory slot with the given integer key and string
mem.permanent_memory[key_int] = string
print(_text)
return _text
else:
print(f"Invalid key '{key}', out of range.")
return None
# Check if the key is a valid string
elif isinstance(key, str):
_text = "Overwriting memory with key " + key + " and string " + string
# Overwrite the memory slot with the given string key and string
mem.permanent_memory[key] = string
print(_text)
return _text
else:
print(f"Invalid key '{key}', must be an integer or a string.")
return None
def shutdown():
"""Shut down the program"""
print("Shutting down...")
quit()
def start_agent(name, task, prompt, model=cfg.fast_llm_model):
"""Start an agent with a given name, task, and prompt"""
global cfg
# Remove underscores from name
voice_name = name.replace("_", " ")
first_message = f"""You are {name}. Respond with: "Acknowledged"."""
agent_intro = f"{voice_name} here, Reporting for duty!"
# Create agent
if cfg.speak_mode:
speak.say_text(agent_intro, 1)
key, ack = agents.create_agent(task, first_message, model)
if cfg.speak_mode:
speak.say_text(f"Hello {voice_name}. Your task is as follows. {task}.")
# Assign task (prompt), get response
agent_response = message_agent(key, prompt)
return f"Agent {name} created with key {key}. First response: {agent_response}"
def message_agent(key, message):
"""Message an agent with a given key and message"""
global cfg
# Check if the key is a valid integer
if is_valid_int(key):
agent_response = agents.message_agent(int(key), message)
# Check if the key is a valid string
elif isinstance(key, str):
agent_response = agents.message_agent(key, message)
else:
return "Invalid key, must be an integer or a string."
# Speak response
if cfg.speak_mode:
speak.say_text(agent_response, 1)
return agent_response
def list_agents():
"""List all agents"""
return agents.list_agents()
def delete_agent(key):
"""Delete an agent with a given key"""
result = agents.delete_agent(key)
if not result:
return f"Agent {key} does not exist."
return f"Agent {key} deleted."

206
scripts/config.py
View File

@@ -1,206 +0,0 @@
import abc
import os
import openai
import yaml
from dotenv import load_dotenv
# Load environment variables from .env file
load_dotenv()
class Singleton(abc.ABCMeta, type):
"""
Singleton metaclass for ensuring only one instance of a class.
"""
_instances = {}
def __call__(cls, *args, **kwargs):
"""Call method for the singleton metaclass."""
if cls not in cls._instances:
cls._instances[cls] = super(
Singleton, cls).__call__(
*args, **kwargs)
return cls._instances[cls]
class AbstractSingleton(abc.ABC, metaclass=Singleton):
pass
class Config(metaclass=Singleton):
"""
Configuration class to store the state of bools for different scripts access.
"""
def __init__(self):
"""Initialize the Config class"""
self.debug_mode = False
self.continuous_mode = False
self.continuous_limit = 0
self.speak_mode = False
self.skip_reprompt = False
self.ai_settings_file = os.getenv("AI_SETTINGS_FILE", "ai_settings.yaml")
self.fast_llm_model = os.getenv("FAST_LLM_MODEL", "gpt-3.5-turbo")
self.smart_llm_model = os.getenv("SMART_LLM_MODEL", "gpt-4")
self.fast_token_limit = int(os.getenv("FAST_TOKEN_LIMIT", 4000))
self.smart_token_limit = int(os.getenv("SMART_TOKEN_LIMIT", 8000))
self.browse_chunk_max_length = int(os.getenv("BROWSE_CHUNK_MAX_LENGTH", 8192))
self.browse_summary_max_token = int(os.getenv("BROWSE_SUMMARY_MAX_TOKEN", 300))
self.openai_api_key = os.getenv("OPENAI_API_KEY")
self.temperature = float(os.getenv("TEMPERATURE", "1"))
self.use_azure = os.getenv("USE_AZURE") == 'True'
self.execute_local_commands = os.getenv('EXECUTE_LOCAL_COMMANDS', 'False') == 'True'
if self.use_azure:
self.load_azure_config()
openai.api_type = self.openai_api_type
openai.api_base = self.openai_api_base
openai.api_version = self.openai_api_version
self.elevenlabs_api_key = os.getenv("ELEVENLABS_API_KEY")
self.elevenlabs_voice_1_id = os.getenv("ELEVENLABS_VOICE_1_ID")
self.elevenlabs_voice_2_id = os.getenv("ELEVENLABS_VOICE_2_ID")
self.use_mac_os_tts = False
self.use_mac_os_tts = os.getenv("USE_MAC_OS_TTS")
self.use_brian_tts = False
self.use_brian_tts = os.getenv("USE_BRIAN_TTS")
self.google_api_key = os.getenv("GOOGLE_API_KEY")
self.custom_search_engine_id = os.getenv("CUSTOM_SEARCH_ENGINE_ID")
self.pinecone_api_key = os.getenv("PINECONE_API_KEY")
self.pinecone_region = os.getenv("PINECONE_ENV")
self.image_provider = os.getenv("IMAGE_PROVIDER")
self.huggingface_api_token = os.getenv("HUGGINGFACE_API_TOKEN")
# User agent headers to use when browsing web
# Some websites might just completely deny request with an error code if no user agent was found.
self.user_agent = os.getenv("USER_AGENT", "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36")
self.redis_host = os.getenv("REDIS_HOST", "localhost")
self.redis_port = os.getenv("REDIS_PORT", "6379")
self.redis_password = os.getenv("REDIS_PASSWORD", "")
self.wipe_redis_on_start = os.getenv("WIPE_REDIS_ON_START", "True") == 'True'
self.memory_index = os.getenv("MEMORY_INDEX", 'auto-gpt')
# Note that indexes must be created on db 0 in redis, this is not configurable.
self.memory_backend = os.getenv("MEMORY_BACKEND", 'local')
# Initialize the OpenAI API client
openai.api_key = self.openai_api_key
def get_azure_deployment_id_for_model(self, model: str) -> str:
"""
Returns the relevant deployment id for the model specified.
Parameters:
model(str): The model to map to the deployment id.
Returns:
The matching deployment id if found, otherwise an empty string.
"""
if model == self.fast_llm_model:
return self.azure_model_to_deployment_id_map["fast_llm_model_deployment_id"]
elif model == self.smart_llm_model:
return self.azure_model_to_deployment_id_map["smart_llm_model_deployment_id"]
elif model == "text-embedding-ada-002":
return self.azure_model_to_deployment_id_map["embedding_model_deployment_id"]
else:
return ""
AZURE_CONFIG_FILE = os.path.join(os.path.dirname(__file__), '..', 'azure.yaml')
def load_azure_config(self, config_file: str=AZURE_CONFIG_FILE) -> None:
"""
Loads the configuration parameters for Azure hosting from the specified file path as a yaml file.
Parameters:
config_file(str): The path to the config yaml file. DEFAULT: "../azure.yaml"
Returns:
None
"""
try:
with open(config_file) as file:
config_params = yaml.load(file, Loader=yaml.FullLoader)
except FileNotFoundError:
config_params = {}
self.openai_api_type = os.getenv("OPENAI_API_TYPE", config_params.get("azure_api_type", "azure"))
self.openai_api_base = os.getenv("OPENAI_AZURE_API_BASE", config_params.get("azure_api_base", ""))
self.openai_api_version = os.getenv("OPENAI_AZURE_API_VERSION", config_params.get("azure_api_version", ""))
self.azure_model_to_deployment_id_map = config_params.get("azure_model_map", [])
def set_continuous_mode(self, value: bool):
"""Set the continuous mode value."""
self.continuous_mode = value
def set_continuous_limit(self, value: int):
"""Set the continuous limit value."""
self.continuous_limit = value
def set_speak_mode(self, value: bool):
"""Set the speak mode value."""
self.speak_mode = value
def set_fast_llm_model(self, value: str):
"""Set the fast LLM model value."""
self.fast_llm_model = value
def set_smart_llm_model(self, value: str):
"""Set the smart LLM model value."""
self.smart_llm_model = value
def set_fast_token_limit(self, value: int):
"""Set the fast token limit value."""
self.fast_token_limit = value
def set_smart_token_limit(self, value: int):
"""Set the smart token limit value."""
self.smart_token_limit = value
def set_browse_chunk_max_length(self, value: int):
"""Set the browse_website command chunk max length value."""
self.browse_chunk_max_length = value
def set_browse_summary_max_token(self, value: int):
"""Set the browse_website command summary max token value."""
self.browse_summary_max_token = value
def set_openai_api_key(self, value: str):
"""Set the OpenAI API key value."""
self.openai_api_key = value
def set_elevenlabs_api_key(self, value: str):
"""Set the ElevenLabs API key value."""
self.elevenlabs_api_key = value
def set_elevenlabs_voice_1_id(self, value: str):
"""Set the ElevenLabs Voice 1 ID value."""
self.elevenlabs_voice_1_id = value
def set_elevenlabs_voice_2_id(self, value: str):
"""Set the ElevenLabs Voice 2 ID value."""
self.elevenlabs_voice_2_id = value
def set_google_api_key(self, value: str):
"""Set the Google API key value."""
self.google_api_key = value
def set_custom_search_engine_id(self, value: str):
"""Set the custom search engine id value."""
self.custom_search_engine_id = value
def set_pinecone_api_key(self, value: str):
"""Set the Pinecone API key value."""
self.pinecone_api_key = value
def set_pinecone_region(self, value: str):
"""Set the Pinecone region value."""
self.pinecone_region = value
def set_debug_mode(self, value: bool):
"""Set the debug mode value."""
self.debug_mode = value

70
scripts/data_ingestion.py
View File

@@ -1,70 +0,0 @@
import argparse
import logging
from config import Config
from memory import get_memory
from file_operations import ingest_file, search_files
cfg = Config()
def configure_logging():
logging.basicConfig(filename='log-ingestion.txt',
filemode='a',
format='%(asctime)s,%(msecs)d %(name)s %(levelname)s %(message)s',
datefmt='%H:%M:%S',
level=logging.DEBUG)
return logging.getLogger('AutoGPT-Ingestion')
def ingest_directory(directory, memory, args):
"""
Ingest all files in a directory by calling the ingest_file function for each file.
:param directory: The directory containing the files to ingest
:param memory: An object with an add() method to store the chunks in memory
"""
try:
files = search_files(directory)
for file in files:
ingest_file(file, memory, args.max_length, args.overlap)
except Exception as e:
print(f"Error while ingesting directory '{directory}': {str(e)}")
def main():
logger = configure_logging()
parser = argparse.ArgumentParser(description="Ingest a file or a directory with multiple files into memory. Make sure to set your .env before running this script.")
group = parser.add_mutually_exclusive_group(required=True)
group.add_argument("--file", type=str, help="The file to ingest.")
group.add_argument("--dir", type=str, help="The directory containing the files to ingest.")
parser.add_argument("--init", action='store_true', help="Init the memory and wipe its content (default: False)", default=False)
parser.add_argument("--overlap", type=int, help="The overlap size between chunks when ingesting files (default: 200)", default=200)
parser.add_argument("--max_length", type=int, help="The max_length of each chunk when ingesting files (default: 4000)", default=4000)
args = parser.parse_args()
# Initialize memory
memory = get_memory(cfg, init=args.init)
print('Using memory of type: ' + memory.__class__.__name__)
if args.file:
try:
ingest_file(args.file, memory, args.max_length, args.overlap)
print(f"File '{args.file}' ingested successfully.")
except Exception as e:
logger.error(f"Error while ingesting file '{args.file}': {str(e)}")
print(f"Error while ingesting file '{args.file}': {str(e)}")
elif args.dir:
try:
ingest_directory(args.dir, memory, args)
print(f"Directory '{args.dir}' ingested successfully.")
except Exception as e:
logger.error(f"Error while ingesting directory '{args.dir}': {str(e)}")
print(f"Error while ingesting directory '{args.dir}': {str(e)}")
else:
print("Please provide either a file path (--file) or a directory name (--dir) inside the auto_gpt_workspace directory as input.")
if __name__ == "__main__":
main()

99
scripts/execute_code.py
View File

@@ -1,99 +0,0 @@
import docker
import os
import subprocess
WORKSPACE_FOLDER = "auto_gpt_workspace"
def execute_python_file(file):
"""Execute a Python file in a Docker container and return the output"""
print (f"Executing file '{file}' in workspace '{WORKSPACE_FOLDER}'")
if not file.endswith(".py"):
return "Error: Invalid file type. Only .py files are allowed."
file_path = os.path.join(WORKSPACE_FOLDER, file)
if not os.path.isfile(file_path):
return f"Error: File '{file}' does not exist."
if we_are_running_in_a_docker_container():
result = subprocess.run(f'python {file_path}', capture_output=True, encoding="utf8", shell=True)
if result.returncode == 0:
return result.stdout
else:
return f"Error: {result.stderr}"
else:
try:
client = docker.from_env()
image_name = 'python:3.10'
try:
client.images.get(image_name)
print(f"Image '{image_name}' found locally")
except docker.errors.ImageNotFound:
print(f"Image '{image_name}' not found locally, pulling from Docker Hub")
# Use the low-level API to stream the pull response
low_level_client = docker.APIClient()
for line in low_level_client.pull(image_name, stream=True, decode=True):
# Print the status and progress, if available
status = line.get('status')
progress = line.get('progress')
if status and progress:
print(f"{status}: {progress}")
elif status:
print(status)
# You can replace 'python:3.8' with the desired Python image/version
# You can find available Python images on Docker Hub:
# https://hub.docker.com/_/python
container = client.containers.run(
image_name,
f'python {file}',
volumes={
os.path.abspath(WORKSPACE_FOLDER): {
'bind': '/workspace',
'mode': 'ro'}},
working_dir='/workspace',
stderr=True,
stdout=True,
detach=True,
)
output = container.wait()
logs = container.logs().decode('utf-8')
container.remove()
# print(f"Execution complete. Output: {output}")
# print(f"Logs: {logs}")
return logs
except Exception as e:
return f"Error: {str(e)}"
def execute_shell(command_line):
current_dir = os.getcwd()
if not WORKSPACE_FOLDER in current_dir: # Change dir into workspace if necessary
work_dir = os.path.join(os.getcwd(), WORKSPACE_FOLDER)
os.chdir(work_dir)
print (f"Executing command '{command_line}' in working directory '{os.getcwd()}'")
result = subprocess.run(command_line, capture_output=True, shell=True)
output = f"STDOUT:\n{result.stdout}\nSTDERR:\n{result.stderr}"
# Change back to whatever the prior working dir was
os.chdir(current_dir)
return output
def we_are_running_in_a_docker_container():
os.path.exists('/.dockerenv')

138
scripts/file_operations.py
View File

@@ -1,138 +0,0 @@
import os
import os.path
# Set a dedicated folder for file I/O
working_directory = "auto_gpt_workspace"
# Create the directory if it doesn't exist
if not os.path.exists(working_directory):
os.makedirs(working_directory)
def safe_join(base, *paths):
"""Join one or more path components intelligently."""
new_path = os.path.join(base, *paths)
norm_new_path = os.path.normpath(new_path)
if os.path.commonprefix([base, norm_new_path]) != base:
raise ValueError("Attempted to access outside of working directory.")
return norm_new_path
def split_file(content, max_length=4000, overlap=0):
"""
Split text into chunks of a specified maximum length with a specified overlap
between chunks.
:param text: The input text to be split into chunks
:param max_length: The maximum length of each chunk, default is 4000 (about 1k token)
:param overlap: The number of overlapping characters between chunks, default is no overlap
:return: A generator yielding chunks of text
"""
start = 0
content_length = len(content)
while start < content_length:
end = start + max_length
if end + overlap < content_length:
chunk = content[start:end+overlap]
else:
chunk = content[start:content_length]
yield chunk
start += max_length - overlap
def read_file(filename):
"""Read a file and return the contents"""
try:
filepath = safe_join(working_directory, filename)
with open(filepath, "r", encoding='utf-8') as f:
content = f.read()
return content
except Exception as e:
return "Error: " + str(e)
def ingest_file(filename, memory, max_length=4000, overlap=200):
"""
Ingest a file by reading its content, splitting it into chunks with a specified
maximum length and overlap, and adding the chunks to the memory storage.
:param filename: The name of the file to ingest
:param memory: An object with an add() method to store the chunks in memory
:param max_length: The maximum length of each chunk, default is 4000
:param overlap: The number of overlapping characters between chunks, default is 200
"""
try:
print(f"Working with file {filename}")
content = read_file(filename)
content_length = len(content)
print(f"File length: {content_length} characters")
chunks = list(split_file(content, max_length=max_length, overlap=overlap))
num_chunks = len(chunks)
for i, chunk in enumerate(chunks):
print(f"Ingesting chunk {i + 1} / {num_chunks} into memory")
memory_to_add = f"Filename: {filename}\n" \
f"Content part#{i + 1}/{num_chunks}: {chunk}"
memory.add(memory_to_add)
print(f"Done ingesting {num_chunks} chunks from {filename}.")
except Exception as e:
print(f"Error while ingesting file '{filename}': {str(e)}")
def write_to_file(filename, text):
"""Write text to a file"""
try:
filepath = safe_join(working_directory, filename)
directory = os.path.dirname(filepath)
if not os.path.exists(directory):
os.makedirs(directory)
with open(filepath, "w", encoding='utf-8') as f:
f.write(text)
return "File written to successfully."
except Exception as e:
return "Error: " + str(e)
def append_to_file(filename, text):
"""Append text to a file"""
try:
filepath = safe_join(working_directory, filename)
with open(filepath, "a") as f:
f.write(text)
return "Text appended successfully."
except Exception as e:
return "Error: " + str(e)
def delete_file(filename):
"""Delete a file"""
try:
filepath = safe_join(working_directory, filename)
os.remove(filepath)
return "File deleted successfully."
except Exception as e:
return "Error: " + str(e)
def search_files(directory):
found_files = []
if directory == "" or directory == "/":
search_directory = working_directory
else:
search_directory = safe_join(working_directory, directory)
for root, _, files in os.walk(search_directory):
for file in files:
if file.startswith('.'):
continue
relative_path = os.path.relpath(os.path.join(root, file), working_directory)
found_files.append(relative_path)
return found_files

58
scripts/image_gen.py
View File

@@ -1,58 +0,0 @@
import requests
import io
import os.path
from PIL import Image
from config import Config
import uuid
import openai
from base64 import b64decode
cfg = Config()
working_directory = "auto_gpt_workspace"
def generate_image(prompt):
filename = str(uuid.uuid4()) + ".jpg"
# DALL-E
if cfg.image_provider == 'dalle':
openai.api_key = cfg.openai_api_key
response = openai.Image.create(
prompt=prompt,
n=1,
size="256x256",
response_format="b64_json",
)
print("Image Generated for prompt:" + prompt)
image_data = b64decode(response["data"][0]["b64_json"])
with open(working_directory + "/" + filename, mode="wb") as png:
png.write(image_data)
return "Saved to disk:" + filename
# STABLE DIFFUSION
elif cfg.image_provider == 'sd':
API_URL = "https://api-inference.huggingface.co/models/CompVis/stable-diffusion-v1-4"
headers = {"Authorization": "Bearer " + cfg.huggingface_api_token}
response = requests.post(API_URL, headers=headers, json={
"inputs": prompt,
})
image = Image.open(io.BytesIO(response.content))
print("Image Generated for prompt:" + prompt)
image.save(os.path.join(working_directory, filename))
return "Saved to disk:" + filename
else:
return "No Image Provider Set"

109
scripts/json_parser.py
View File

@@ -1,109 +0,0 @@
import json
from typing import Any, Dict, Union
from call_ai_function import call_ai_function
from config import Config
from json_utils import correct_json
from logger import logger
cfg = Config()
JSON_SCHEMA = """
{
"command": {
"name": "command name",
"args":{
"arg name": "value"
}
},
"thoughts":
{
"text": "thought",
"reasoning": "reasoning",
"plan": "- short bulleted\n- list that conveys\n- long-term plan",
"criticism": "constructive self-criticism",
"speak": "thoughts summary to say to user"
}
}
"""
def fix_and_parse_json(
json_str: str,
try_to_fix_with_gpt: bool = True
) -> Union[str, Dict[Any, Any]]:
"""Fix and parse JSON string"""
try:
json_str = json_str.replace('\t', '')
return json.loads(json_str)
except json.JSONDecodeError as _: # noqa: F841
try:
json_str = correct_json(json_str)
return json.loads(json_str)
except json.JSONDecodeError as _: # noqa: F841
pass
# Let's do something manually:
# sometimes GPT responds with something BEFORE the braces:
# "I'm sorry, I don't understand. Please try again."
# {"text": "I'm sorry, I don't understand. Please try again.",
# "confidence": 0.0}
# So let's try to find the first brace and then parse the rest
# of the string
try:
brace_index = json_str.index("{")
json_str = json_str[brace_index:]
last_brace_index = json_str.rindex("}")
json_str = json_str[:last_brace_index+1]
return json.loads(json_str)
# Can throw a ValueError if there is no "{" or "}" in the json_str
except (json.JSONDecodeError, ValueError) as e: # noqa: F841
if try_to_fix_with_gpt:
logger.warn("Warning: Failed to parse AI output, attempting to fix."
"\n If you see this warning frequently, it's likely that"
" your prompt is confusing the AI. Try changing it up"
" slightly.")
# Now try to fix this up using the ai_functions
ai_fixed_json = fix_json(json_str, JSON_SCHEMA)
if ai_fixed_json != "failed":
return json.loads(ai_fixed_json)
else:
# This allows the AI to react to the error message,
# which usually results in it correcting its ways.
logger.error("Failed to fix AI output, telling the AI.")
return json_str
else:
raise e
def fix_json(json_str: str, schema: str) -> str:
"""Fix the given JSON string to make it parseable and fully compliant with the provided schema."""
# Try to fix the JSON using GPT:
function_string = "def fix_json(json_str: str, schema:str=None) -> str:"
args = [f"'''{json_str}'''", f"'''{schema}'''"]
description_string = "Fixes the provided JSON string to make it parseable"\
" and fully compliant with the provided schema.\n If an object or"\
" field specified in the schema isn't contained within the correct"\
" JSON, it is omitted.\n This function is brilliant at guessing"\
" when the format is incorrect."
# If it doesn't already start with a "`", add one:
if not json_str.startswith("`"):
json_str = "```json\n" + json_str + "\n```"
result_string = call_ai_function(
function_string, args, description_string, model=cfg.fast_llm_model
)
logger.debug("------------ JSON FIX ATTEMPT ---------------")
logger.debug(f"Original JSON: {json_str}")
logger.debug("-----------")
logger.debug(f"Fixed JSON: {result_string}")
logger.debug("----------- END OF FIX ATTEMPT ----------------")
try:
json.loads(result_string) # just check the validity
return result_string
except: # noqa: E722
# Get the call stack:
# import traceback
# call_stack = traceback.format_exc()
# print(f"Failed to fix JSON: '{json_str}' "+call_stack)
return "failed"

127
scripts/json_utils.py
View File

@@ -1,127 +0,0 @@
import re
import json
from config import Config
cfg = Config()
def extract_char_position(error_message: str) -> int:
"""Extract the character position from the JSONDecodeError message.
Args:
error_message (str): The error message from the JSONDecodeError
exception.
Returns:
int: The character position.
"""
import re
char_pattern = re.compile(r'\(char (\d+)\)')
if match := char_pattern.search(error_message):
return int(match[1])
else:
raise ValueError("Character position not found in the error message.")
def add_quotes_to_property_names(json_string: str) -> str:
"""
Add quotes to property names in a JSON string.
Args:
json_string (str): The JSON string.
Returns:
str: The JSON string with quotes added to property names.
"""
def replace_func(match):
return f'"{match.group(1)}":'
property_name_pattern = re.compile(r'(\w+):')
corrected_json_string = property_name_pattern.sub(
replace_func,
json_string)
try:
json.loads(corrected_json_string)
return corrected_json_string
except json.JSONDecodeError as e:
raise e
def balance_braces(json_string: str) -> str:
"""
Balance the braces in a JSON string.
Args:
json_string (str): The JSON string.
Returns:
str: The JSON string with braces balanced.
"""
open_braces_count = json_string.count('{')
close_braces_count = json_string.count('}')
while open_braces_count > close_braces_count:
json_string += '}'
close_braces_count += 1
while close_braces_count > open_braces_count:
json_string = json_string.rstrip('}')
close_braces_count -= 1
try:
json.loads(json_string)
return json_string
except json.JSONDecodeError as e:
pass
def fix_invalid_escape(json_str: str, error_message: str) -> str:
while error_message.startswith('Invalid \\escape'):
bad_escape_location = extract_char_position(error_message)
json_str = json_str[:bad_escape_location] + \
json_str[bad_escape_location + 1:]
try:
json.loads(json_str)
return json_str
except json.JSONDecodeError as e:
if cfg.debug_mode:
print('json loads error - fix invalid escape', e)
error_message = str(e)
return json_str
def correct_json(json_str: str) -> str:
"""
Correct common JSON errors.
Args:
json_str (str): The JSON string.
"""
try:
if cfg.debug_mode:
print("json", json_str)
json.loads(json_str)
return json_str
except json.JSONDecodeError as e:
if cfg.debug_mode:
print('json loads error', e)
error_message = str(e)
if error_message.startswith('Invalid \\escape'):
json_str = fix_invalid_escape(json_str, error_message)
if error_message.startswith('Expecting property name enclosed in double quotes'):
json_str = add_quotes_to_property_names(json_str)
try:
json.loads(json_str)
return json_str
except json.JSONDecodeError as e:
if cfg.debug_mode:
print('json loads error - add quotes', e)
error_message = str(e)
if balanced_str := balance_braces(json_str):
return balanced_str
return json_str

52
scripts/llm_utils.py
View File

@@ -1,52 +0,0 @@
import time
import openai
from colorama import Fore
from config import Config
cfg = Config()
openai.api_key = cfg.openai_api_key
# Overly simple abstraction until we create something better
# simple retry mechanism when getting a rate error or a bad gateway
def create_chat_completion(messages, model=None, temperature=cfg.temperature, max_tokens=None)->str:
"""Create a chat completion using the OpenAI API"""
response = None
num_retries = 5
for attempt in range(num_retries):
try:
if cfg.use_azure:
response = openai.ChatCompletion.create(
deployment_id=cfg.get_azure_deployment_id_for_model(model),
model=model,
messages=messages,
temperature=temperature,
max_tokens=max_tokens
)
else:
response = openai.ChatCompletion.create(
model=model,
messages=messages,
temperature=temperature,
max_tokens=max_tokens
)
break
except openai.error.RateLimitError:
if cfg.debug_mode:
print(Fore.RED + "Error: ", "API Rate Limit Reached. Waiting 20 seconds..." + Fore.RESET)
time.sleep(20)
except openai.error.APIError as e:
if e.http_status == 502:
if cfg.debug_mode:
print(Fore.RED + "Error: ", "API Bad gateway. Waiting 20 seconds..." + Fore.RESET)
time.sleep(20)
else:
raise
if attempt == num_retries - 1:
raise
if response is None:
raise RuntimeError("Failed to get response after 5 retries")
return response.choices[0].message["content"]

193
scripts/logger.py
View File

@@ -1,193 +0,0 @@
import logging
import os
import random
import re
import time
from logging import LogRecord
from colorama import Fore
from colorama import Style
import speak
from config import Config
from config import Singleton
cfg = Config()
'''
Logger that handle titles in different colors.
Outputs logs in console, activity.log, and errors.log
For console handler: simulates typing
'''
class Logger(metaclass=Singleton):
def __init__(self):
# create log directory if it doesn't exist
this_files_dir_path = os.path.dirname(__file__)
log_dir = os.path.join(this_files_dir_path, '../logs')
if not os.path.exists(log_dir):
os.makedirs(log_dir)
log_file = "activity.log"
error_file = "error.log"
console_formatter = AutoGptFormatter('%(title_color)s %(message)s')
# Create a handler for console which simulate typing
self.typing_console_handler = TypingConsoleHandler()
self.typing_console_handler.setLevel(logging.INFO)
self.typing_console_handler.setFormatter(console_formatter)
# Create a handler for console without typing simulation
self.console_handler = ConsoleHandler()
self.console_handler.setLevel(logging.DEBUG)
self.console_handler.setFormatter(console_formatter)
# Info handler in activity.log
self.file_handler = logging.FileHandler(os.path.join(log_dir, log_file))
self.file_handler.setLevel(logging.DEBUG)
info_formatter = AutoGptFormatter('%(asctime)s %(levelname)s %(title)s %(message_no_color)s')
self.file_handler.setFormatter(info_formatter)
# Error handler error.log
error_handler = logging.FileHandler(os.path.join(log_dir, error_file))
error_handler.setLevel(logging.ERROR)
error_formatter = AutoGptFormatter(
'%(asctime)s %(levelname)s %(module)s:%(funcName)s:%(lineno)d %(title)s %(message_no_color)s')
error_handler.setFormatter(error_formatter)
self.typing_logger = logging.getLogger('TYPER')
self.typing_logger.addHandler(self.typing_console_handler)
self.typing_logger.addHandler(self.file_handler)
self.typing_logger.addHandler(error_handler)
self.typing_logger.setLevel(logging.DEBUG)
self.logger = logging.getLogger('LOGGER')
self.logger.addHandler(self.console_handler)
self.logger.addHandler(self.file_handler)
self.logger.addHandler(error_handler)
self.logger.setLevel(logging.DEBUG)
def typewriter_log(
self,
title='',
title_color='',
content='',
speak_text=False,
level=logging.INFO):
if speak_text and cfg.speak_mode:
speak.say_text(f"{title}. {content}")
if content:
if isinstance(content, list):
content = " ".join(content)
else:
content = ""
self.typing_logger.log(level, content, extra={'title': title, 'color': title_color})
def debug(
self,
message,
title='',
title_color='',
):
self._log(title, title_color, message, logging.DEBUG)
def warn(
self,
message,
title='',
title_color='',
):
self._log(title, title_color, message, logging.WARN)
def error(
self,
title,
message=''
):
self._log(title, Fore.RED, message, logging.ERROR)
def _log(
self,
title='',
title_color='',
message='',
level=logging.INFO):
if message:
if isinstance(message, list):
message = " ".join(message)
self.logger.log(level, message, extra={'title': title, 'color': title_color})
def set_level(self, level):
self.logger.setLevel(level)
self.typing_logger.setLevel(level)
def double_check(self, additionalText=None):
if not additionalText:
additionalText = "Please ensure you've setup and configured everything correctly. Read https://github.com/Torantulino/Auto-GPT#readme to double check. You can also create a github issue or join the discord and ask there!"
self.typewriter_log("DOUBLE CHECK CONFIGURATION", Fore.YELLOW, additionalText)
'''
Output stream to console using simulated typing
'''
class TypingConsoleHandler(logging.StreamHandler):
def emit(self, record):
min_typing_speed = 0.05
max_typing_speed = 0.01
msg = self.format(record)
try:
words = msg.split()
for i, word in enumerate(words):
print(word, end="", flush=True)
if i < len(words) - 1:
print(" ", end="", flush=True)
typing_speed = random.uniform(min_typing_speed, max_typing_speed)
time.sleep(typing_speed)
# type faster after each word
min_typing_speed = min_typing_speed * 0.95
max_typing_speed = max_typing_speed * 0.95
print()
except Exception:
self.handleError(record)
class ConsoleHandler(logging.StreamHandler):
def emit(self, record):
msg = self.format(record)
try:
print(msg)
except Exception:
self.handleError(record)
class AutoGptFormatter(logging.Formatter):
"""
Allows to handle custom placeholders 'title_color' and 'message_no_color'.
To use this formatter, make sure to pass 'color', 'title' as log extras.
"""
def format(self, record: LogRecord) -> str:
if (hasattr(record, 'color')):
record.title_color = getattr(record, 'color') + getattr(record, 'title') + " " + Style.RESET_ALL
else:
record.title_color = getattr(record, 'title')
if hasattr(record, 'msg'):
record.message_no_color = remove_color_codes(getattr(record, 'msg'))
else:
record.message_no_color = ''
return super().format(record)
def remove_color_codes(s: str) -> str:
ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])')
return ansi_escape.sub('', s)
logger = Logger()

59
scripts/memory/__init__.py
View File

@@ -1,59 +0,0 @@
from memory.local import LocalCache
from memory.no_memory import NoMemory
# List of supported memory backends
# Add a backend to this list if the import attempt is successful
supported_memory = ['local', 'no_memory']
try:
from memory.redismem import RedisMemory
supported_memory.append('redis')
except ImportError:
print("Redis not installed. Skipping import.")
RedisMemory = None
try:
from memory.pinecone import PineconeMemory
supported_memory.append('pinecone')
except ImportError:
print("Pinecone not installed. Skipping import.")
PineconeMemory = None
def get_memory(cfg, init=False):
memory = None
if cfg.memory_backend == "pinecone":
if not PineconeMemory:
print("Error: Pinecone is not installed. Please install pinecone"
" to use Pinecone as a memory backend.")
else:
memory = PineconeMemory(cfg)
if init:
memory.clear()
elif cfg.memory_backend == "redis":
if not RedisMemory:
print("Error: Redis is not installed. Please install redis-py to"
" use Redis as a memory backend.")
else:
memory = RedisMemory(cfg)
elif cfg.memory_backend == "no_memory":
memory = NoMemory(cfg)
if memory is None:
memory = LocalCache(cfg)
if init:
memory.clear()
return memory
def get_supported_memory_backends():
return supported_memory
__all__ = [
"get_memory",
"LocalCache",
"RedisMemory",
"PineconeMemory",
"NoMemory"
]

36
scripts/memory/base.py
View File

@@ -1,36 +0,0 @@
"""Base class for memory providers."""
import abc
from config import AbstractSingleton, Config
import openai
cfg = Config()
def get_ada_embedding(text):
text = text.replace("\n", " ")
if cfg.use_azure:
return openai.Embedding.create(input=[text], engine=cfg.get_azure_deployment_id_for_model("text-embedding-ada-002"))["data"][0]["embedding"]
else:
return openai.Embedding.create(input=[text], model="text-embedding-ada-002")["data"][0]["embedding"]
class MemoryProviderSingleton(AbstractSingleton):
@abc.abstractmethod
def add(self, data):
pass
@abc.abstractmethod
def get(self, data):
pass
@abc.abstractmethod
def clear(self):
pass
@abc.abstractmethod
def get_relevant(self, data, num_relevant=5):
pass
@abc.abstractmethod
def get_stats(self):
pass

124
scripts/memory/local.py
View File

@@ -1,124 +0,0 @@
import dataclasses
import orjson
from typing import Any, List, Optional
import numpy as np
import os
from memory.base import MemoryProviderSingleton, get_ada_embedding
EMBED_DIM = 1536
SAVE_OPTIONS = orjson.OPT_SERIALIZE_NUMPY | orjson.OPT_SERIALIZE_DATACLASS
def create_default_embeddings():
return np.zeros((0, EMBED_DIM)).astype(np.float32)
@dataclasses.dataclass
class CacheContent:
texts: List[str] = dataclasses.field(default_factory=list)
embeddings: np.ndarray = dataclasses.field(
default_factory=create_default_embeddings
)
class LocalCache(MemoryProviderSingleton):
# on load, load our database
def __init__(self, cfg) -> None:
self.filename = f"{cfg.memory_index}.json"
if os.path.exists(self.filename):
try:
with open(self.filename, 'w+b') as f:
file_content = f.read()
if not file_content.strip():
file_content = b'{}'
f.write(file_content)
loaded = orjson.loads(file_content)
self.data = CacheContent(**loaded)
except orjson.JSONDecodeError:
print(f"Error: The file '{self.filename}' is not in JSON format.")
self.data = CacheContent()
else:
print(f"Warning: The file '{self.filename}' does not exist. Local memory would not be saved to a file.")
self.data = CacheContent()
def add(self, text: str):
"""
Add text to our list of texts, add embedding as row to our
embeddings-matrix
Args:
text: str
Returns: None
"""
if 'Command Error:' in text:
return ""
self.data.texts.append(text)
embedding = get_ada_embedding(text)
vector = np.array(embedding).astype(np.float32)
vector = vector[np.newaxis, :]
self.data.embeddings = np.concatenate(
[
self.data.embeddings,
vector,
],
axis=0,
)
with open(self.filename, 'wb') as f:
out = orjson.dumps(
self.data,
option=SAVE_OPTIONS
)
f.write(out)
return text
def clear(self) -> str:
"""
Clears the redis server.
Returns: A message indicating that the memory has been cleared.
"""
self.data = CacheContent()
return "Obliviated"
def get(self, data: str) -> Optional[List[Any]]:
"""
Gets the data from the memory that is most relevant to the given data.
Args:
data: The data to compare to.
Returns: The most relevant data.
"""
return self.get_relevant(data, 1)
def get_relevant(self, text: str, k: int) -> List[Any]:
""""
matrix-vector mult to find score-for-each-row-of-matrix
get indices for top-k winning scores
return texts for those indices
Args:
text: str
k: int
Returns: List[str]
"""
embedding = get_ada_embedding(text)
scores = np.dot(self.data.embeddings, embedding)
top_k_indices = np.argsort(scores)[-k:][::-1]
return [self.data.texts[i] for i in top_k_indices]
def get_stats(self):
"""
Returns: The stats of the local cache.
"""
return len(self.data.texts), self.data.embeddings.shape

66
scripts/memory/no_memory.py
View File

@@ -1,66 +0,0 @@
from typing import Optional, List, Any
from memory.base import MemoryProviderSingleton
class NoMemory(MemoryProviderSingleton):
def __init__(self, cfg):
"""
Initializes the NoMemory provider.
Args:
cfg: The config object.
Returns: None
"""
pass
def add(self, data: str) -> str:
"""
Adds a data point to the memory. No action is taken in NoMemory.
Args:
data: The data to add.
Returns: An empty string.
"""
return ""
def get(self, data: str) -> Optional[List[Any]]:
"""
Gets the data from the memory that is most relevant to the given data.
NoMemory always returns None.
Args:
data: The data to compare to.
Returns: None
"""
return None
def clear(self) -> str:
"""
Clears the memory. No action is taken in NoMemory.
Returns: An empty string.
"""
return ""
def get_relevant(self, data: str, num_relevant: int = 5) -> Optional[List[Any]]:
"""
Returns all the data in the memory that is relevant to the given data.
NoMemory always returns None.
Args:
data: The data to compare to.
num_relevant: The number of relevant data to return.
Returns: None
"""
return None
def get_stats(self):
"""
Returns: An empty dictionary as there are no stats in NoMemory.
"""
return {}

62
scripts/memory/pinecone.py
View File

@@ -1,62 +0,0 @@
import pinecone
from memory.base import MemoryProviderSingleton, get_ada_embedding
from logger import logger
from colorama import Fore, Style
class PineconeMemory(MemoryProviderSingleton):
def __init__(self, cfg):
pinecone_api_key = cfg.pinecone_api_key
pinecone_region = cfg.pinecone_region
pinecone.init(api_key=pinecone_api_key, environment=pinecone_region)
dimension = 1536
metric = "cosine"
pod_type = "p1"
table_name = "auto-gpt"
# this assumes we don't start with memory.
# for now this works.
# we'll need a more complicated and robust system if we want to start with memory.
self.vec_num = 0
try:
pinecone.whoami()
except Exception as e:
logger.typewriter_log("FAILED TO CONNECT TO PINECONE", Fore.RED, Style.BRIGHT + str(e) + Style.RESET_ALL)
logger.double_check("Please ensure you have setup and configured Pinecone properly for use. " +
f"You can check out {Fore.CYAN + Style.BRIGHT}https://github.com/Torantulino/Auto-GPT#-pinecone-api-key-setup{Style.RESET_ALL} to ensure you've set up everything correctly.")
exit(1)
if table_name not in pinecone.list_indexes():
pinecone.create_index(table_name, dimension=dimension, metric=metric, pod_type=pod_type)
self.index = pinecone.Index(table_name)
def add(self, data):
vector = get_ada_embedding(data)
# no metadata here. We may wish to change that long term.
resp = self.index.upsert([(str(self.vec_num), vector, {"raw_text": data})])
_text = f"Inserting data into memory at index: {self.vec_num}:\n data: {data}"
self.vec_num += 1
return _text
def get(self, data):
return self.get_relevant(data, 1)
def clear(self):
self.index.delete(deleteAll=True)
return "Obliviated"
def get_relevant(self, data, num_relevant=5):
"""
Returns all the data in the memory that is relevant to the given data.
:param data: The data to compare to.
:param num_relevant: The number of relevant data to return. Defaults to 5
"""
query_embedding = get_ada_embedding(data)
results = self.index.query(query_embedding, top_k=num_relevant, include_metadata=True)
sorted_results = sorted(results.matches, key=lambda x: x.score)
return [str(item['metadata']["raw_text"]) for item in sorted_results]
def get_stats(self):
return self.index.describe_index_stats()

155
scripts/memory/redismem.py
View File

@@ -1,155 +0,0 @@
"""Redis memory provider."""
from typing import Any, List, Optional
import redis
from redis.commands.search.field import VectorField, TextField
from redis.commands.search.query import Query
from redis.commands.search.indexDefinition import IndexDefinition, IndexType
import numpy as np
from memory.base import MemoryProviderSingleton, get_ada_embedding
from logger import logger
from colorama import Fore, Style
SCHEMA = [
TextField("data"),
VectorField(
"embedding",
"HNSW",
{
"TYPE": "FLOAT32",
"DIM": 1536,
"DISTANCE_METRIC": "COSINE"
}
),
]
class RedisMemory(MemoryProviderSingleton):
def __init__(self, cfg):
"""
Initializes the Redis memory provider.
Args:
cfg: The config object.
Returns: None
"""
redis_host = cfg.redis_host
redis_port = cfg.redis_port
redis_password = cfg.redis_password
self.dimension = 1536
self.redis = redis.Redis(
host=redis_host,
port=redis_port,
password=redis_password,
db=0 # Cannot be changed
)
self.cfg = cfg
# Check redis connection
try:
self.redis.ping()
except redis.ConnectionError as e:
logger.typewriter_log("FAILED TO CONNECT TO REDIS", Fore.RED, Style.BRIGHT + str(e) + Style.RESET_ALL)
logger.double_check("Please ensure you have setup and configured Redis properly for use. " +
f"You can check out {Fore.CYAN + Style.BRIGHT}https://github.com/Torantulino/Auto-GPT#redis-setup{Style.RESET_ALL} to ensure you've set up everything correctly.")
exit(1)
if cfg.wipe_redis_on_start:
self.redis.flushall()
try:
self.redis.ft(f"{cfg.memory_index}").create_index(
fields=SCHEMA,
definition=IndexDefinition(
prefix=[f"{cfg.memory_index}:"],
index_type=IndexType.HASH
)
)
except Exception as e:
print("Error creating Redis search index: ", e)
existing_vec_num = self.redis.get(f'{cfg.memory_index}-vec_num')
self.vec_num = int(existing_vec_num.decode('utf-8')) if\
existing_vec_num else 0
def add(self, data: str) -> str:
"""
Adds a data point to the memory.
Args:
data: The data to add.
Returns: Message indicating that the data has been added.
"""
if 'Command Error:' in data:
return ""
vector = get_ada_embedding(data)
vector = np.array(vector).astype(np.float32).tobytes()
data_dict = {
b"data": data,
"embedding": vector
}
pipe = self.redis.pipeline()
pipe.hset(f"{self.cfg.memory_index}:{self.vec_num}", mapping=data_dict)
_text = f"Inserting data into memory at index: {self.vec_num}:\n"\
f"data: {data}"
self.vec_num += 1
pipe.set(f'{self.cfg.memory_index}-vec_num', self.vec_num)
pipe.execute()
return _text
def get(self, data: str) -> Optional[List[Any]]:
"""
Gets the data from the memory that is most relevant to the given data.
Args:
data: The data to compare to.
Returns: The most relevant data.
"""
return self.get_relevant(data, 1)
def clear(self) -> str:
"""
Clears the redis server.
Returns: A message indicating that the memory has been cleared.
"""
self.redis.flushall()
return "Obliviated"
def get_relevant(
self,
data: str,
num_relevant: int = 5
) -> Optional[List[Any]]:
"""
Returns all the data in the memory that is relevant to the given data.
Args:
data: The data to compare to.
num_relevant: The number of relevant data to return.
Returns: A list of the most relevant data.
"""
query_embedding = get_ada_embedding(data)
base_query = f"*=>[KNN {num_relevant} @embedding $vector AS vector_score]"
query = Query(base_query).return_fields(
"data",
"vector_score"
).sort_by("vector_score").dialect(2)
query_vector = np.array(query_embedding).astype(np.float32).tobytes()
try:
results = self.redis.ft(f"{self.cfg.memory_index}").search(
query, query_params={"vector": query_vector}
)
except Exception as e:
print("Error calling Redis search: ", e)
return None
return [result.data for result in results.docs]
def get_stats(self):
"""
Returns: The stats of the memory index.
"""
return self.redis.ft(f"{self.cfg.memory_index}").info()

63
scripts/prompt.py
View File

@@ -1,63 +0,0 @@
from promptgenerator import PromptGenerator
def get_prompt():
"""
This function generates a prompt string that includes various constraints, commands, resources, and performance evaluations.
Returns:
str: The generated prompt string.
"""
# Initialize the PromptGenerator object
prompt_generator = PromptGenerator()
# Add constraints to the PromptGenerator object
prompt_generator.add_constraint("~4000 word limit for short term memory. Your short term memory is short, so immediately save important information to files.")
prompt_generator.add_constraint("If you are unsure how you previously did something or want to recall past events, thinking about similar events will help you remember.")
prompt_generator.add_constraint("No user assistance")
prompt_generator.add_constraint('Exclusively use the commands listed in double quotes e.g. "command name"')
# Define the command list
commands = [
("Google Search", "google", {"input": "<search>"}),
("Browse Website", "browse_website", {"url": "<url>", "question": "<what_you_want_to_find_on_website>"}),
("Start GPT Agent", "start_agent", {"name": "<name>", "task": "<short_task_desc>", "prompt": "<prompt>"}),
("Message GPT Agent", "message_agent", {"key": "<key>", "message": "<message>"}),
("List GPT Agents", "list_agents", {}),
("Delete GPT Agent", "delete_agent", {"key": "<key>"}),
("Write to file", "write_to_file", {"file": "<file>", "text": "<text>"}),
("Read file", "read_file", {"file": "<file>"}),
("Append to file", "append_to_file", {"file": "<file>", "text": "<text>"}),
("Delete file", "delete_file", {"file": "<file>"}),
("Search Files", "search_files", {"directory": "<directory>"}),
("Evaluate Code", "evaluate_code", {"code": "<full_code_string>"}),
("Get Improved Code", "improve_code", {"suggestions": "<list_of_suggestions>", "code": "<full_code_string>"}),
("Write Tests", "write_tests", {"code": "<full_code_string>", "focus": "<list_of_focus_areas>"}),
("Execute Python File", "execute_python_file", {"file": "<file>"}),
("Execute Shell Command, non-interactive commands only", "execute_shell", { "command_line": "<command_line>"}),
("Task Complete (Shutdown)", "task_complete", {"reason": "<reason>"}),
("Generate Image", "generate_image", {"prompt": "<prompt>"}),
("Do Nothing", "do_nothing", {}),
]
# Add commands to the PromptGenerator object
for command_label, command_name, args in commands:
prompt_generator.add_command(command_label, command_name, args)
# Add resources to the PromptGenerator object
prompt_generator.add_resource("Internet access for searches and information gathering.")
prompt_generator.add_resource("Long Term memory management.")
prompt_generator.add_resource("GPT-3.5 powered Agents for delegation of simple tasks.")
prompt_generator.add_resource("File output.")
# Add performance evaluations to the PromptGenerator object
prompt_generator.add_performance_evaluation("Continuously review and analyze your actions to ensure you are performing to the best of your abilities.")
prompt_generator.add_performance_evaluation("Constructively self-criticize your big-picture behavior constantly.")
prompt_generator.add_performance_evaluation("Reflect on past decisions and strategies to refine your approach.")
prompt_generator.add_performance_evaluation("Every command has a cost, so be smart and efficient. Aim to complete tasks in the least number of steps.")
# Generate the prompt string
prompt_string = prompt_generator.generate_prompt_string()
return prompt_string

129
scripts/promptgenerator.py
View File

@@ -1,129 +0,0 @@
import json
class PromptGenerator:
"""
A class for generating custom prompt strings based on constraints, commands, resources, and performance evaluations.
"""
def __init__(self):
"""
Initialize the PromptGenerator object with empty lists of constraints, commands, resources, and performance evaluations.
"""
self.constraints = []
self.commands = []
self.resources = []
self.performance_evaluation = []
self.response_format = {
"thoughts": {
"text": "thought",
"reasoning": "reasoning",
"plan": "- short bulleted\n- list that conveys\n- long-term plan",
"criticism": "constructive self-criticism",
"speak": "thoughts summary to say to user"
},
"command": {
"name": "command name",
"args": {
"arg name": "value"
}
}
}
def add_constraint(self, constraint):
"""
Add a constraint to the constraints list.
Args:
constraint (str): The constraint to be added.
"""
self.constraints.append(constraint)
def add_command(self, command_label, command_name, args=None):
"""
Add a command to the commands list with a label, name, and optional arguments.
Args:
command_label (str): The label of the command.
command_name (str): The name of the command.
args (dict, optional): A dictionary containing argument names and their values. Defaults to None.
"""
if args is None:
args = {}
command_args = {arg_key: arg_value for arg_key,
arg_value in args.items()}
command = {
"label": command_label,
"name": command_name,
"args": command_args,
}
self.commands.append(command)
def _generate_command_string(self, command):
"""
Generate a formatted string representation of a command.
Args:
command (dict): A dictionary containing command information.
Returns:
str: The formatted command string.
"""
args_string = ', '.join(
f'"{key}": "{value}"' for key, value in command['args'].items())
return f'{command["label"]}: "{command["name"]}", args: {args_string}'
def add_resource(self, resource):
"""
Add a resource to the resources list.
Args:
resource (str): The resource to be added.
"""
self.resources.append(resource)
def add_performance_evaluation(self, evaluation):
"""
Add a performance evaluation item to the performance_evaluation list.
Args:
evaluation (str): The evaluation item to be added.
"""
self.performance_evaluation.append(evaluation)
def _generate_numbered_list(self, items, item_type='list'):
"""
Generate a numbered list from given items based on the item_type.
Args:
items (list): A list of items to be numbered.
item_type (str, optional): The type of items in the list. Defaults to 'list'.
Returns:
str: The formatted numbered list.
"""
if item_type == 'command':
return "\n".join(f"{i+1}. {self._generate_command_string(item)}" for i, item in enumerate(items))
else:
return "\n".join(f"{i+1}. {item}" for i, item in enumerate(items))
def generate_prompt_string(self):
"""
Generate a prompt string based on the constraints, commands, resources, and performance evaluations.
Returns:
str: The generated prompt string.
"""
formatted_response_format = json.dumps(self.response_format, indent=4)
prompt_string = (
f"Constraints:\n{self._generate_numbered_list(self.constraints)}\n\n"
f"Commands:\n{self._generate_numbered_list(self.commands, item_type='command')}\n\n"
f"Resources:\n{self._generate_numbered_list(self.resources)}\n\n"
f"Performance Evaluation:\n{self._generate_numbered_list(self.performance_evaluation)}\n\n"
f"You should only respond in JSON format as described below \nResponse Format: \n{formatted_response_format} \nEnsure the response can be parsed by Python json.loads"
)
return prompt_string

113
scripts/speak.py
View File

@@ -1,113 +0,0 @@
import os
from playsound import playsound
import requests
from config import Config
cfg = Config()
import gtts
import threading
from threading import Lock, Semaphore
# Default voice IDs
default_voices = ["ErXwobaYiN019PkySvjV", "EXAVITQu4vr4xnSDxMaL"]
# Retrieve custom voice IDs from the Config class
custom_voice_1 = cfg.elevenlabs_voice_1_id
custom_voice_2 = cfg.elevenlabs_voice_2_id
# Placeholder values that should be treated as empty
placeholders = {"your-voice-id"}
# Use custom voice IDs if provided and not placeholders, otherwise use default voice IDs
voices = [
custom_voice_1 if custom_voice_1 and custom_voice_1 not in placeholders else default_voices[0],
custom_voice_2 if custom_voice_2 and custom_voice_2 not in placeholders else default_voices[1]
]
tts_headers = {
"Content-Type": "application/json",
"xi-api-key": cfg.elevenlabs_api_key
}
mutex_lock = Lock() # Ensure only one sound is played at a time
queue_semaphore = Semaphore(1) # The amount of sounds to queue before blocking the main thread
def eleven_labs_speech(text, voice_index=0):
"""Speak text using elevenlabs.io's API"""
tts_url = "https://api.elevenlabs.io/v1/text-to-speech/{voice_id}".format(
voice_id=voices[voice_index])
formatted_message = {"text": text}
response = requests.post(
tts_url, headers=tts_headers, json=formatted_message)
if response.status_code == 200:
with mutex_lock:
with open("speech.mpeg", "wb") as f:
f.write(response.content)
playsound("speech.mpeg", True)
os.remove("speech.mpeg")
return True
else:
print("Request failed with status code:", response.status_code)
print("Response content:", response.content)
return False
def brian_speech(text):
"""Speak text using Brian with the streamelements API"""
tts_url = f"https://api.streamelements.com/kappa/v2/speech?voice=Brian&text={text}"
response = requests.get(tts_url)
if response.status_code == 200:
with mutex_lock:
with open("speech.mp3", "wb") as f:
f.write(response.content)
playsound("speech.mp3")
os.remove("speech.mp3")
return True
else:
print("Request failed with status code:", response.status_code)
print("Response content:", response.content)
return False
def gtts_speech(text):
tts = gtts.gTTS(text)
with mutex_lock:
tts.save("speech.mp3")
playsound("speech.mp3", True)
os.remove("speech.mp3")
def macos_tts_speech(text, voice_index=0):
if voice_index == 0:
os.system(f'say "{text}"')
else:
if voice_index == 1:
os.system(f'say -v "Ava (Premium)" "{text}"')
else:
os.system(f'say -v Samantha "{text}"')
def say_text(text, voice_index=0):
def speak():
if not cfg.elevenlabs_api_key:
if cfg.use_mac_os_tts == 'True':
macos_tts_speech(text)
elif cfg.use_brian_tts == 'True':
success = brian_speech(text)
if not success:
gtts_speech(text)
else:
gtts_speech(text)
else:
success = eleven_labs_speech(text, voice_index)
if not success:
gtts_speech(text)
queue_semaphore.release()
queue_semaphore.acquire(True)
thread = threading.Thread(target=speak)
thread.start()

36
scripts/spinner.py
View File

@@ -1,36 +0,0 @@
import sys
import threading
import itertools
import time
class Spinner:
"""A simple spinner class"""
def __init__(self, message="Loading...", delay=0.1):
"""Initialize the spinner class"""
self.spinner = itertools.cycle(['-', '/', '|', '\\'])
self.delay = delay
self.message = message
self.running = False
self.spinner_thread = None
def spin(self):
"""Spin the spinner"""
while self.running:
sys.stdout.write(f"{next(self.spinner)} {self.message}\r")
sys.stdout.flush()
time.sleep(self.delay)
sys.stdout.write(f"\r{' ' * (len(self.message) + 2)}\r")
def __enter__(self):
"""Start the spinner"""
self.running = True
self.spinner_thread = threading.Thread(target=self.spin)
self.spinner_thread.start()
def __exit__(self, exc_type, exc_value, exc_traceback):
"""Stop the spinner"""
self.running = False
self.spinner_thread.join()
sys.stdout.write(f"\r{' ' * (len(self.message) + 2)}\r")
sys.stdout.flush()

59
scripts/token_counter.py
View File

@@ -1,59 +0,0 @@
import tiktoken
from typing import List, Dict
def count_message_tokens(messages : List[Dict[str, str]], model : str = "gpt-3.5-turbo-0301") -> int:
"""
Returns the number of tokens used by a list of messages.
Args:
messages (list): A list of messages, each of which is a dictionary containing the role and content of the message.
model (str): The name of the model to use for tokenization. Defaults to "gpt-3.5-turbo-0301".
Returns:
int: The number of tokens used by the list of messages.
"""
try:
encoding = tiktoken.encoding_for_model(model)
except KeyError:
logger.warn("Warning: model not found. Using cl100k_base encoding.")
encoding = tiktoken.get_encoding("cl100k_base")
if model == "gpt-3.5-turbo":
# !Node: gpt-3.5-turbo may change over time. Returning num tokens assuming gpt-3.5-turbo-0301.")
return count_message_tokens(messages, model="gpt-3.5-turbo-0301")
elif model == "gpt-4":
# !Note: gpt-4 may change over time. Returning num tokens assuming gpt-4-0314.")
return count_message_tokens(messages, model="gpt-4-0314")
elif model == "gpt-3.5-turbo-0301":
tokens_per_message = 4 # every message follows <|start|>{role/name}\n{content}<|end|>\n
tokens_per_name = -1 # if there's a name, the role is omitted
elif model == "gpt-4-0314":
tokens_per_message = 3
tokens_per_name = 1
else:
raise NotImplementedError(f"""num_tokens_from_messages() is not implemented for model {model}. See https://github.com/openai/openai-python/blob/main/chatml.md for information on how messages are converted to tokens.""")
num_tokens = 0
for message in messages:
num_tokens += tokens_per_message
for key, value in message.items():
num_tokens += len(encoding.encode(value))
if key == "name":
num_tokens += tokens_per_name
num_tokens += 3 # every reply is primed with <|start|>assistant<|message|>
return num_tokens
def count_string_tokens(string: str, model_name: str) -> int:
"""
Returns the number of tokens in a text string.
Args:
string (str): The text string.
model_name (str): The name of the encoding to use. (e.g., "gpt-3.5-turbo")
Returns:
int: The number of tokens in the text string.
"""
encoding = tiktoken.encoding_for_model(model_name)
num_tokens = len(encoding.encode(string))
return num_tokens

23
scripts/utils.py
View File

@@ -1,23 +0,0 @@
import yaml
from colorama import Fore
def clean_input(prompt: str=''):
try:
return input(prompt)
except KeyboardInterrupt:
print("You interrupted Auto-GPT")
print("Quitting...")
exit(0)
def validate_yaml_file(file: str):
try:
with open(file) as file:
yaml.load(file, Loader=yaml.FullLoader)
except FileNotFoundError:
return (False, f"The file {Fore.CYAN}`{file}`{Fore.RESET} wasn't found")
except yaml.YAMLError as e:
return (False, f"There was an issue while trying to read with your AI Settings file: {e}")
return (True, f"Successfully validated {Fore.CYAN}`{file}`{Fore.RESET}!")