Python Versions Discord Twitter

OpenAGI aims to make human-like agents accessible to everyone, thereby paving the way towards open agents and, eventually, AGI for everyone. We strongly believe in the transformative power of AI and are confident that this initiative will significantly contribute to solving many real-life problems. Currently, OpenAGI is designed to offer developers a framework for creating autonomous human-like agents.

Installation

# For Mac and Linux users
python3 -m venv venv
source venv/bin/activate
# For Windows users
python -m venv venv
venv/scripts/activate

pip install openagi

git clone https://github.com/aiplanethub/openagi.git
pip install -e .

Example (Manual Agent Execution)

from openagi.agent import Admin
from openagi.planner.task_decomposer import TaskPlanner
from openagi.actions.tools.ddg_search import DuckDuckGoSearch
from openagi.llms.openai import OpenAIModel
from openagi.worker import Worker
plan = TaskPlanner(human_intervene=False)
action = DuckDuckGoSearch
import os
os.environ['OPENAI_API_KEY'] = "sk-xxxx"
config = OpenAIModel.load_from_env_config()
llm = OpenAIModel(config=config)
trip_plan = Worker(
 role="Trip Planner",
 instructions="""
 User loves calm places, suggest the best itinerary accordingly.
 """,
 actions=[action],
 max_iterations=10)
admin = Admin(
 llm=llm,
 actions=[action],
 planner=plan,
)
admin.assign_workers([trip_plan])
res = admin.run(
 query="Give me total 3 Days Trip to San francisco Bay area",
 description="You are a knowledgeable local guide with extensive information about the city, it's attractions and customs",
)
print(res)

Example (Autonomous Multi-Agent Execution)

from openagi.planner.task_decomposer import TaskPlanner
from openagi.actions.tools.tavilyqasearch import TavilyWebSearchQA
from openagi.agent import Admin
from openagi.llms.gemini import GeminiModel
import os
os.environ['TAVILY_API_KEY'] = "<replace with Tavily key>"
os.environ['GOOGLE_API_KEY'] = "<replace with Gemini key>"
os.environ['Gemini_MODEL'] = "gemini-1.5-flash"
os.environ['Gemini_TEMP'] = "0.1"
gemini_config = GeminiModel.load_from_env_config()
llm = GeminiModel(config=gemini_config)
# define the planner
plan = TaskPlanner(autonomous=True,human_intervene=True)
admin = Admin(
 actions = [TavilyWebSearchQA],
 planner = plan,
 llm = llm,
)
res = admin.run(
 query="I need cricket updates from India vs Sri lanka 2024 ODI match in Sri Lanka",
 description=f"give me the results of India vs Sri Lanka ODI and respective Man of the Match",
)
print(res)

Long Term Memory like never before

import os
from openagi.agent import Admin
from openagi.llms.openai import OpenAIModel
from openagi.memory import Memory
from openagi.planner.task_decomposer import TaskPlanner
from openagi.worker import Worker
from openagi.actions.tools.ddg_search import DuckDuckGoSearch
memory = Memory(long_term=True)
os.environ['OPENAI_API_KEY'] = "-"
config = OpenAIModel.load_from_env_config()
llm = OpenAIModel(config=config)
web_searcher = Worker(
 role="Web Researcher",
 instructions="""
 You are tasked with conducting web searches using DuckDuckGo.
 Find the most relevant and accurate information based on the user's query.
 """,
 actions=[DuckDuckGoSearch], 
)
admin = Admin(
 actions=[DuckDuckGoSearch],
 planner=TaskPlanner(human_intervene=False),
 memory=memory,
 llm=llm,
)
admin.assign_workers([web_searcher])
query = input("Enter your search query: ")
description = f"Find accurate and relevant information for the query: {query}"
res = admin.run(query=query,description=description)
print(res)

Documentation

Use Cases:

Get in Touch

Contribution guidelines