[DOCS][CLEANUP]

multion_example.py

@@ -0,0 +1,27 @@
+import os
+import threading
+from swarms.agents.multion_wrapper import MultiOnAgent
+
+def run_model(api_key):
+    model = MultiOnAgent(api_key=api_key, max_steps=500, url="https://x.com")
+    out = model.run(
+        """
+        click on the 'Tweet' button to start a new tweet and post it saying: $pip3 install swarms
+
+        """
+    )
+    print(out)
+
+# Create a list to store the threads
+threads = []
+
+# Run 100 instances using multithreading
+for _ in range(10):
+    api_key = os.getenv("MULTION_API_KEY")
+    thread = threading.Thread(target=run_model, args=(api_key,))
+    thread.start()
+    threads.append(thread)
+
+# Wait for all threads to finish
+for thread in threads:
+    thread.join()

buy_abunch_of_cybertrucks.py

@@ -0,0 +1,72 @@
+
+from swarms import Agent, AgentRearrange, OpenAIChat
+from swarms.agents.multion_wrapper import MultiOnAgent
+
+model = MultiOnAgent(
+    url="https://tesla.com",
+)
+
+
+llm = OpenAIChat()
+
+
+def browser_automation(task: str):
+    """
+    Run a task on the browser automation agent.
+
+    Args:
+        task (str): The task to be executed on the browser automation agent.
+    """
+    out = model.run(task)
+    return out
+
+
+# Purpose = To detect email spam using three different agents
+agent1 = Agent(
+    agent_name="CyberTruckBuyer1",
+    system_prompt="Find the best deal on a Cyber Truck and provide your reasoning",
+    llm=llm,
+    max_loops=1,
+    # output_type=str,
+    metadata="json",
+    function_calling_format_type="OpenAI",
+    function_calling_type="json",
+    streaming_on=True,
+    tools=[browser_automation],
+)
+
+agent2 = Agent(
+    agent_name="CyberTruckBuyer2",
+    system_prompt="Find the best deal on a Cyber Truck and provide your reasoning",
+    llm=llm,
+    max_loops=1,
+    # output_type=str,
+    metadata="json",
+    function_calling_format_type="OpenAI",
+    function_calling_type="json",
+    streaming_on=True,
+    tools=[browser_automation],
+)
+
+agent3 = Agent(
+    agent_name="CyberTruckBuyer3",
+    system_prompt="Find the best deal on a Cyber Truck and provide your reasoning",
+    llm=llm,
+    max_loops=1,
+    # output_type=str,
+    metadata="json",
+    function_calling_format_type="OpenAI",
+    function_calling_type="json",
+    streaming_on=True,
+    tools=[browser_automation],
+)
+
+swarm = AgentRearrange(
+    flow="CyberTruckBuyer1 -> CyberTruckBuyer2 -> CyberTruckBuyer3",
+    agents=[agent1, agent2, agent3],
+    logging_enabled=True,
+    max_loops=1,
+)
+
+# Run all the agents
+swarm.run("Let's buy a cyber truck")

docs/swarms/structs/agent_rearrange.md

@@ -1,5 +1,4 @@
-# Documentation for `AgentRearrange` Class
------
+# `AgentRearrange` Class
 
 The `AgentRearrange` class represents a swarm of agents for rearranging tasks. It allows you to create a swarm of agents, add or remove agents from the swarm, and run the swarm to process tasks based on a specified flow pattern.
 
@@ -119,8 +118,8 @@ from typing import List
 
 # Initialize the director agent
 director = Agent(
-    agent_name="Director",
-    system_prompt="Directs the tasks for the workers",
+    agent_name="Accounting Director",
+    system_prompt="Directs the accounting tasks for the workers",
     llm=Anthropic(),
     max_loops=1,
     dashboard=False,
@@ -128,13 +127,13 @@ director = Agent(
     verbose=True,
     stopping_token="<DONE>",
     state_save_file_type="json",
-    saved_state_path="director.json",
+    saved_state_path="accounting_director.json",
 )
 
 # Initialize worker 1
 worker1 = Agent(
-    agent_name="Worker1",
-    system_prompt="Generates a transcript for a youtube video on what swarms are",
+    agent_name="Accountant 1",
+    system_prompt="Processes financial transactions and prepares financial statements",
     llm=Anthropic(),
     max_loops=1,
     dashboard=False,
@@ -142,13 +141,13 @@ worker1 = Agent(
     verbose=True,
     stopping_token="<DONE>",
     state_save_file_type="json",
-    saved_state_path="worker1.json",
+    saved_state_path="accountant1.json",
 )
 
 # Initialize worker 2
 worker2 = Agent(
-    agent_name="Worker2",
-    system_prompt="Summarizes the transcript generated by Worker1",
+    agent_name="Accountant 2",
+    system_prompt="Performs audits and ensures compliance with financial regulations",
     llm=Anthropic(),
     max_loops=1,
     dashboard=False,
@@ -156,22 +155,22 @@ worker2 = Agent(
     verbose=True,
     stopping_token="<DONE>",
     state_save_file_type="json",
-    saved_state_path="worker2.json",
+    saved_state_path="accountant2.json",
 )
 
 # Create a list of agents
 agents = [director, worker1, worker2]
 
 # Define the flow pattern
-flow = "Director -> Worker1 -> Worker2"
+flow = "Accounting Director -> Accountant 1 -> Accountant 2"
 
 # Using AgentRearrange class
 agent_system = AgentRearrange(agents=agents, flow=flow)
-output = agent_system.run("Create a format to express and communicate swarms of llms in a structured manner for youtube")
+output = agent_system.run("Process monthly financial statements")
 print(output)
 
 # Using rearrange function
-output = rearrange(agents, flow, "Create a format to express and communicate swarms of llms in a structured manner for youtube")
+output = rearrange(agents, flow, "Process monthly financial statements")
 print(output)
 
 ```

example.py

@@ -1,4 +1,10 @@
 from swarms import Agent, OpenAIChat
+from swarms.agents.multion_wrapper import MultiOnAgent
+
+model = MultiOnAgent(
+    url="https://tesla.com",
+)
+
 
 
 def calculate_profit(revenue: float, expenses: float):
@@ -28,6 +34,16 @@ def generate_report(company_name: str, profit: float):
     """
     return f"The profit for {company_name} is ${profit}."
 
+def browser_automation(task: str = None):
+    """
+    Run a task on the browser automation agent.
+
+    Args:
+        task (str): The task to be executed on the browser automation agent.
+    """
+    out = model.run(task)
+    return out
+
 
 # Initialize the agent
 agent = Agent(
@@ -46,8 +62,9 @@ agent = Agent(
     # tools=[calculate_profit, generate_report],
     # docs_folder="docs",
     # pdf_path="docs/accounting_agent.pdf",
+    # tools=[browser_automation],
 )
 
 agent.run(
-    "We're the Swarm Corporation, our total revenue is $100,000 and our total expenses are $50,000, is our revenue good?"
+    "Calculate the profit for Tesla with a revenue of $100,000 and expenses of $50,000."
 )

phising_my_parents.py

@@ -0,0 +1,53 @@
+import os
+
+from swarms import OpenAIChat, Agent, AgentRearrange
+
+# Purpose = To detect email spam using three different agents
+agent1 = Agent(
+    agent_name="SpamDetector1",
+    system_prompt="Detect if the email is spam or not, and provide your reasoning",
+    llm=OpenAIChat(openai_api_key=os.getenv("OPENAI_API_KEY")),
+    max_loops=1,
+    output_type=str,
+    # tools=[],
+    metadata="json",
+    function_calling_format_type="OpenAI",
+    function_calling_type="json",
+    streaming_on=True,
+)
+
+agent2 = Agent(
+    agent_name="SpamDetector2",
+    system_prompt="Detect if the email is spam or not, and provide your reasoning",
+    llm=OpenAIChat(openai_api_key=os.getenv("OPENAI_API_KEY")),
+    max_loops=1,
+    output_type=str,
+    # tools=[],
+    metadata="json",
+    function_calling_format_type="OpenAI",
+    function_calling_type="json",
+    streaming_on=True,
+)
+
+agent3 = Agent(
+    agent_name="SpamDetector3",
+    system_prompt="Detect if the email is spam or not, and provide your reasoning",
+    llm=OpenAIChat(openai_api_key=os.getenv("OPENAI_API_KEY")),
+    max_loops=1,
+    output_type=str,
+    # tools=[],
+    metadata="json",
+    function_calling_format_type="OpenAI",
+    function_calling_type="json",
+    streaming_on=True,
+)
+
+swarm = AgentRearrange(
+    flow="SpamDetector1 -> SpamDetector2 -> SpamDetector3",
+    agents=[agent1, agent2, agent3],
+    logging_enabled=True,
+    max_loops=1,
+)
+
+# Run all the agents
+swarm.run("Find YSL bag with the biggest discount")

scripts/cleanup/json_log_cleanup.py

@@ -31,4 +31,4 @@ def cleanup_json_logs(name: str = None):
 
 
 # Call the function
-cleanup_json_logs("artifacts_five")
+cleanup_json_logs("artifacts_six")

swarms/agents/multion_wrapper.py

@@ -0,0 +1,75 @@
+# pip3 install multion
+# pip3 install swarms
+from multion.client import MultiOn
+import os
+from swarms.models.base_llm import BaseLLM
+
+
+def check_multion_api_key():
+    """
+    Checks if the MultiOn API key is available in the environment variables.
+
+    Returns:
+        str: The MultiOn API key.
+    """
+    api_key = os.getenv("MULTION_API_KEY")
+    return api_key
+
+
+class MultiOnAgent(BaseLLM):
+    """
+    Represents an agent that interacts with the MultiOn API to run tasks on a remote session.
+
+    Args:
+        api_key (str): The API key for accessing the MultiOn API.
+        url (str): The URL of the remote session.
+        *args: Variable length argument list.
+        **kwargs: Arbitrary keyword arguments.
+
+    Attributes:
+        client (MultiOn): The MultiOn client instance.
+        url (str): The URL of the remote session.
+        session_id (str): The ID of the current session.
+
+    Methods:
+        run: Runs a task on the remote session.
+    """
+
+    def __init__(
+        self,
+        name: str = None,
+        system_prompt: str = None,
+        api_key: str = check_multion_api_key,
+        url: str = "https://huggingface.co/papers",
+        max_steps: int = 1,
+        *args,
+        **kwargs,
+    ):
+        super().__init__(*args, **kwargs)
+        self.name = name
+        self.client = MultiOn(api_key=api_key)
+        self.url = url
+        self.system_prompt = system_prompt
+        self.max_steps = max_steps
+
+    def run(self, task: str, *args, **kwargs):
+        """
+        Runs a task on the remote session.
+
+        Args:
+            task (str): The task to be executed on the remote session.
+            *args: Variable length argument list.
+            **kwargs: Arbitrary keyword arguments.
+        """
+        response = self.client.browse(
+            cmd=task,
+            url=self.url,
+            local=True,
+            max_steps=self.max_steps,
+        )
+        
+        # response = response.json()
+        
+        # print(response.message)
+        return str(response.message)
+