Embodied Agents#
You can also check this cookbook in colab here
⭐ Star us on Github, join our Discord or follow our X
Philosophical Bits#
We believe the essence of intelligence emerges from its dynamic interactions with the external environment, where the use of various tools becomes a pivotal factor in its development and manifestation.
The EmbodiedAgent() in CAMEL is an advanced conversational agent that leverages code interpreters and tool agents (e.g., HuggingFaceToolAgent()) to execute diverse tasks efficiently. This agent represents a blend of advanced programming and AI capabilities, and is able to interact and respond within a dynamic environment.
Quick Start#
Let’s first play with a ChatAgent instance by simply initialize it with a system message and interact with user messages.
🕹 Step 0: Preparations#
[ ]:
%pip install "camel-ai==0.2.16"
[2]:
from camel.agents import EmbodiedAgent
from camel.generators import SystemMessageGenerator as sys_msg_gen
from camel.messages import BaseMessage as bm
from camel.types import RoleType
Setting Up API Keys#
You’ll need to set up your API keys for OpenAI.
[3]:
import os
from getpass import getpass
# Prompt for the API key securely
openai_api_key = getpass('Enter your API key: ')
os.environ["OPENAI_API_KEY"] = openai_api_key
Enter your API key: ··········
Alternatively, if running on Colab, you could save your API keys and tokens as Colab Secrets, and use them across notebooks.
To do so, comment out the above manual API key prompt code block(s), and uncomment the following codeblock.
⚠️ Don’t forget granting access to the API key you would be using to the current notebook.
[ ]:
# import os
# from google.colab import userdata
# os.environ["OPENAI_API_KEY"] = userdata.get("OPENAI_API_KEY")
🕹 Step 1: Define the Role#
We first need to set up the necessary information.
[4]:
# Set the role name and the task
role = 'Programmer'
task = 'Writing and executing codes.'
# Create the meta_dict and the role_tuple
meta_dict = dict(role=role, task=task)
role_tuple = (role, RoleType.EMBODIMENT)
The meta_dict and role_type will be used to generate the system message.
[5]:
# Generate the system message based on this
sys_msg = sys_msg_gen().from_dict(meta_dict=meta_dict, role_tuple=role_tuple)
🕹 Step 2: Initialize the Agent 🐫#
Based on the system message, we are ready to initialize our embodied agent.
[6]:
embodied_agent = EmbodiedAgent(system_message=sys_msg,
tool_agents=None,
code_interpreter=None,
verbose=True)
Be aware that the default argument values for tool_agents and code_interpreter are None, and the underlying code interpreter is using the SubProcessInterpreter(), which handles the execution of code in Python and Bash within a subprocess.
🕹 Step 3: Interact with the Agent with .step()#
Use the base message wrapper to generate the user message.
[7]:
usr_msg = bm.make_user_message(
role_name='user',
content=('1. write a bash script to install numpy. '
'2. then write a python script to compute '
'the dot product of [8, 9] and [5, 4], '
'and print the result. '
'3. then write a script to search for '
'the weather at london with wttr.in/london.'))
And feed that into your agents:
[8]:
response = embodied_agent.step(usr_msg)
> Explanation:
To accomplish the tasks you've outlined, I will perform the following actions:
1. **Write a Bash script** to install NumPy using `pip`. This will ensure that the necessary library is available for the Python script that follows.
2. **Write a Python script** to compute the dot product of the two lists `[8, 9]` and `[5, 4]`. The dot product is calculated as the sum of the products of the corresponding entries of the two sequences.
3. **Write a script** to fetch the weather information for London using the `wttr.in` service. This will provide a simple way to get weather data from the command line.
### Step 1: Bash Script to Install NumPy
I'll create a Bash script named `install_numpy.sh` that will check if `pip` is installed and then use it to install NumPy.
### Step 2: Python Script for Dot Product Calculation
I'll create a Python script named `dot_product.py` that will compute the dot product of the two lists and print the result.
### Step 3: Script to Fetch Weather Information
I'll create a Bash script named `get_weather.sh` that will use `curl` to fetch the weather information for London from `wttr.in`.
Now, let's implement these actions in code.
> Code:
# Step 1: Bash script to install NumPy
echo '#!/bin/bash' > install_numpy.sh
echo 'if command -v pip &> /dev/null; then' >> install_numpy.sh
echo ' echo "Installing NumPy..."' >> install_numpy.sh
echo ' pip install numpy' >> install_numpy.sh
echo 'else' >> install_numpy.sh
echo ' echo "pip is not installed. Please install pip first."' >> install_numpy.sh
echo 'fi' >> install_numpy.sh
chmod +x install_numpy.sh
# Step 2: Python script for dot product calculation
echo 'import numpy as np' > dot_product.py
echo 'a = np.array([8, 9])' >> dot_product.py
echo 'b = np.array([5, 4])' >> dot_product.py
echo 'dot_product = np.dot(a, b)' >> dot_product.py
echo 'print("Dot product:", dot_product)' >> dot_product.py
# Step 3: Bash script to fetch weather information
echo '#!/bin/bash' > get_weather.sh
echo 'curl wttr.in/london' >> get_weather.sh
chmod +x get_weather.sh
> Explanation:
### Explanation of the Code
1. **install_numpy.sh**: This script checks if `pip` is installed and installs NumPy if it is. It provides feedback if `pip` is not found.
2. **dot_product.py**: This Python script uses NumPy to calculate the dot product of two arrays and prints the result.
3. **get_weather.sh**: This script uses `curl` to fetch and display the weather information for London from `wttr.in`.
### Next Steps
I will execute these scripts in the appropriate order to ensure that NumPy is installed before running the Python script. After that, I will fetch the weather information for London.
Let's execute the installation script first.
> Code:
./install_numpy.sh
> Explanation:
After confirming that NumPy is installed, I will run the Python script:
> Code:
python3 dot_product.py
> Explanation:
Finally, I will run the weather script:
> Code:
./get_weather.sh
> Explanation:
Now, I will perform these actions.
The following bash code will run on your computer:
# Step 1: Bash script to install NumPy
echo '#!/bin/bash' > install_numpy.sh
echo 'if command -v pip &> /dev/null; then' >> install_numpy.sh
echo ' echo "Installing NumPy..."' >> install_numpy.sh
echo ' pip install numpy' >> install_numpy.sh
echo 'else' >> install_numpy.sh
echo ' echo "pip is not installed. Please install pip first."' >> install_numpy.sh
echo 'fi' >> install_numpy.sh
chmod +x install_numpy.sh
# Step 2: Python script for dot product calculation
echo 'import numpy as np' > dot_product.py
echo 'a = np.array([8, 9])' >> dot_product.py
echo 'b = np.array([5, 4])' >> dot_product.py
echo 'dot_product = np.dot(a, b)' >> dot_product.py
echo 'print("Dot product:", dot_product)' >> dot_product.py
# Step 3: Bash script to fetch weather information
echo '#!/bin/bash' > get_weather.sh
echo 'curl wttr.in/london' >> get_weather.sh
chmod +x get_weather.sh
Running code? [Y/n]:y
The following bash code will run on your computer:
./install_numpy.sh
Running code? [Y/n]:y
The following bash code will run on your computer:
python3 dot_product.py
Running code? [Y/n]:y
The following bash code will run on your computer:
./get_weather.sh
Running code? [Y/n]:y
======stderr======
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
100 8906 100 8906 0 0 9068 0 --:--:-- --:--:-- --:--:-- 9069
100 8906 100 8906 0 0 9068 0 --:--:-- --:--:-- --:--:-- 9069
==================
Under the hood, the agent will perform multiple actions within its action space in the OS to fulfill the user request. It will compose code to implement the action – no worries, it will ask for your permission before execution.
Ideally you should get the output similar to this, if you allow the agent to perform actions:
[ ]:
print(response.msg.content)
1. write a bash script to install numpy. 2. then write a python script to compute the dot product of [8, 9] and [5, 4], and print the result. 3. then write a script to search for the weather at london with wttr.in/london.
> Embodied Actions:
> Executed Results:
Executing code block 0: {
Installing NumPy...
Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (1.26.4)
}
Executing code block 1: {
The dot product of [8, 9] and [5, 4] is: 76
}
Executing code block 2: {
Fetching weather information for London...
Weather report: London
.-. Light rain shower
( ). 15 °C
(___(__) ↗ 12 km/h
‘ ‘ ‘ ‘ 3 km
‘ ‘ ‘ ‘ 3.4 mm
┌─────────────┐
┌──────────────────────────────┬───────────────────────┤ Thu 26 Sep ├───────────────────────┬──────────────────────────────┐
│ Morning │ Noon └──────┬──────┘ Evening │ Night │
├──────────────────────────────┼──────────────────────────────┼──────────────────────────────┼──────────────────────────────┤
│ _`/"".-. Patchy light d…│ _`/"".-. Patchy light d…│ _`/"".-. Light rain sho…│ _`/"".-. Patchy rain ne…│
│ ,\_( ). 16 °C │ ,\_( ). 17 °C │ ,\_( ). +14(12) °C │ ,\_( ). +12(10) °C │
│ /(___(__) ↗ 19-24 km/h │ /(___(__) ↗ 20-23 km/h │ /(___(__) ↗ 23-35 km/h │ /(___(__) ↗ 20-29 km/h │
│ ‘ ‘ ‘ ‘ 5 km │ ‘ ‘ ‘ ‘ 5 km │ ‘ ‘ ‘ ‘ 10 km │ ‘ ‘ ‘ ‘ 10 km │
│ ‘ ‘ ‘ ‘ 0.5 mm | 100% │ ‘ ‘ ‘ ‘ 0.2 mm | 100% │ ‘ ‘ ‘ ‘ 0.6 mm | 100% │ ‘ ‘ ‘ ‘ 0.0 mm | 65% │
└──────────────────────────────┴──────────────────────────────┴──────────────────────────────┴──────────────────────────────┘
┌─────────────┐
┌──────────────────────────────┬───────────────────────┤ Fri 27 Sep ├───────────────────────┬──────────────────────────────┐
│ Morning │ Noon └──────┬──────┘ Evening │ Night │
├──────────────────────────────┼──────────────────────────────┼──────────────────────────────┼──────────────────────────────┤
│ .-. Light rain │ _`/"".-. Patchy rain ne…│ \ / Sunny │ \ / Clear │
│ ( ). +9(7) °C │ ,\_( ). +11(9) °C │ .-. +11(8) °C │ .-. +9(8) °C │
│ (___(__) ↘ 18-26 km/h │ /(___(__) ↘ 23-31 km/h │ ― ( ) ― ↘ 16-24 km/h │ ― ( ) ― ↘ 10-16 km/h │
│ ‘ ‘ ‘ ‘ 9 km │ ‘ ‘ ‘ ‘ 10 km │ `-’ 10 km │ `-’ 10 km │
│ ‘ ‘ ‘ ‘ 0.8 mm | 100% │ ‘ ‘ ‘ ‘ 0.1 mm | 100% │ / \ 0.0 mm | 0% │ / \ 0.0 mm | 0% │
└──────────────────────────────┴──────────────────────────────┴──────────────────────────────┴──────────────────────────────┘
┌─────────────┐
┌──────────────────────────────┬───────────────────────┤ Sat 28 Sep ├───────────────────────┬──────────────────────────────┐
│ Morning │ Noon └──────┬──────┘ Evening │ Night │
├──────────────────────────────┼──────────────────────────────┼──────────────────────────────┼──────────────────────────────┤
│ \ / Sunny │ _`/"".-. Patchy rain ne…│ Cloudy │ \ / Partly Cloudy │
│ .-. +11(9) °C │ ,\_( ). +14(13) °C │ .--. 13 °C │ _ /"".-. 12 °C │
│ ― ( ) ― → 11-14 km/h │ /(___(__) → 12-14 km/h │ .-( ). → 6-10 km/h │ \_( ). ↗ 5-8 km/h │
│ `-’ 10 km │ ‘ ‘ ‘ ‘ 10 km │ (___.__)__) 10 km │ /(___(__) 10 km │
│ / \ 0.0 mm | 0% │ ‘ ‘ ‘ ‘ 0.1 mm | 100% │ 0.0 mm | 0% │ 0.0 mm | 0% │
└──────────────────────────────┴──────────────────────────────┴──────────────────────────────┴──────────────────────────────┘
Location: London, Greater London, England, UK [51.5073219,-0.1276474]
Follow @igor_chubin for wttr.in updates
(stderr: % Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
0 0 0 0 0 0 0 0 --:--:-- 0:00:01 --:--:-- 0
0 0 0 0 0 0 0 0 --:--:-- 0:00:02 --:--:-- 0
100 8977 100 8977 0 0 3495 0 0:00:02 0:00:02 --:--:-- 3495
)}
Executing code block 3: {
}