AI Agents: Workshop @ Hyperskill 2025-03-27

Table of Contents

Assignment	Description	Make Target	Script File
1	Basic Agent Setup	make setup-langgraph	langgraph/langgraph_agent.py
2	Multi-Agent Conversation	make setup-autogen	autogen/autogen_agent.py
3	Specialized Crew Creation	make setup-crewai	crewai/crewai_agent.py
4	Vector Database with ChromaDB	make setup-rag	rag/rag_agent.py
5	Browser Use	make setup-browseruse	browseruse/browseruse_agent.py
6	Voice Assistant with Vapi	make setup-voice	voice/voice_agent.py

Installation

Prerequisites

• Python 3.11.x (tested with 3.11.7)
• pip (Python package installer)
• virtualenv
• make (for Linux/macOS users)

Setup

For Linux/macOS Users

1. Clone the repository:

   git clone https://github.com/yourusername/workshop-ai-agents-20250327.git
   cd workshop-ai-agents-20250327

2. Setup environment for an assignment:

   make setup-langgraph  # Replace langgraph with the assignment number

3. Activate the environment:

   source .venv-langgraph/bin/activate  # Replace langgraph with the assignment number

4. Run the assignment: Follow the specific instructions provided for each assignment.

5. Clean environments:

   make clean

For Windows Users

1. Clone the repository:

   git clone https://github.com/yourusername/workshop-ai-agents-20250327.git
   cd workshop-ai-agents-20250327

2. Create and activate a virtual environment:

   python -m venv .venv-langgraph  # Replace langgraph with the assignment number
   .venv-langgraph\Scripts\activate

3. Install requirements:

   pip install --upgrade pip
   pip install -r langgraph/requirements.txt # Replace langgraph with the assignment number

4. Run the assignment: Follow the specific instructions provided for each assignment.

5. Clean environments manually:

   rmdir /s /q .venv-langgraph  # Replace langgraph with the assignment number

Optional: Makefile Alternative for Windows

Consider creating a make.bat file with equivalent commands for easier setup on Windows.

@echo off
:: Usage: make.bat setup-langgraph
if "%1"=="setup-langgraph" (
    python -m venv .venv-langgraph
    .venv-langgraph\Scripts\activate
    pip install --upgrade pip
    pip install -r requirements-langgraph.txt
)
if "%1"=="clean" (
    rmdir /s /q .venv-*
)

Recommended Workflow (per assignment)

Setup

make setup-langgraph  # Setup for Assignment 1
source .venv-langgraph/bin/activate
make dev       # Install development tools (optional)

Running the Assignment

python langgraph/langgraph_agent.py  # Assignment 1
deactivate

Development & Testing

To install development tools, which include pytest, use:

make dev # Recommended

or manually:

pip install -r requirements-dev.txt

Testing

To run tests for the assignments, use pytest. Ensure you have installed the development requirements (which include pytest) either by running make dev or manually with:

pip install -r requirements-dev.txt

Run tests with:

pytest

Using AI models:

OpenAI

from langchain_openai import ChatOpenAI

OPENAI_API_KEY="sk-..."  # Obtain from https://platform.openai.com/api-keys
llm = ChatOpenAI(model="gpt-4o-mini", api_key=OPENAI_API_KEY, temperature=0)

GitHub's OpenAI

from langchain_openai import ChatOpenAI

GITHUB_TOKEN = "..." # Obtain from: https://github.com/marketplace/models/azure-openai/gpt-4o/playground
llm = ChatOpenAI(model="gpt-4o-mini", api_key=GITHUB_TOKEN, base_url="https://models.inference.ai.azure.com/", temperature=0)

OpenRounter's OpenAI

from langchain_openai import ChatOpenAI

OPENROUTER_API_KEY = "..." # Obtain from: https://openrouter.ai/settings/keys
llm = ChatOpenAI(model="gpt-4o-mini", api_key=OPENROUTER_API_KEY, base_url="https://openrouter.ai/api/v1", temperature=0)

Google Gemini

from langchain_community import ChatGoogleGenerativeAI

GOOGLE_API_KEY = "..." # Obtain from: https://aistudio.google.com/apikey
llm = ChatGoogleGenerativeAI(model="gemini-2.0-flash-lite", api_key=GOOGLE_API_KEY, temperature=0)

Usage

python langgraph/langgraph_agent.py
python autogen/autogen_agent.py
python crewai/crewai_agent.py
python rag/rag_agent.py
python browseruse/browseruse_agent.py
python voice/voice_agent.py

Notes on Project Structure

This workshop contains independent assignments, each demonstrating a different AI agent framework:

Assignment	Agent Framework
01	LangGraph
02	AutoGen
03	CrewAI
04	LangChain + ChromaDB
05	Browser-Use Agent
06	Vapi Voice Assistant

⚠ Important: Each assignment may have incompatible dependency versions due to rapidly evolving AI agent libraries.

To avoid conflicts, it is recommended to use a separate virtual environment for each assignment.

Code Formatting

This project uses Black for code formatting and isort for import sorting. The configuration is in pyproject.toml.

VSCode Setup

1. Install the required extensions:

   • Black Formatter (ms-python.black-formatter)
   • Python (ms-python.python)

2. Select the correct Python interpreter:

   • Open any Python file
   • Press Cmd+Shift+P (Mac) or Ctrl+Shift+P (Windows)
   • Type "Python: Select Interpreter"
   • Choose the interpreter from your current assignment's environment (e.g., .venv-langgraph/bin/python)

3. The following settings are already configured in the project:

   • Format on save is enabled
   • Black is set as the default formatter
   • Import sorting is configured to use Google style

IDE Setup

This project uses standard configuration files to ensure consistent formatting across different IDEs:

• .editorconfig: Basic editor settings (works in most IDEs)
• pyproject.toml: Black and isort configuration

VSCode

1. Install extensions:
   • Black Formatter (ms-python.black-formatter)
   • Python (ms-python.python)
   • EditorConfig (editorconfig.editorconfig)

PyCharm

1. Enable EditorConfig:
   • Settings → Editor → Code Style
   • Check "Enable EditorConfig support"
2. Set Black as formatter:
   • Settings → Tools → Actions on Save
   • Enable "Reformat code"
   • Set Python formatter to "Black"

Sublime Text

1. Install packages:
   • EditorConfig
   • Python Black
2. Enable format on save:
   • Preferences → Package Settings → Python Black → Settings
   • Add: "on_save": true

Command Line Usage

To format files from the command line, first activate your virtual environment:

source .venv-langgraph/bin/activate   # Match the assignment folder name (e.g., autogen, rag, etc.)
black your_file.py                    # Format a single file
black .                               # Format all Python files in the project
isort .                               # Sort imports in all Python files

Making Changes

1. Before starting work:

   git pull                   # Get latest changes
   source .venv-langgraph/bin/activate

2. While working:

   • Let your IDE format on save (VSCode, PyCharm)
   • Or format manually:

     black .
     isort .

3. Before committing:

   # Format all files
   black .
   isort .
   
   # Review changes
   git status
   git diff
   
   # Important: Always commit formatting configs together
   # - .editorconfig: Basic editor settings
   # - pyproject.toml: Black and isort settings
   # This ensures all developers use the same formatting
   
   # Commit
   git add .
   git commit -m "descriptive message"
   git push