Agents SDK Mini Projects

🤖 A curated collection of AI agent mini-projects built with the OpenAI Agents SDK and Google's Gemini models.

A hands-on exploration of AI agent implementations, demonstrating various configurations, use cases, and integration patterns. Each project is a standalone application showcasing different aspects of conversational AI development.

---

📁 Projects Overview

Core Assistants

Project	Description	Version	Status
assistant/	General-purpose AI assistant with friendly, brief responses	v2.0.0	✅ Complete
fast_assistant/	Optimized for quick, concise answers	v2.0.0	✅ Complete
first_agent/	Introductory agent - perfect for learning AI agent basics	v2.0.0	✅ Complete

Specialized Assistants

Project	Description	Version	Status
gemini_assistant/	Concise responses using Gemini 2.0 Flash	v1.0.0	✅ Complete
local_assistant/	Locally-running assistant with full customization	v1.0.0	✅ Complete
news_agent/	News aggregation and summarization agent	-	🚧 In Progress
support_desk/	Customer support desk simulation agent	-	🚧 In Progress
task_manager/	Task management and productivity agent	-	🚧 In Progress
time_agent/	Time-related queries and scheduling agent	-	🚧 In Progress

---

🎯 Purpose

This collection serves as:

• Learning Resource: Understand AI agent architecture and implementation patterns
• Reference Code: Quick starting point for building your own agents
• Experimentation Platform: Test different models, configurations, and instructions
• Portfolio: Showcase practical AI integration skills

---

🚀 Quick Start

Prerequisites

1. Python 3.13+ installed
2. uv package manager (recommended)
3. Gemini API Key from Google AI Studio

Setup Any Project

# 1. Clone the repository
git clone https://github.com/sheikh-mohammad/agents-sdk-mini-projects.git
cd agents-sdk-mini-projects

# 2. Navigate to a project
cd assistant

# 3. Install dependencies
uv sync

# 4. Configure environment
cp .env.example .env

# 5. Edit .env and add your API key
# GEMINI_API_KEY=your_actual_api_key_here

# 6. Run the project
uv run python main.py

Alternative: Using pip

# Create virtual environment
python -m venv .venv
source .venv/bin/activate  # Windows: .venv\Scripts\activate

# Install dependencies
pip install openai openai-agents python-dotenv

# Run the project
python main.py

---

🏗️ Project Structure

Each project follows a consistent structure:

project_name/
├── main.py              # Main agent implementation
├── pyproject.toml       # Dependencies and project metadata
├── .env.example         # Environment variable template
├── .gitignore           # Git ignore rules
└── README.md            # Project-specific documentation

Typical Agent Implementation

from agents import Agent, Runner, OpenAIChatCompletionsModel
from openai import AsyncOpenAI
from dotenv import load_dotenv
import os

load_dotenv()

# Initialize client with Gemini API
client = AsyncOpenAI(
    api_key=os.getenv("GEMINI_API_KEY"),
    base_url="https://generativelanguage.googleapis.com/v1beta/openai/",
)

# Configure model
model = OpenAIChatCompletionsModel(
    model="gemini-2.0-flash", openai_client=client
)

# Create agent with custom instructions
agent = Agent(
    name="My Assistant",
    instructions="You are a helpful assistant.",
    model=model,
)

# Run the agent
result = Runner.run_sync(agent, "Your prompt here")
print(result.final_output)

---

🔧 Configuration

Environment Variables

Variable	Description	Required
GEMINI_API_KEY	Your Google Gemini API key	Yes

Available Models

Model	Description	Best For
gemini-2.0-flash	Fast, efficient model	General tasks, quick responses
gemini-2.5-flash	Latest flash model	Improved accuracy and speed

---

📚 Technologies Used

Technology	Purpose
OpenAI Agents SDK	Agent framework and runtime
Google Gemini API	LLM backend
uv	Fast Python package management
python-dotenv	Environment variable management
OpenAI Python Client	API client compatibility

---

📖 Learning Resources

• OpenAI Agents SDK Documentation
• Google AI Studio
• Gemini API Documentation
• Getting Your Gemini API Key

---

⚠️ License and Usage Terms

This is a personal project collection created by Sheikh Mohammad Ahmed.

Permitted Use

• ✅ Review: Study the code for learning purposes
• ✅ Inspiration: Use as inspiration for your own projects
• ✅ Educational Use: Understand AI agent implementation patterns

Restrictions

• ❌ No Direct Copying: Copying code or ideas without attribution is not allowed
• ❌ No Plagiarism: Presenting this work as your own is prohibited
• ❌ No Commercial Use: Commercial use requires explicit permission

Attribution Requirement

If you draw inspiration from this collection, you must:

1. Credit Sheikh Mohammad Ahmed as the original author
2. Link to this repository: https://github.com/sheikh-mohammad/agents-sdk-mini-projects
3. State that your work was inspired by this project collection

Example Attribution:

Inspired by agents-sdk-mini-projects by Sheikh Mohammad Ahmed
https://github.com/sheikh-mohammad/agents-sdk-mini-projects

See LICENSE for complete terms.

---

🤝 Contributing

This is a personal learning project. However, feel free to:

• ⭐ Star the repository if you find it useful
• 📢 Share with others learning AI development
• 💡 Open issues for suggestions or improvements

---

📬 Contact

• Author: Sheikh Mohammad Ahmed
• GitHub: @sheikh-mohammad
• Repository: agents-sdk-mini-projects

---

Copyright © 2026 Sheikh Mohammad Ahmed. All Rights Reserved.