Getting Started with AI Foundry Agents

Just my experiments with AI foundry agents.

Features

• I added a devcontainer with the needed tools for execute/debug this examples.

Getting Started

1. Clone the repository:

   git clone <repo-url>
   cd getting-started-with-ai-foundry-agents

2. Open the devcontainer or Install dependencies:
   • For Python:

     make install

3. Modify the .env file The endpoint can be found in the Azure Foundry portal The format is : https://.services.ai.azure.com/api/projects/.

For the example 01 and 02 you will need to create a "Grounding with Bing Custom Search" resource, and add a connection to that service in the AI Foundry project.

4. Run or debug your agent:
   • Python:

     uv run agent_example_xx.py

Examples

• Example 0: This is the simplest example of using the Azure AI Agent service to create an agent, send a message, and receive a response.
• Example 1: Example of using the Azure AI Agent service to create an agent, and add a knowledge tool (Bing Custom Search) to it. Send a message, and receive a response.
• Example 2: Example of using the Azure AI Agent service to create two connected agents. One of the agents has a knowledge tool (Bing Custom Search). The other agent is an orchestrator that can call the first agent.
• Example 3: Create a Semantic Kernel agent that interacts with the Azure AI Foundry Agent service.
• Example 4: Semantic Kernel agent with a SK plugin used by the agent. The agent uses Azure AI Foundry Agent service to generate the answer and access to the plugin.
• Example 5 and 6: Creates a Semantic Kernel Group Chat Orchestration where two agents chat. We are defining a Termination Strategy (when one of the agents approves the work of the other one), and a callback function to log the conversation.

Contributing

Contributions are welcome! Please open issues or submit pull requests.

License

This project is licensed under the MIT License.

---

May the Force be with your code!