chore: Add Azure OpenAI TAP example using LLMTarget

[rdheekonda]

Summary

Adds a new TAP attack example for Azure OpenAI using the built-in LLMTarget and renames the existing task-based example for clarity.

Changes

• New: tap_azure_openai_llm_target.ipynb - Uses built-in LLMTarget with LiteLLM (recommended approach)
• Renamed: tap_azure_openai_custom_target.ipynb → tap_azure_openai_task_target.ipynb - Uses @dn.task + CustomTarget for full control
• Removed all secrets/API keys, replaced with placeholders
• Added clear distinction in markdown headers explaining each approach
• Cleaned up comments to be concise

Approach Comparison

LLMTarget (Recommended)

• Simplest approach
• Uses LiteLLM for automatic Azure OpenAI setup
• No manual task creation needed
• Best for standard LLM endpoints

Task-based CustomTarget

• Full control over API client
• Custom request/response handling
• Best when you need specific logic

Both examples demonstrate TAP attack with transforms on Azure OpenAI.

Generated Summary:

• Added a new Jupyter notebook tap_azure_openai_llm_target.ipynb for demonstrating TAP attacks using the built-in LLMTarget with Azure OpenAI.
• The new notebook provides a simplified approach for standard LLM endpoints, handling API client setup and message conversion automatically.
• Renamed and updated tap_azure_openai_custom_target.ipynb to tap_azure_openai_task_target.ipynb, reflecting its focus on a task-based method for more control over API interactions.
• Improved documentation and readability in the task-based notebook, highlighting the distinction between basic and advanced usage.
• Both notebooks now reflect current Azure OpenAI configuration practices and coding style.

This summary was generated with ❤️ by rigging