Add AG2 framework examples with native Bedrock support

[faridun-ag2]

Summary

• Add AG2 (formerly AutoGen) to agents-and-function-calling/open-source-agents/ with 3 Jupyter notebooks demonstrating native Amazon Bedrock integration
• ag2-single-agent-bedrock.ipynb — basic conversational agent
• ag2-tool-use-bedrock.ipynb — function calling with dual registration pattern (register_for_llm + register_for_execution)
• ag2-multi-agent-bedrock.ipynb — GroupChat multi-agent orchestration with automatic speaker selection

Why AG2?

AG2 has native Bedrock support via LLMConfig(api_type="bedrock") — no LangChain wrapper or OpenAI API key required. It uses the standard AWS credential chain (IAM roles, env vars, ~/.aws/credentials) for authentication.

Feature	Other Frameworks	AG2
Bedrock access	Requires ChatBedrock wrapper (LangChain) or boto3 directly	Native LLMConfig(api_type="bedrock")
Authentication	Manual boto3 session setup	AWS credential chain (automatic)
Multi-agent	Framework-specific patterns	GroupChat with auto speaker selection
Package	Various	pip install ag2 (500K+ monthly downloads)

Test plan

• All 3 notebooks are valid JSON
• Code validated: imports, LLMConfig, agents, tool registration, GroupChat all work without errors
• Notebooks run end-to-end with Bedrock model access enabled
• pip install ag2[openai] installs successfully
• Existing framework examples untouched
• Parent README updated with AG2 entry
• Markdown docs with HTML headers, YAML tags, and "Open in GitHub" CTA links

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[faridun-ag2]

Hi @akinfaa! 👋 Hope you're doing well. Could you please take a look at this PR when you get a chance? It adds AG2 framework examples with native Bedrock support to the open-source agents section. Would really appreciate your review. Thanks!