Summary

Adds AGENTS.md to provide coding guidelines for AI agents (e.g., GitHub Copilot, Cursor, Codex) working in this repository.

What's included

• Dependencies: Poetry, Pydantic, and Pydantic-settings usage guidelines
• Testing Guidelines: pytest conventions, mocking with monkeypatch/pytest-mock, and honesty rules
• Make Targets: Workflow automation via make targets
• Notes: Python 3.12+ requirement, pyproject.toml/poetry.lock management
• Coding Conventions:
  • Pydantic Field(description=...) on every field
  • camelCase alias convention with populate_by_name=True and alias_generator=to_camel
  • No model_config class attribute — use keyword arguments on base class
  • Explicit from x import y import style
  • Required if __name__ == "__main__": main() block in every test file