Agent Design and Tool Definition

Design the core structure of your OpenAI agent. Define the roles and goals for a 'Code Architect' agent and a 'Code Refiner' agent. Create the initial tool definitions for simulating a code linter/static analyzer (e.g., 'check_code_style', 'identify_security_vulnerabilities') and a mock 'apply_fix_to_code' tool, ensuring they adhere to MCP principles for input/output schemas. Use the OpenAI Agents SDK's `Tool` or `tool` decorator. Explain how these tools will interact with the agent's thought process.