This thesis describes the design, implementation and validation of an intelligent agentic workflow that acts as an automated bridge between natural language security requirements and the Otupy implementation of the OpenC2 standard. The main objective is to create a robust system for translating human intent into executable firewall policies within the Otupy framework. The proposed work aims to bridge the gap between high-level natural language requests, which are often ambiguous, and the rigorous, structured input format required by security orchestration tools. The system is specifically designed to manage the translation process, interpreting user inputs and converting them into precise Otupy commands for firewall configuration that are compliant with the target device specifications.