Understanding program structure, control flow, and problem decomposition is essential in programming education, but many students find algorithmic thinking challenging. Conventional educational approaches frequently prioritize syntax at the expense of logical reasoning and provide minimal personalized feedback. This paper outlines the design and initial assessment of FlowBro, an AI-driven educational tool intended to enhance algorithmic thinking in early programming education. The tool integrates a flowchart visualization, sequential code simulation, and an intelligent assistant that offers contextual hints, debugging tips, and educational support while maintaining a focus on minimalism by not disclosing complete solutions. The system enables students to engage with Python code, follow its execution, and obtain tailored assistance aligned with their specific objectives. A qualitative evaluation with educators in the field revealed the tool's potential advantages, especially in large classes and self-directed learning contexts, while also pointing out challenges concerning the reliability of AI feedback and the necessity for more extensive testing. This study presents a streamlined framework for merging AI with visual learning in programming education, while also suggesting pathways for its future enhancement and empirical testing.