SOP-GPT: A Framework for AI Agents Based on Artificial Intelligence-Generated Content

The advancements in automated problem-solving were explored by the agents of Artificial Intelligence (AI)Generated Content (AIGC). Although existing AIGC-based AI Agent systems can solve simple tasks, it is complicated to handle complex tasks due to inconsistent work logic. In this research, the operating mechanism of AI Agents was dissected to elucidate why AIGC is suitable as the foundation for AI Agents. Based on this technology, we proposed an AI Agent framework named SOP-GPT which has integrated the concept of Standard Operating Procedures (SOP) from human workflow into AI Agents. SOP-GPT comprises three main elements: process, role, and skills. By combining the generative capability of AIGC with human SOP mechanisms, the quality of generation was enhanced to meet the expectations of human work requirements. We further explored how to assemble an AIGC engineering process for the division of labor and collaboration methods between various functional departments or units in enterprises in this framework. This research results provide a perspective for human management experience to enhance the capabilities of AI Agents and provide direction for future research.