The article explores a light-hearted comparison of various AI coding agents by sorting them into Hogwarts houses based on their performance in solving programming challenges from this year’s Advent of Code. Each agent was tasked with completing twelve executable programs under specific constraints. All agents finished within 20 minutes, but none achieved perfect accuracy. Codex and Gemini were closely matched in their performance metrics, while Claude took longer due to struggles with a specific problem. Mistral, on the other hand, approached solutions with an over-engineering mindset, utilizing classes in every implementation. The authors categorized each agent into coding archetypes based on their solution styles. Most agents fell under the "Pragmatist" category, producing clear and efficient code. Claude exhibited tendencies of an "Over-Engineer," which aligned with its higher complexity score. Codex displayed characteristics of a "Wizard," focusing on cleverness over comments, while Gemini showed traits of a "Professor," with extensive annotations throughout its code. Mistral’s approach was heavily focused on over-engineering, which hindered its practicality. In a secondary test using a different orchestrator, results showed Codex remained a Slytherin and improved its accuracy. Claude's personality shifted depending on the tool, switching from Hufflepuff to Ravenclaw, showcasing how the interface influences coding style. The aim of the comparison wasn’t to declare a superior agent but rather to highlight the distinct personalities and problem-solving approaches of each AI. The experiment provided insights into the diverse capabilities of these tools while keeping the tone light and fun.