From Tools to Teammates: Evaluating LLMs in Multi-Session Coding Interactions

Large Language Models (LLMs) are increasingly used in working environments for a wide range of tasks, excelling at solving individual problems in isolation. However, are they also able to effectively collaborate over longterm interactions? To investigate this, we introduce MEMORYCODE, a synthetic multisession dataset designed to test LLMs' ability to track and execute simple coding instructions amid irrelevant information, simulating a realistic setting. While all the models we tested handle isolated instructions well, even stateof-the-art models like GPT-4o and the reasoning model DeepSeek-R1 show degraded performance when instructions are spread across sessions. Our analysis suggests this is due to their failure to retrieve and integrate information over long instruction chains. Our results highlight a fundamental limitation of current LLMs, restricting their ability to collaborate effectively in long interactions.