LLM Code Smells: A Taxonomy and Detection Approach

📰 ArXiv cs.AI

Learn to identify and detect LLM code smells to improve software system quality

intermediate Published 25 May 2026

Action Steps

Identify potential LLM code smells in your source code using the proposed taxonomy
Analyze your code for inadequate LLM integration practices
Apply detection approaches to mitigate LLM code smells
Refactor your code to improve software system quality
Test and validate your refactored code to ensure LLM integration is proper

Who Needs to Know This

Developers and software engineers integrating LLMs into their systems can benefit from this knowledge to ensure high-quality code

Key Insight

💡 Poor LLM integration can undermine software system quality, but a taxonomy and detection approach can help

Key Takeaways

Learn to identify and detect LLM code smells to improve software system quality

Full Article

Title: LLM Code Smells: A Taxonomy and Detection Approach

Abstract:
arXiv:2605.22976v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly integrated into software systems for diverse purposes, due to their versatility, flexibility, and ability to simulate human reasoning to some extent. However, poor integration of LLM inference in source code can undermine software system quality. Therefore, inadequate LLM integration coding practices must be documented to help developers mitigate such issues. Following our earlier work on LLM code sme

Read full paper → ← Back to Reads