Skip to content

SmeLLM - Automatic code smell detection and refactoring using LLMs.

Choose a tag to compare

@vedant130701 vedant130701 released this 10 Jul 16:29
· 12 commits to main since this release
3b152d6

SmeLLM was built with the intent to facilitate code smell detection in student code in coding based courses. The following are the features present in v2.0.0 -

Current Features

  • Code smell detection and report generation for individual files.
  • Support for OpenAI large language models (LLMs) as the analyzer.
  • Automated test suites benchmarking to evaluate detection capabilities.
  • Command line capabilities supporting various arguments.
  • Support to process, files, folders and zipfiles.

Upcoming featues

  • Support for more models from different LLM providers.
  • Support for custom fine-tuned models.
  • Expansive dataset covering more code smell sub-types based on various treatments.