CodeScene ACE is an AI-powered service from CodeScene, designed to assist with code refactoring without storing any of your source code. For more details on our commitment to privacy, please review our Privacy Principles here.
Additionally, it is important to mention that we don't train AI models based on your code. You can verify that also in the Privacy Principles mentioned above.
Currently, CodeScene ACE supports the following programming languages:
- JavaScript
- TypeScript
- React (.jsx & .tsx)
- Java
CodeScene ACE supports the identification and refactoring of the following code smells:
- Complex Conditional
- Bumpy Road Ahead
- Complex Method
- Deep, Nested Complexity
- Large Method
This list highlights the key areas where ACE can help improve code quality.
We use a combination of LLMs from various vendors, selecting the most suitable ones based on the type of code smell and programming language to deliver the best results. Currently, our LLMs include models from OpenAI, Anthropic, and Google.
When using CodeScene ACE, you may encounter the terms 'Quick Inspection' and 'Detailed Inspection'. Here’s what they mean:
Quick Inspection - The auto-refactored suggestion has passed semantic and syntactic validation, so we are highly confident that the change can be applied without further changes.
Detailed Inspection - The auto-refactored suggestion does have some deviation in semantic or syntactic validation, so we are less confident that it may be applied without some further alteration. It may still be valid and/or serve as an inspiration to the user.
If the 'Quick Inspection' is recommended, you can trust the refactoring suggestions provided. We also make sure that the refactoring suggestion indeed improves code health.
A 'Detailed Inspection' may require some refinement, but it typically provides a strong starting point for improvement. If the AI generates a result that doesn't meet our validation criteria, it will not be suggested to the developer.
CodeScene ACE is free to use under our freemium plan, which includes limited usage. A credit cap applies to the number of lines of code (LoCs) that can be refactored within a specific timeframe. Once the limit is reached, credits are automatically replenished after a few days.
If you're on the freemium plan, simply wait for your credits to refresh. Future updates will offer expanded features and the option to purchase unlimited refactoring credits