The Self-Corrective Coding Agent is an automated tool designed to assist developers in writing code and fixing errors automatically. It utilizes a language model to generate code snippets and employs a self-corrective mechanism to address any errors encountered during code execution.
- Automated Code Generation: Generates code snippets based on user specifications and requirements.
- Error Detection and Correction: Identifies and resolves errors in the generated code using a self-corrective mechanism.
- Language Model Integration: Utilizes Mixtral language model for code generation and error correction.
- Conda Environment Integration: Executes generated code within a specified Conda environment for dependency management.
- Interactive Feedback: Provides interactive feedback to the user during code execution and error correction.
To use the Self-Corrective Coding Agent, follow these steps:
- Clone the repository to your local machine.
- Set up a Conda environment with the required dependencies.(groq for LLM inference )
- Run the
main.pyscript to start the self-corrective coding agent. - Provide input specifications or requirements for code generation.
- Review the generated code and any errors detected by the agent.
- Optionally, provide feedback to improve the agent's performance.
- Iterate as needed to refine the generated code and achieve the desired outcome.