Code Together

Formerly known as DevGPT, rebranded due to OpenAI policy changes.

Latest Updates

• New ML-Focused Version: Discover our enhanced version for building machine learning projects and pipelines here.
• OpenWebUI Prompt: Access the OpenWebUI prompt here, compatible with models like Stable-Code-3B, Llama-3-8B, WizardLM2-7B, Mistral, GPT3.5, and GPT4o.

Try It Now

How to Use

Follow this guide to get started with Code Together:

1. Initialize a Project: Use the /project command followed by the project name and description.
   Example:

   /project MyApp "My application is a simple chatbot using the GPT2 Italian model"

2. Complete Your Files: Review the project structure and use /complete filename to generate the necessary files.
3. Finish and Refine: Use additional commands like /improve, /logs, /debug, or /dim to enhance and finalize your project.

Additional Tips

• Improve a File: Type /improve followed by the file content to enhance a specific file.
• Add Logging: Use /logs to insert logging and debugging snippets.
• Resource Estimation: Estimate compute resources required with /dim.
• Generate a README: Create a project README with /repo.

...and much more! Code Together is continuously evolving to meet all your development needs.

Full Command List

Command	Description	Example Use
/project	Initialize a new project with a name and description.	/project MyApp "My application is a simple chatbot"
/list	List all needed files for the project in a Linux environment.	/list
/setup	Propose main and requirements.txt files with full code.	/setup
/env	List required imports, libraries, and environment setup steps (Linux only).	/env
/dim	Calculate the minimum RAM and HDD space required to run the project.	/dim
/split	Modularize a single code snippet or file into multiple files.	/split (after providing a large code block)
/log	Add logging to the existing code.	/log (after selecting or providing code)
/docs	Explain the functions and libraries used in the latest code snippet.	/docs (after providing code)
/debug	Add debugging code to troubleshoot issues.	/debug (after selecting or providing code)
/exec	Simulate running the code and display the output.	/exec (after setting up the required files and env.)
/improve	Optimize and enhance the existing code.	/improve (after selecting or providing code)
/complete	Generate a complete script for a specific file.	/complete my_file.py
/order	Provide the order of file execution to test the project.	/order
/adapt	Update all existing files to accommodate recent changes.	/adapt (after modifying one file)
/use	Explain how to run, use, or test the project.	/use
/repo	Generate a README.md file for GitHub with all necessary project information.	/repo
/fix	Attempt to make the script work using an alternative approach.	/fix (when existing code doesn't work)
/thanks	Encourage users to support the project by starring the GitHub repository.	/thanks

Example Uses

Share your creations! I love to feature them in the "Wall of Fame" 😊

Starter Apps:

• Create a Simple Blockchain
• Manage DNS with DNSControl and GitHub Actions
• Fine-Tune GPT2 LLM Model
• ScamTracker: Pattern Matching in the Stock Market (PoC)
• HTML Bootstrap Skeleton for Beginners
• AI-Enhanced Real Estate Valuation (AIREV)

Built in Minutes with DevGPT:

• Ethical AI Automated Assessment using GitHub Actions and OpenAI API

Contributing

Contributions are welcome! Please read the contributing guidelines for more details on how to get involved.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contact

For questions, issues, or feedback, please open an issue.