CoPrompt 2024
Writing prompts with a team can mean a lot of copying, checking, and asking what changed. We built CoPrompt to make that back-and-forth easier when coding with AI. You can see what your teammates are working on, build on their prompts, and link them to your own so you don't have to copy every update by hand. And when you get stuck, you can leave a request for help right alongside the prompt.
Recognition & Outreach
- Accepted to and presented at CHI 2024. Read the paper .
Core Features
- Share a prompt or build on one a teammate has already written.
- See how prompts fit together, then open one for a closer look.
- Link prompts so changes stay in sync.
- Ask for help where you need it, whether your teammate is working alongside you or checking in later.
Links
Read the CoPrompt paper
View on ACM Digital Library.
Skills
- Python
- PyTorch
- scikit-learn
- Machine Learning
- Evasion Attacks
- D3.js
- JavaScript
- OpenAI API
Keywords
- HCI
- Collaborative Programming
- Natural Language Programming
- Prompt Engineering
- Large Language Models (LLMs)
Authors
Li Feng, Ryan Yen, Yuzhe You, Mingming Fan, Jian Zhao, Zhicong Lu