Made in London by
Create a GPT chatbot for any GitHub repo in
30 seconds
Every major programming language supported:
...and more
1
Enter your GitHub repository URL:
We will crawl your repo and email you an index file you can upload directly to your GPT.
2
Upload your GPT codebase file:
You will receive an email with a link to a file repo.md. You can then upload that file to OpenAI to create your custom codebase GPT.
Frequently Asked Questions
How does this work?
CodebaseChat clones and crawls your repo to create a file that you can upload to OpenAI's new GPTs or Assistants. OpenAI GPTs and Assistants currently have a 20 file limit when using the retrieval features, so for now you need to upload a single indexed file to allow them to operate over large amounts of data.
How should I prompt my GPT or assistant to make full use of the file?
We recommend using a prompt that tells the GPT about the contents of your codebase.
For example, at Context.ai we use the following prompt in our GPT:
You are a code assistant for the Git repo of Context AI. Context AI is a startup that is building product analytics software for natural language interfaces where users are interacting with LLMs. The product ingests transcripts of its customers, then provides analytics about those interactions, such as topics, sentiment analysis, languages spoken and more. The contents of the codebase are stored in repo.md, with the contents of each file under a new Markdown header.
Can I modify the code or suggest changes to CodebaseChat?
Yes! CodebaseChat is open-source and we welcome contributions. You can find the repo here.
I don't have ChatGPT Plus, can I still use this?
Yes, you can instead upload the repo file in the OpenAI Assistants Playground and ask questions of your repo with the GPT API free tier.
Do you support private repos?
We don't yet support private repos. Contact us here if you'd like to be notified when we launch this.
I have another feature request, how can I contact you?
You can add your feature requests here.