A number of days in the past, I confirmed you an incredible new ChatGPT characteristic accessible to paying customers. Plus, Professional, and Workforce tier customers can now level Deep Analysis at a complete GitHub repo and get again evaluation experiences.
As I confirmed, this functionality helps velocity up the method of coming up to the mark on present codebases. You would possibly want to do that if you happen to purchase a product from one other developer or if you happen to’re introduced onto a venture and have to be taught the codebase rapidly. It is also good for reviewing your individual codebase and refreshing your self on how sections work — particularly if you happen to’ve moved on to different issues for some time and at the moment are coming again to the unique code.
I promised I might present you the best way to carry a codebase into GitHub particularly for evaluation by Deep Analysis. That is what we’re about to do on this article.
Transferring my code into GitHub
To display this, I am shifting My Non-public Web site into GitHub. My Non-public Web site is a freemium WordPress plugin with about 20,000 lively customers I have been engaged on for a couple of decade. WordPress, for historic causes, makes use of SVN as an alternative of GitHub as a code repository, so I have not actually had a have to put My Non-public Web site into GitHub.
However given the chance to carry out deep evaluation on it, I made a decision to set it up. I am going to undergo that course of with you right here.
Getting began with GitHub Desktop
Earlier than we begin, let’s make clear some issues. Git is a distributed model management system that runs on a programmer’s native laptop. GitHub is a cloud-based service that shops an infinite library of open-source and proprietary coding initiatives. These initiatives are moved into GitHub (the cloud service) utilizing Git (the software).
Actual programmers solely use Git on the command line, the place it is referred to as git. No actual programmer would dare to capitalize git. Actual programmers command git through a spread of choices, creating specialised command traces that do their bidding. Failure to make use of git on the command line will end in your real-programmer card being revoked by the Worldwide Society of Programmers Who Are Smarter Than You.
I’m apparently not an actual programmer. I’d as nicely get that out of the best way earlier than the feedback erupt in disdainful RPs (actual programmers) mocking my lack of command-line acuity. I do not use Git through the command line. I do not prefer it. I consider people left the cave way back and adopted graphical person interfaces as instruments of civilized society.
I, due to this fact, favor utilizing GitHub Desktop, which is a point-and-click model of Git for these undeserving of the title actual programmer. And sure, my official real-programmer card has been revoked. I can stay with it.
You may obtain GitHub Desktop right here.
As soon as you’ve got launched GitHub Desktop, both check in to your GitHub account or create one. I’ve lengthy had a GitHub account for different initiatives, so I simply signed in.
How one can create a GitHub repository
Subsequent, I created a repository within the GitHub cloud for my codebase. Right here it may be a bit complicated. Despite the fact that I did not have an present repo for My Non-public Web site, I selected “Add an Present Repository out of your Native Drive…” as a result of I used to be going to take that codebase and switch it right into a repo.
GitHub Desktop is definitely fairly good about this. As soon as it realizes there isn’t any GitHub information for the folder chosen, it offers you an error and give you the choice to create a repo. Click on the hyperlink highlighted by the inexperienced arrow proven under.
That may current the Create a New Repository dialog. Right here, I named my repo (all lowercase, with dashes between phrases), added a brief description, informed it the native path to the code on my laptop, and left the remainder as-is.
I did not have to play with the README, license, or ignore choices as a result of I am utilizing this repo for AI evaluation, not for supply management and collaboration.
It is right here I ought to be aware this text describes what it is advisable do to let your code be examined by ChatGPT Deep Analysis. That is undoubtedly not a complete how-to-set-up-GitHub article.
How one can transfer the codebase to GitHub
It is time to transfer your code as much as GitHub.
This is a cautionary be aware: In the event you’ve saved your code personal, importing it to GitHub is sending your code to a cloud service. GitHub presents each personal and public repositories, however you are technically giving Microsoft entry to your code. Microsoft owns GitHub.
Now, go forward and hit Publish.
At this level, you will have the chance to make your repo public or personal. Whenever you join ChatGPT to your repo, you will be passing alongside your entry rights, so you possibly can let ChatGPT study a non-public repository.
That mentioned, I bumped into some points with Deep Analysis accessing my code, and one of many issues ChatGPT requested me was whether or not my code was public. My tackle that’s: in case your code is personal and you’ve got all of your credentials and connector arrange (extra on that later), you possibly can most likely work on a non-public repo.
Since My Non-public Web site is open supply, I unchecked “Preserve this code personal.”
Taking a look at your new repository
If every thing labored, you will see a brand new choice: “View on GitHub.” Click on it.
That may carry you to your newly created GitHub repo. This is mine.
Now that your repo is up, be aware of its designation. Yow will discover that within the higher left nook of the GitHub display. For My Non-public Web site, it is davidgewirtz/my-private-site (with none areas).
How one can arrange the ChatGPT connection
Now it is time to change to ChatGPT. The subsequent two screenshots are the identical as what I confirmed you on this article on the characteristic. However to get to the subsequent configuration step, you will have to do what’s proven within the screenshots.
First, change your mannequin to o3 and kind within the immediate precisely as I did. You may most likely tweak this over time, however if in case you have the $20-per-month Plus tier, you are solely going to be allowed 10 queries into Deep Analysis per 30 days, so reducing and pasting is your buddy.
Subsequent, click on the little caret subsequent to Deep Analysis.
Now, create the hyperlink between your ChatGPT account and your GitHub account. Go forward (if you happen to dare) and provides Skynet — uh, I imply the AI — permission to entry your GitHub account options.
Subsequent, you will be requested which GitHub account ought to get the ChatGPT connector. I’ve two, so I acquired this alternative display. You would possibly skip this display if you happen to solely have one account.
Now it is time for extra permissions. This time, you are giving permission to entry both all of your account’s repos or only one. I chosen solely the my-private-site repo.
And now, theoretically, Deep Analysis in ChatGPT will likely be linked to your repo. Theoretically. In apply, mine required one other step.
What to do if ChatGPT cannot discover your repo
GitHub indexes repositories, and if ChatGPT would not present your repo as accessible, it most likely means GitHub hasn’t listed your new repository but. That is what occurred right here.
I ought to have been capable of choose or kind in my full repo identify (bear in mind, davidgewirtz/my-private-site), however ChatGPT wasn’t capable of find it.
To repair this, return to your GitHub account and kind within the command string proven on the prime of this screenshot. Clearly, change the textual content in blue to match your repo identify.
The command is mainly repo:(repo adopted by a colon), adopted by the total identify of your repo, adopted by an area and the phrase import. This may inform GitHub you’d prefer it to index your repository.
As you possibly can see, GitHub confirmed it was now indexing my repository.
I brewed myself a well-deserved cup of espresso as a method to give GitHub time to index my repo. As soon as I completed the final drop, I went again to ChatGPT, dropped down the Deep Analysis menu, and located my newly created repository.
Have enjoyable with Deep Analysis
You are prepared to start out utilizing Deep Analysis in your repo. For an in depth information on how that labored for my repo, level your self to my earlier article on the subject.
Have enjoyable. I used to be fairly blown away. You may be, as nicely.
Have you ever tried utilizing ChatGPT Deep Analysis with your individual code but? What was your expertise connecting a GitHub repo? Did you run into any indexing points or permission snags alongside the best way? Do you like utilizing GitHub Desktop or the command line when establishing your repositories? Tell us within the feedback under.
You may observe my day-to-day venture updates on social media. Make sure you subscribe to my weekly replace e-newsletter, and observe me on Twitter/X at @DavidGewirtz, on Fb at Fb.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.
Get the morning’s prime tales in your inbox every day with our Tech As we speak e-newsletter.