By Girish Gupta — 13 Mar 2025

GitHub Contributor Analysis, code via LLM

A reporter got a tip about a small tech company working close to a government department into which their team was looking, and wanted to find out who was working there—beyond what was available on LinkedIn.

The company had an active GitHub account, and some public repositories. That could be a good source of gaining more information on at least their engineers, I thought.

I knew GitHub had an API through which I could pull a list of a repository's contributors and details about them. Writing a script to do this wasn't going to be particularly difficult, perhaps an hour or two's work, but Claude, or another LLM-based product, would be able to do it much faster.

So I instructed Claude with an incredibly simple prompt:

I want Python code to look at a GitHub repo (public) using the API, find all the contributors, get their names, links to social media, etc., and order them by most recent commits. I'd like this saved as a CSV.

Its output was fantastic. A couple of minutes of finessing gave me methods I could call to get a list of contributors to a code repository, as well as biographical information and links to their online profiles:

REPO_URL = "https://github.com/openai/whisper"

class GitHubContributorAnalyzer:
    """
    A class to analyze contributors of a GitHub repository,
    including fetching their names, social media links, and sorting by most recent commits.
    """
   
...

REPO_OWNER, REPO_NAME = extract_repo_info(REPO_URL)

# Optional: Add your GitHub token here for higher rate limits
GITHUB_TOKEN = os.getenv("GITHUB_TOKEN")

# Create analyzer and run analysis
analyzer = GitHubContributorAnalyzer(GITHUB_TOKEN)
contributors = analyzer.analyze_repository(REPO_OWNER, REPO_NAME)

Here's the full, locally-runable code

I entered the organization's GitHub URL into REPO_URL and got my list as a spreadsheet: (This is for OpenAI's Whisper repository; OpenAI is not the organization into which the reporter was looking!)

Username	Name	Email	Bio	Company	Location	Profile URL	Contributions	Latest Commit Date	Social Website	Social GitHub
jongwook	Jong Wook Kim	jongwook@nyu.edu		OpenAI	San Francisco, CA	GitHub	69	2025-01-04T20:56:16Z	Website	GitHub
cclauss	Christian Clauss	cclauss@me.com	Working hard to find and fix bugs in software...	Christian Clauss	Switzerland	GitHub	3	2025-01-04T09:38:35Z		GitHub
Purfview			Pro observer of the earthlings... Contact: purfview@protonmail.com		Aldebaran	GitHub	1	2024-12-01T05:47:01Z		GitHub
lvaughn	Lowell Vaughn	lowell@vaughnresearch.com				GitHub	1	2024-11-26T17:37:01Z		GitHub
YuZekai	YuZekai	yuzekai@hdu.edu.cn		Hangzhou Dianzi University	Hangzhou, Zhejiang	GitHub	1	2024-11-13T00:35:54Z		GitHub
BotMaster3000	BotMaster3000					GitHub	1	2024-11-04T07:00:30Z		GitHub
kittsil						GitHub	1	2024-10-26T14:17:31Z		GitHub
xingjianan	Jianan Xing					GitHub	1	2024-09-10T16:53:08Z		GitHub
ryanheise					Sydney, Australia	GitHub	4	2023-12-18T20:11:16Z		GitHub

The whole thing took just a few minutes and gave the reporter's team a list of all contributors to the target company's top repository, as well as social and contact details.

Going further, an LLM could be used to parse blogposts, social content, etc., to find stories, draw knowledge graphs of links between people, etc.

To do this yourself, first clone (try GitHub Desktop) my repository, open main.py and enter the URL of the GitHub repository about which you're interested in learning in REPO_URL, as well as a GitHub API token in your .env, and then run uv run main.py in your Terminal.

Subscribe to AI for Investigation