What I Built
This one is for me. I'm a third-year student making my first
open-source contributions, and the hardest part isn't the code, it's
understanding an issue well enough to start.
Issue Explainer: paste a GitHub issue link and get a summary,
type, difficulty (based on my own skills), skills needed, where to
look in the repo, first 3 steps, and one question to ask the
maintainer. It also warns me if the model names a file that doesn't
exist in the repo.
I tested it on a real issue I wanted to understand: https://github.com/agentrhq/webcmd/issues/534.
Demo
It runs locally, so there is no deployed link. Setup steps are in the
README.
Code
https://github.com/Shree-Spirit/01_Hacktoberfest-2026
How I Built It
-
Model: Gemma (
gemma3:4b) running locally through Ollama -
Also compared:
qwen2.5:3bandqwen2.5:7b(dropdown in the UI) - App: Python, GitHub REST API, Streamlit UI plus a CLI
- Safety check: the app compares the files the model suggests against the repo's real file tree
One function, explain(), does the work. The CLI and the UI both
call it.
Why Does Open Innovation Matter?
- Cost: no API bill, so I can analyze as many issues as I like.
- Swappable: I changed models from a dropdown and compared them on the same issue: [your times and one quality note].
- Local: the model step ran with Wi-Fi off: [yes or no, after you test it].
- Honest limits: my first run took 332 seconds on a laptop, and "Where to look" sometimes points to docs instead of code. [Your fix, if any.]
A closed API would be faster, but I'd pay per request and couldn't
compare models this easily.
Top comments (1)
This is for first week,
More to come :)