DEV Community

Cover image for GitHub Contribution Agent | Agent.ai challenge
programORdie
programORdie

Posted on

GitHub Contribution Agent | Agent.ai challenge

This is a submission for the Agent.ai Challenge: Full-Stack Agent (See Details)

What I Built

I built a GitHub Contributions Finder using agent.ai. This Agent helps developers quickly analyze GitHub profiles to identify their most notable contributions, projects, and areas of expertise.

I built this Agent to save time when reviewing profiles for collaboration, hiring, or personal inspiration. It's designed for developers, recruiters, and open-source enthusiasts who want to explore GitHub activity without manually digging through repositories.


Demo

GitHub Contributions Finder Agent

Here’s a quick video showing the Agent in action:


How it Works

  1. User Input: The user specifies preferences like programming language, skill level, and time availability.
  2. GitHub API Integration: The Agent fetches relevant GitHub issues based on the selected language.
  3. Issue Filtering: Using Python, issues are filtered based on the user’s skill level and available time.
  4. LLM Ranking: A language model ranks the filtered issues based on relevance and importance.
  5. Results Presentation: The Agent presents a ranked list of issues with details, including:
    • Issue specifics.
    • Repository information.
    • Why the issue is a good match for the user.

How it works flowchart


Agent.ai Experience

Building with agent.ai was both exciting and educational. The platform made it easy to handle complex tasks like querying the GitHub API and processing the data.

Initially, I faced some challenges with radio inputs not saving correctly. Thankfully, the agent.ai team resolved the issue quickly after I reached out. Another hurdle was figuring out how to run Python code within the Agent. It turns out you just need to deploy an AWS Lambda function—which the platform handles for you completely free!

The most delightful moment was watching the Agent deliver results in seconds, but the learning curve of fine-tuning prompts and understanding workflows made the process even more rewarding. Overall, it was a smooth and enjoyable experience.

Top comments (2)

Collapse
 
taj_malnas profile image
Taj malnas

great idea i would like to connect with you

Collapse
 
programordie profile image
programORdie

Thanks!