Vision and what it does
I constantly want to discover cool new open-source projects, but I don't always have the time to manually browse GitHub's trending pages every day. To solve this, I built the Daily GitHub Explorer.
Instead of me opening an app and manually clicking "Generate", this always-on AI agent automatically wakes up every morning, finds a trending repository from the past 7 days, uses AI to summarize why it's cool, and publishes a fresh HTML dashboard.
To make it even more "agentic", I integrated the AI directly into Telegram! Not only does it update the dashboard, but the agent proactively sends a broadcast message to my Telegram group every morning so I don't even have to open a browser to get my daily tech news.
How I built it
I designed this to run entirely autonomously on AWS Free Tier services using the Serverless Application Model (SAM). The core logic is written in Python.
My two favorite parts of this build:
- The UI Design: I wanted it to look like a modern, premium news app (like Medium or Apple News) rather than a basic text dump. The agent dynamically generates and injects beautiful CSS styling directly into the HTML payload before uploading it to S3!
- The Messaging Integration: Using the Telegram Bot API natively within Python to instantly push notifications.
AWS services used and architecture overview
- Amazon EventBridge: Acts as the brain's clock, triggering a cron schedule every morning at 6:00 AM UTC.
- AWS Lambda: The compute engine that runs the Python script to fetch the GitHub API, generate the UI, and ping the Telegram Bot API.
- Amazon Bedrock: Used to invoke the
amazon.nova-lite-v1:0model to generate the fun, enthusiastic summary. - Amazon S3: Hosted as a static website to serve the dynamically generated HTML dashboard to the public.
What I learned
I learned how to use the Bedrock Converse API to seamlessly integrate generative AI into an autonomous workflow. I also learned how to use Lambda to write directly to a public S3 website bucket without any human intervention, and how easy it is to hook up messaging platforms like Telegram to an AWS backend.
Project Links
Live Demo: http://daily-github-explorer-websitebucket-manb48k0shdu.s3-website-us-east-1.amazonaws.com
Source Code: https://github.com/rasik2002/weekend-creative-agent-challenge
Source Code: https://github.com/rasik2002/weekend-creative-agent-challenge
(Screenshots of the UI and Telegram bot attached!)
Enjoyed reading this content? Let the author know!
Your likes, comments, shares, and saves help creators reach more builders.
Loading recommendations
Loading article