AWS Builder Center
Weekend Creative Agent Challenge: The Morning Micro-Comic Agent

Weekend Creative Agent Challenge: The Morning Micro-Comic Agent

An autonomous, serverless application that utilizes AWS Lambda, EventBridge, and Generative AI to generate and publish a daily satirical tech comic. It operates entirely on its own to bring a morning dose of humor and wisdom to developers.

Community Builder || Full Stack Engineer || BSCS’28 @ VU

Vision & What It Does

Working in software engineering is very busy. We always have so many things to do, like writing code, fixing bugs, and deploying apps. Sometimes, mornings can feel very stressful. So, I thought, why not make something that makes developers smile in the morning?
My idea is "The Morning Micro-Comic Agent". It is a completely automatic agent. Every single day at 7:00 AM, it wakes up on its own. It does not need me to click any button to start. It thinks of a funny story about coding life, writes a 3-panel comic script, and gives a good developer quote for daily motivation. Then, it saves this as an HTML webpage on the internet. It is the perfect tool because I do not have to do anything. It just runs in the background and gives me a fun comic to read before I start my hard work.

How You Built It

To build this project, I used Amazon Web Services (AWS). First, I created an Amazon S3 bucket. I turned on the Static Website Hosting feature so anyone on the internet can visit the link and see my comic. I also changed the bucket policy to make it public.
Next, I needed a place to run my code. I used AWS Lambda with Python 3.14. Inside my Lambda function, I wrote a simple script. I used the standard  urllib  library in Python to connect to the Gemini 3.6 Flash AI model using an API key. I did this because it is easy and does not need extra heavy zip files in Lambda. I gave the AI a very clear prompt. I told it to write the comic in pure HTML format, not Markdown. After the AI gives the HTML response, my Lambda code uses  boto3  to upload a file named  index.html  to my S3 bucket. It does this every day, replacing the old file with the new comic.

AWS Services Used / Architecture Overview

I used these AWS Free Tier services to make a complete serverless architecture:
  • Amazon EventBridge: This is like an alarm clock. I set a cron schedule so it triggers my code every morning at 7:00 AM. This makes the agent "always-on".
  • AWS Lambda: This is where my Python code lives. It is the brain. It calls the AI model, gets the creative text, and makes the HTML file.
  • Amazon S3: This is used to store my HTML file and show it as a live website to the world.
  • AWS IAM: This gives secure permissions. I made a role so Lambda can only upload files to my specific S3 bucket and write logs to CloudWatch.

Challenges Encountered & Overcome

When I was building this, I faced some big problems. First, I wanted to use Amazon Bedrock for the AI models. But when I checked my account, I had zero quota (TPM and RPM) for the models I needed. I opened a support ticket, but I didn't want to wait and miss the deadline. So, I quickly changed my plan and used the Gemini API instead. This kept my project moving fast.
Another problem was the text format. At first, the AI gave me Markdown text. When S3 showed it on the website, it looked very bad with lots of stars and hash symbols. I fixed this by writing a better prompt to force the AI to only output clean HTML tags. Also, my browser was saving the old page in the cache, so I had to learn to do a hard refresh to see my new code working.

What You Learned

I learned a lot of new things in this weekend challenge. I learned that combining EventBridge and Lambda is a very powerful way to make automatic jobs. It feels like magic when the code runs on its own schedule. I also learned that writing prompts for AI is a real skill. You have to be very strict with the AI if you want a specific output like HTML instead of Markdown. Finally, I learned how to use standard Python libraries to make API calls in Lambda. This makes the code simple and easy to deploy without needing to build custom layers. It was a great learning experience.

Link to App or Repo

Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article