Red Flag Detector - Group #11

The AI solution for people who love overthinking. This Chrome Extension will scan your chats for "toxic" patterns that definitely aren't there, giving you the digital confirmation bias you deserve.

Comment

Inspiration

We realized that the biggest problem facing humanity today isn't climate change or the economy, but the fact that our friends might actually be mad at us when they send a 'thumbs up' emoji or use proper punctuation. We decided to stop trusting our intuition and start trusting a cold, unfeeling AI response to reveal the truth behind toxic behavior. Why have peace of mind when you can have a Chrome Extension that validates your deepest insecurities?

What it does

This Chrome Extension informs users on what exactly red, beige, and green flags are, before completely abusing the definition of a red flag. Users can choose to scan their current screen or input a text/file for the AI to brutally analyze and judge whether the user should swipe left or right, like a dating profile.

How we built it

We built our project through Visual Studio Code with HTML/CSS for the visuals, JavaScript to get the technical aspects working like the buttons and flashcards, and Google AI Studio for the AI implementation.

Challenges we ran into

Implementing the Gemini API was the largest challenge for us, given we have little experience with the implementation of AI Agents. Throughout the process, we learned the importance of keeping our personalized API Key secure and hidden, and how to effectively prompt the AI to respond with snarky, yet brief output.

Accomplishments that we're proud of

We are proud that we were able to create a working web extension and properly implement AI to read copied text, uploaded files, and scan your screen. We’ve never made a project like this before, so it was a great learning experience. We were able to take an API key and integrate it securely through a backend, which taught us about real world development practices and protecting user data. We’re also proud of how quickly we learned new tools and adapted when things broke. Each challenge pushed us to understand how all the moving parts fit together.

What we learned

We learned how to create a working web extension and utilize Gemini’s AI to be snarky and dramatic when analyzing the user’s text for red flags.

What's next for Red Flag Detector

While the current version of Red Flag Detector is excellent for destroying your self-esteem in real-time, we have ambitious plans to further disrupt the social fabric of the internet. Our primary goal is to implement the “Delusion Enabler” toggle, which will be a feature that does the exact opposite of our current AI by feeding into your delusions and validating every objectively bad decision you make. We also plan to introduce ”Audio Gaslighting”, which will use AI voice-overs (ideally in the tone of a “disappointed parent” or “sassy best friend”) to read your detected red flags back to you with maximum judgment. We hope to build a full ecosystem of digital chaos beyond this point.

Share this project:

Updates