Guardian Angel AI

An autonomous cybersecurity agent detecting leaked repository credentials and executing instant post-leak recovery workflows powered by Gemini.

Comment

💡 Inspiration In modern DevOps workflows, a single hardcoded credential leaked into a public repository can compromise an entire enterprise infrastructure within seconds. While passive scanners alert developers about leaks, they leave a dangerous "response gap" where the token remains active and vulnerable. We were inspired to build Guardian Angel AI—not just to scan for leaks, but to act as an autonomous, real-time cyber first responder that bridges the gap between detection and remediation instantly.

🛠️ How We Built It Guardian Angel AI is built entirely with a robust, cloud-ready architecture designed for rapid scaling:

  • Core Logic: Built using Python to simulate a repository environment.
  • AI Engine: Integrated with the Gemini API via advanced prompting to act as an autonomous incident responder that instantly evaluates the context of leaked tokens.
  • User Interface: Developed a sleek, interactive frontend using Streamlit, allowing users to paste code snippets, enter an API key, and instantly see a simulated response work flow

🚧 Challenges We Faced Building a complex security agent completely from a mobile environment presented massive workflow hurdles. We encountered clipboard syncing bugs, network formatting constraints, and unexpected API traffic limits (such as HTTP 503 backend rate-limiting). Overcoming these issues forced us to optimize our exception handling within the Streamlit codebase, ensuring that even when the AI brain faces extreme demand, the local application fails gracefully and remains secure.

🎓 What We Learned This project deepened our understanding of building self-healing software systems and using Large Language Models for deterministic security workflows. We learned how to structure defensive code against API bottlenecks and, more importantly, proved that a fully functioning, interactive cybersecurity application can be successfully engineered, tested, and deployed entirely using mobile layouts under tight, high-pressure constraints.

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for Guardian Angel AI

Built With

Share this project:

Updates