*) Inspiration: “Every developer dreams of writing flawless code — but even the best can overlook hidden vulnerabilities. One small mistake, like a leaked API key or weak encryption, can bring down an entire system. We wanted to create something that doesn’t just scan for problems, but understands them. Something that thinks like a security expert, acts like a teammate, and never sleeps. That’s how CodeGuardian AI was born — an intelligent security companion that empowers developers to code confidently, knowing AI has their back.”
*) What it does: CodeGuardian AI automatically scans your code for security vulnerabilities using Semgrep, then uses AWS Bedrock Agents to explain the risks, suggest secure fixes, and even auto-patch them — all through a simple, interactive dashboard. It’s like having an AI-powered security engineer that reviews, reasons, and protects your code in real time.
*) How we built it: Using various programming language and most important using semgrep for security and AWS Bedrock Agents
*) Challenges we ran into: Integrating Semgrep’s static analysis with eigenlayer reasoning while maintaining data security and smooth API flow was our biggest challenge.
*) Accomplishments that we're proud of: We successfully built a fully functional AI security agent that detects, explains, and auto-fixes vulnerabilities — all while keeping developer data 100% secure.
*) What we learned: Nothing is impossible if you start doing it.
*) What's next for CodeGuardian AI – Your Secure Code Agent: 🚀 Integration with GitHub Actions for automated PR scans 🧠 Custom Semgrep rule generation using Bedrock reasoning 📈 Security analytics dashboard with historical vulnerability tracking 💬 Chat interface for AI security consultation
Built With
- css3
- typescript
- vite
Log in or sign up for Devpost to join the conversation.