AI hunter

Inspiration

Smart contract vulnerabilities cause huge financial losses, but writing security-focused tests is time-consuming and requires deep expertise. I wanted to create a tool that helps developers quickly understand and reproduce potential issues in their contracts.

What it does

AI Hunter takes a smart contract project and a vulnerability description, then automatically generates a runnable unit test that attempts to reproduce the issue. It saves developers time, improves code security, and makes auditing more accessible.

How we built it

We combined AI reasoning with automated code analysis and a secure sandbox environment. The AI analyzes the project, gathers the relevant context, and creates a test file that developers can immediately run.

Challenges we ran into

Managing different smart contract structures

Ensuring the AI focuses on the right parts of the code

Getting generated tests to run reliably

Handling compilation, setup, and environment issues

Accomplishments that we're proud of

Built an end-to-end workflow that generates real, executable tests

Successfully reproduced vulnerabilities in sample projects

Created a tool that reduces hours of manual auditing work

Designed a simple interface that hides all technical complexity from users

What we learned

Small, well-designed context helps AI perform much better

Security testing benefits greatly from automation

Developers prefer tools that remove friction and “just work”

What's next for AI hunter

Support for more vulnerability types

A simple web UI

Automated reports and suggested fixes

Integration with more developer tools

Built With

• daytona