As project teams become more aware of security, the demand for security audits before deploying contracts on-chain has also increased. Consequently, audit platforms and related services have become more diverse. However, existing static analysis tools have limitations in detecting certain common vulnerabilities, often leading developers to discover issues only after receiving audit reports. This extends the remediation timeline and increases costs. Additionally, such vulnerabilities are frequently classified as medium to high risk in audit competitions or security platforms, potentially resulting in financial losses, further underscoring the need to enhance detection capabilities.
This session will explore several types of these vulnerabilities and introduce a corresponding Prompt-based vulnerability detection process, demonstrating how AI can be leveraged to enhance security audits. Using the slippage minAmountOut vulnerability as an example, we will showcase the effectiveness of AI-assisted detection, ultimately optimizing security lifecycle management—from development and auditing to continuous monitoring.