Introduction to the SuiAudit Project

SuiAudit is an AI-driven audit tool focused on the Sui blockchain ecosystem, aimed at enhancing smart contract security and code quality through advanced machine learning models. As an emerging open-source project, it combines artificial intelligence with blockchain security auditing to help developers detect vulnerabilities and optimize Move language code. The project’s official website is suiaudit.wal.app, and it has released some infrastructure on GitHub (such as SuiAudit Raw1) to facilitate community contributions and reproducibility. Additionally, the project shares model demos on Kaggle and has published related research papers on TechRxiv.

What We Are Doing

SuiAudit is currently focused on developing and deploying AI audit solutions specifically for Sui Move code. The core product includes SuiAudit Pure1, a small language model (LLM) fine-tuned with extremely low resources, dedicated to smart contract auditing. It can automatically detect vulnerabilities, assess code risks, and generate audit reports. The project integrates the Walrus protocol for decentralized storage and runs POCs (such as Sui Wallet Cleanup) in the Sui ecosystem, allowing users to upload code for real-time analysis. The team also participates in hackathons, such as the 2025 Walrus Haulout Hackathon, demonstrating how AI models can quickly generate risk assessment reports. Through Kaggle notebooks, SuiAudit publicly shares experimental code, enabling developers to compare its performance with other LLMs. Overall, SuiAudit is building an offline-running, privacy-focused audit tool that covers everything from code scanning to comprehensive ecosystem auditing.

Why We Are Doing This

The Sui blockchain is renowned for its high throughput and parallel execution, but the complexity of the Move language makes smart contracts prone to vulnerabilities, leading to potential financial losses and security risks. Traditional auditing relies on human experts, which is costly, time-consuming, and unable to keep up with the rapid iteration of Web3 development. SuiAudit was created to address these pain points: by using AI to achieve automated, cost-effective auditing, especially optimized for small models that can run in resource-constrained environments without relying on cloud services, thereby protecting developer privacy. The team’s mission is to provide qualified AI audit services for the Sui ecosystem, lower the barriers to hacker attacks, and promote the standardization of security for decentralized applications. As an open-source-oriented project, it also aims to fill market gaps, allowing small and medium-sized developers to easily access professional-level auditing, rather than being limited to large enterprises.

Expected Results or Impacts

In the short term, we expect SuiAudit Pure1 to continue outperforming large LLMs like Claude Opus in benchmark tests, achieving a significant 31.4% improvement in accuracy. This will lead to more reliable vulnerability detection, reducing security incidents in the Sui ecosystem and saving audit costs for developers. In terms of impact, the project may become a standard tool in the Sui community, integrated into development workflows to enhance overall ecosystem security. Through hackathons and paper publications, SuiAudit has already attracted attention and is expected to promote the widespread adoption of AI in blockchain auditing, with potential impacts including reducing risks in DeFi protocols, accelerating project launches, and establishing Sui as a secure platform in the global Web3 community. In the long term, this could drive a shift in industry standards from manual auditing to AI-assisted transformations, benefiting thousands of developers and users.

Expected Future Developments

In the future, SuiAudit plans to expand to full blockchain network auditing, integrating more AI technologies such as reinforcement learning (Reinforcement Learning), and developing dedicated LLMs to handle complex ecosystems. The team will compile high-quality audit datasets, support hybrid modes of manual and automated auditing, and open free online services to attract users. Potential developments include collaborating with the Sui Foundation to integrate into official toolchains; expanding to other chains like Ethereum or Aptos; and launching paid advanced features, such as real-time monitoring and custom rules. Through GitHub and Kaggle openness, the community can contribute to model improvements, evolving into a crowdsourced audit platform. Ultimately, SuiAudit aims to become a global leader in AI blockchain auditing, driving Web3 toward a more secure and intelligent direction.

SuiAudit is not just a tool, but a catalyst for advancing security in the Sui ecosystem—welcome community participation!