SolAgent: A Specialized Multi-Agent Framework for Solidity Code Generation

TL;DR

SolAgent tackles the critical challenge of secure and correct Solidity code generation by introducing a specialized, tool-augmented multi-agent framework with a dual-loop refinement: an inner Forge-based correctness loop and an outer Slither-based security loop. The approach is complemented by file-system context tools and a workflow-distillation pathway that distills high-quality agent trajectories into smaller open-source models, enabling efficient deployment. Empirical evaluation on the SolEval+ Benchmark shows SolAgent achieving a Pass@1 of up to 64.39% and reducing security vulnerabilities by up to 39.77% relative to human-written baselines, outperforming vanilla LLMs and existing agent frameworks. The work demonstrates the practical impact of integrating domain-specific tooling into agent workflows and provides a viable distillation route to democratize secure smart-contract generation.

Abstract

Smart contracts are the backbone of the decentralized web, yet ensuring their functional correctness and security remains a critical challenge. While Large Language Models (LLMs) have shown promise in code generation, they often struggle with the rigorous requirements of smart contracts, frequently producing code that is buggy or vulnerable. To address this, we propose SolAgent, a novel tool-augmented multi-agent framework that mimics the workflow of human experts. SolAgent integrates a \textbf{dual-loop refinement mechanism}: an inner loop using the \textit{Forge} compiler to ensure functional correctness, and an outer loop leveraging the \textit{Slither} static analyzer to eliminate security vulnerabilities. Additionally, the agent is equipped with file system capabilities to resolve complex project dependencies. Experiments on the SolEval+ Benchmark, a rigorous suite derived from high-quality real-world projects, demonstrate that SolAgent achieves a Pass@1 rate of up to \textbf{64.39\%}, significantly outperforming state-of-the-art LLMs ($\sim$25\%), AI IDEs (e.g., GitHub Copilot), and existing agent frameworks. Moreover, it reduces security vulnerabilities by up to \textbf{39.77\%} compared to human-written baselines. Finally, we demonstrate that the high-quality trajectories generated by SolAgent can be used to distill smaller, open-source models, democratizing access to secure smart contract generation. We release our data and code at https://github.com/openpaperz/SolAgent.

Figures (4)

• Figure 1: Overview of SolAgent
• Figure 2: The SolEval+ Benchmark
• Figure 3: Dataset Token Distribution
• Figure 4: Example: Reentrancy fix in claimThrone by SolAgent versus a baseline.