SeedAIchemy: LLM-Driven Seed Corpus Generation for Fuzzing

TL;DR

SeedAIchemy leverages large language models to automate the creation of high-quality seed corpora for fuzzing, improving effectiveness over naive methods and matching manual curation.

Contribution

This paper presents SeedAIchemy, a novel LLM-driven tool that automates seed corpus generation for fuzzing, enhancing efficiency and quality compared to traditional approaches.

Findings

SeedAIchemy-generated corpora outperform naive corpora in fuzzing effectiveness.

Corpora produced by SeedAIchemy are comparable to manually curated ones.

The tool utilizes multiple modules with LLM workflows to optimize search term construction.

Abstract

We introduce SeedAIchemy, an automated LLM-driven corpus generation tool that makes it easier for developers to implement fuzzing effectively. SeedAIchemy consists of five modules which implement different approaches at collecting publicly available files from the internet. Four of the five modules use large language model (LLM) workflows to construct search terms designed to maximize corpus quality. Corpora generated by SeedAIchemy perform significantly better than a naive corpus and similarly to a manually-curated corpus on a diverse range of target programs and libraries.