Kaikai Zhang

Kaikai Zhang

Auto-generated profile, based on publicly available information

Kaikai Zhang

Hong Kong University of Science and Technology, Department of Computer Science and Engineering, Hong Kong, Hong Kong

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
11 September 2024
Verified
Conference Paper
Open access
HFHaoran Fang
Haoran Fang
KZKaikai Zhang
Kaikai Zhang
DYDonghui Yu
Donghui Yu
YZYuanyuan Zhang
Yuanyuan Zhang
Coverage-Guided Fuzzing (CGF) has become the most popular and effective method for vulnerability detection. It is usually designed as an automated “black-box” tool. Security auditors start it and then just wait for the results. However, after a period of testing, CGF struggles to find new coverage gradually, thus making it inefficient. It is difficult for users to explain reasons that prevent fuzzing from making further progress and to determine whether the existing coverage is sufficient. In addition, there is no way to interact and direct the fuzzing process. In this paper, we design the dynamic directed greybox fuzzing (DDGF) to facilitate collaboration between the user and fuzzer. By leveraging Ball-Larus path profiling algorithm, we propose two new techniques: dynamic introspection and dynamic direction. Dynamic introspection reveals the significant imbalance in the distribution of path frequency through encoding and decoding. Based on the insight from introspection, users can dynamically direct the fuzzer to focus testing on the selected paths in real time. We implement DDGF based on AFL++. Experiments on Magma show that DDGF is effective in helping the fuzzer to reproduce vulnerabilities faster, with up to 100x speedup and only 13% performance overhead. DDGF shows the great potential of human-in-the-loop for fuzzing.

10 April 2026 in Proceedings of the ACM on Programming Languages
Verified
Journal Article
Open access
YXYuchong Xie
Yuchong Xie
KZKaikai Zhang
Kaikai Zhang
YLYu Liu
Yu Liu
RYRundong Yang
Rundong Yang
PCPing Chen
Ping Chen
DSDongdong She
Dongdong She
Seed explosion is a fundamental problem in fuzzing seed scheduling. It occurs when a fuzzer maintains a corpus with a huge number of seeds and fails to choose a promising one. Existing seed scheduling works focus on seed prioritization but suffer from the seed explosion since the seed corpus size is still huge. We tackle seed explosion from a new perspective, corpus reduction, i.e., compute a seed corpus subset. Corpus reduction can eliminate redundant seeds in the corpus and significantly reduce corpus size. However, this could lead to poor diversity in seed selection and severely impact the fuzzing performance. Meanwhile, effective corpus reduction incurs large runtime overhead. In practice, it’s challenging to adopt corpus reduction in fuzzing seed scheduling. Prior techniques like cull_queue, AFLCmin and MinSet all suffer from poor seed diversity. AFL-Cmin and MinSet incur prohibitive runtime overhead and are hence only applicable to one-time task initial seed selection rather than high-frequency seed scheduling. We propose a novel randomized corpus reduction technique, RandSet, that can reduce the corpus size and yield diverse seed selection simultaneously. Meanwhile, the runtime overhead of RandSet is minimal, suiting a high-frequency seed scheduling task. Our key insight is to introduce randomness into corpus reduction so as to enjoy the two benefits of a randomized algorithm: randomized output (i.e., diverse seed selection) and low runtime overhead. Specifically, we formulate the corpus reduction in seed scheduling as a classic set cover problem and compute a randomized subset of seed corpus as a set cover to cover all features of the entire corpus. We then develop a novel seed scheduling approach using the randomized corpus subset. Our technique can effectively mitigate seed explosion by scheduling a small and randomized subset of the corpus rather than the entire corpus. We implement RandSet on three popular fuzzers: AFL++, LibAFL and Centipede to showcase its general algorithmic design. We perform a comprehensive evaluation of RandSet on three benchmarks: standalone programs, FuzzBench and Magma. Our evaluation results show that RandSet can achieve significantly more diverse seed selection compared with other corpus reduction techniques. RandSet also yields high reduction ratio, achieving an average subset ratio of 4.03% and 5.99% after corpus reduction in terms of standalone programs and FuzzBench programs. In terms of fuzzing performance gain from our randomized corpus reduction, RandSet achieves a 16.58% gain on standalone programs and up to 3.57% gain on FuzzBench programs in AFL++. RandSet triggers up to 7 more ground-truth bugs than the state-of-the-art fuzzer on Magma, while introducing only 3.93% overhead on standalone programs and as low as 1.17% overhead on FuzzBench.

1 October 2026 in Proceedings of the ACM on Software Engineering
Verified
Journal Article
Open access
YXYuchong Xie
Yuchong Xie
MLMingyu Luo
Mingyu Luo
ZLZesen Liu
Zesen Liu
ZZZhixiang Zhang
Zhixiang Zhang
KZKaikai Zhang
Kaikai Zhang
DSDongdong She
Dongdong She
Coding agents powered by large language models are becoming central modules of modern IDEs. They help users to perform various complex coding tasks by invoking tools. Although powerful, tool-invocation operation in coding agents opens a substantial attack surface for adversaries. Prior work has demonstrated attacks against both general-purpose LLM agents and domain-specific agents. However, to our knowledge, no previous works focus on the security risks of tool-invocation in coding agents. To fill this gap, we conduct the first systematic, in-depth red-teaming from a tool-invocation perspective in six popular real-world coding agents: Cursor, Claude Code, Copilot, Windsurf, Cline, and Trae. Our red-teaming proceeds in two phases. In Phase 1, we conduct a prompt leakage as reconnaissance to recover system prompts and related context. Specifically, we identify a mode gap between chat generation and tool-call argument generation: during schema-driven argument completion, the model behaves as if it is performing benign structured filling and may copy hidden agent context into tool-call arguments. We instantiate this gap as ToolLeak, which exfiltrates agent-internal prompts (e.g., system prompts and tool metadata) via required tool parameters. In Phase 2, we hijack the tool-invocation behavior of the coding agent with a novel two-channel prompt injection in the tool description and the tool return. Our hijacking achieves remote code execution (RCE) on major real-world coding agents. We adaptively construct the malicious payload using leaked security information in Phase 1. Our evaluation shows that ToolLeak substantially outperforms strong prompt-leak baselines in both emulated and real-world settings. In the emulated setting, ToolLeak achieves the best overall prompt-exfiltration performance across all six simulated coding agents. On real-world coding agents, ToolLeak achieves the best pseudo-recall on 18 of 25 evaluated agent-LLM pairs. Furthermore, our red-teaming successfully hijacks all six real-world coding agents for RCE and consistently yields higher attack success rates than baseline attacks. Lastly, we present two case studies on Cursor and Claude Code to demonstrate the real-world impact of our red-teaming.