The Model Context Protocol (MCP) has recently emerged as a standardized interface for connecting language models with external tools and data. As the ecosystem rapidly expands, the lack of a structured, comprehensive view of existing MCP artifacts presents challenges for research. To bridge this gap, we introduce MCPCORPUS, a large-scale dataset containing around 14K MCP servers and 300 MCP clients. Each artifact is annotated with 20+ normalized attributes capturing its identity, interface configuration, GitHub activity, and metadata. MCPCORPUS provides a reproducible snapshot of the real-world MCP ecosystem, enabling studies of adoption trends, ecosystem health, and implementation diversity. To keep pace with the rapid evolution of the MCP ecosystem, we provide utility tools for automated data synchronization, normalization, and inspection. Furthermore, to support efficient exploration and exploitation, we release a lightweight web-based search interface. MCPCORPUS is publicly available at: https://github.com/Snakinya/MCPCorpus. The video is at https://youtu.be/2a9WrHMcfxU.

Identifying the impact scope and scale is critical for software supply chain vulnerability assessment. However, existing studies face substantial limitations. First, prior studies either work at coarse package-level granularity—producing many false positives—or fail to accomplish whole-ecosystem vulnerability propagation analysis. Second, although vulnerability assessment indicators like CVSS characterize individual vulnerabilities, no metric exists to specifically quantify the dynamic impact of vulnerability propagation across software supply chains. To address these limitations and enable accurate and comprehensive vulnerability impact assessment, we propose a novel approach: (i) a hierarchical worklist-based algorithm for whole-ecosystem and call-graph-level vulnerability propagation analysis and (ii) the Vulnerability Propagation Scoring System (VPSS), a dynamic metric to quantify the scope and evolution of vulnerability impacts in software supply chains. We implement a prototype of our approach in the Java Maven ecosystem and evaluate it on 100 real-world vulnerabilities. Experimental results demonstrate that our approach enables effective ecosystem-wide vulnerability propagation analysis, and provides a practical, quantitative measure of vulnerability impact through VPSS.

Graph Database Management Systems (GDBMSs) store graphs as data. They are used naturally in applications such as social networks, recommendation systems and program analysis. However, they can be affected by logic bugs, which cause the GDBMSs to compute incorrect results and subsequently affect the applications relying on them. In this work, we propose injective and surjective Graph Query Transformation (GQT) to detect logic bugs in GDBMSs. Given a query Q, we derive a mutated query Q', so that either their result sets are: (i) semantically equivalent; or (ii) variant based on the mutation to be either a subset or superset of each other. When the expected relationship between the results does not hold, a logic bug in the GDBMS is detected. The key insight to mutate Q is that the graph pattern in graph queries enables systematic query transformations derived from injective and surjective mappings of the directed edge sets between Q and Q'. We implemented injective and surjective Graph Query Transformation (GQT) as a tool called GraphGenie and evaluated it on 6 popular and mature GDBMSs. GraphGenie has found 25 unknown bugs, comprising 16 logic bugs, 3 internal errors, and 6 performance issues. Our results demonstrate the practicality and effectiveness of GraphGenie in detecting logic bugs in GDBMSs which has the potential for improving the reliability of applications relying on these GDBMSs.

Software vulnerabilities pose critical security and risk concerns. Many techniques are proposed to assess and prioritize vulnerabilities. To evaluate their performance, researchers often craft datasets from limited data sources, lacking a global overview of broad vulnerability intelligence. The repetitive data preparation process complicates the evaluation of new solutions. To solve this issue, we propose VulZoo, a comprehensive vulnerability intelligence dataset that covers 17 vulnerability data sources. We also construct connections among these sources, enabling more straightforward configuration and adaptation for different tasks. VulZoo provides utility scripts for automatic data synchronization and cleaning, relationship mining, and statistics generation. We make VulZoo publicly available and maintain it with incremental updates. We believe that VulZoo serves as a valuable input to vulnerability assessment and prioritization studies. The video is at https://youtu.be/EvoxQmUAHtw. The dataset is at https://github.com/NUS-Curiosity/VulZoo.

Android malware attacks have posed a severe threat to mobile users, necessitating a significant demand for the automated detection system. Among the various tools employed in malware detection, graph representations (e.g., function call graphs) have played a pivotal role in characterizing the behaviors of Android apps. However, though achieving impressive performance in malware detection, current state-of-the-art graph-based malware detectors are vulnerable to adversarial examples. These adversarial examples are meticulously crafted by introducing specific perturbations to normal malicious inputs. To defend against adversarial attacks, existing defensive mechanisms are typically supplementary additions to detectors and exhibit significant limitations, often relying on prior knowledge of adversarial examples and failing to defend against unseen types of attacks effectively. In this paper, we propose MASKDROID, a powerful detector with a strong discriminative ability to identify malware and remarkable robustness against adversarial attacks. Specifically, we introduce a masking mechanism into the Graph Neural Network (GNN) based framework, forcing MASKDROID to recover the whole input graph using a small portion (e.g., 20%) of randomly selected nodes.This strategy enables the model to understand the malicious semantics and learn more stable representations, enhancing its robustness against adversarial attacks. While capturing stable malicious semantics in the form of dependencies inside the graph structures, we further employ a contrastive module to encourage MASKDROID to learn more compact representations for both the benign and malicious classes to boost its discriminative power in detecting malware from benign apps and adversarial examples.

Linux kernel vulnerability reproduction is a critical task in system security. To reproduce a kernel vulnerability, the vulnerable environment and the Proof of Concept (PoC) program are needed. Most existing research focuses on the generation of PoC, while the construction of environment is overlooked. However, establishing an effective vulnerable environment to trigger a vulnerability is challenging. Firstly, it is hard to guarantee that the selected kernel version for reproduction is vulnerable, as the vulnerability version claims in online databases can occasionally be incorrect. Secondly, many vulnerabilities cannot be reproduced in kernels built with default configurations. Intricate non-default kernel configurations must be set to include and trigger a kernel vulnerability, but less information is available on how to recognize these configurations.

Detecting code functional similarity forms the basis of various software engineering tasks. However, the detection is challenging as functionally similar code fragments can be implemented differently, e.g., with irrelevant syntax. Recent studies incorporate program dependencies as semantics to identify syntactically different yet semantically similar programs, but they often focus only on local neighborhoods (e.g., one-hop dependencies), limiting the expressiveness of program semantics in modeling functionalities. In this paper, we present Tailor that explicitly exploits deep graph-structured code features for functional similarity detection. Given source-level programs, Tailor first represents them into code property graphs (CPGs) - which combine abstract syntax trees, control flow graphs, and data flow graphs - to collectively reason about program syntax and semantics. Then, Tailor learns representations of CPGs by applying a CPG-based neural network (CPGNN) to iteratively propagate information on them. It improves over prior work on code representation learning through a new graph neural network (GNN) tailored to CPG structures instead of the off-the-shelf GNNs used previously. We systematically evaluate Tailor on C and Java programs using two public benchmarks. Experimental results show that Tailor outperforms the state-of-the-art approaches, achieving 99.8% and 99.9% F-scores in code clone detection and 98.3% accuracy in source code classification.

Developers insert logging statements into source code to monitor system execution, which forms the basis for software debugging and maintenance. For distinguishing diverse runtime information, each software log is assigned with a separate verbosity level (e.g., trace and error). However, choosing an appropriate verbosity level is a challenging and error-prone task due to the lack of specifications for log level usages. Prior solutions aim to suggest log levels based on the code block in which a logging statement resides (i.e., intra-block features). Such suggestions, however, do not consider information from surrounding blocks (i.e., inter-block features), which also plays an important role in revealing logging characteristics.

System auditing provides a low-level view into cyber threats by monitoring system entity interactions. In response to advanced cyber-attacks, one prevalent solution is to apply data provenance analysis on audit records to search for anomalies (anomalous behaviors) or specifications of known attacks. However, existing approaches suffer from several limitations: 1) generating high volumes of false alarms, 2) relying on expert knowledge, or 3) producing coarse-grained detection signals. In this paper, we recognize the structural similarity between threat detection in cybersecurity and recommendation in information retrieval. By mapping security concepts of system entity interactions to recommendation concepts of user-item interactions, we identify cyber threats by predicting the preferences of a system entity on its interactive entities. Furthermore, inspired by the recent advances in modeling high-order connectivity via item side information in the recommendation, we transfer the insight to cyber threat analysis and customize an automated detection system, SHADEWATCHER. It fulfills the potential of high-order information in audit records via graph neural networks to improve detection effectiveness. Besides, we equip SHADEWATCHER with dynamic updates towards better generalization to false alarms. In our evaluation against both real-life and simulated cyber-attack scenarios, SHADEWATCHER shows its advantage in identifying threats with high precision and recall rates. Moreover, SHADEWATCHER is capable of pinpointing threats from nearly a million system entity interactions within seconds.