Malicious code in open-source repositories such as PyPI poses a growing threat to software supply chains. Traditional rule-based tools often overlook the semantic patterns in source code that are crucial for identifying adversarial components. Large language models (LLMs) show promise for software analysis, yet their use in interpretable and modular security pipelines remains limited. This paper presents LAMPS , a multi-agent system that employs collaborative LLMs to detect malicious PyPI packages. The system consists of four role-specific agents for package retrieval, file extraction, classification , and verdict aggregation , coordinated through the CrewAI framework. A prototype combines a fine-tuned CodeBERT model for classification with LLaMA 3 agents for contextual reasoning. LAMPS has been evaluated on two complementary datasets: D 1 , a balanced collection of 6,000 setup.py files, and D 2 , a realistic multi-file dataset with 1,296 files and natural class imbalance. On D 1 , LAMPS achieves 97.7% accuracy, surpassing MPHunter and TD-IDF stacking models–two state-of-the-art approaches. On D 2 , it reaches 99.5% accuracy and 99.5% balanced accuracy, outperforming RAG-based approaches and fine-tuned single-agent baselines. McNemar’s test confirmed these improvements as highly significant. The results demonstrate the feasibility of distributed LLM reasoning for malicious code detection and highlight the benefits of modular multi-agent designs in software supply chain security.

Malicious software packages in open-source ecosystems, such as PyPI, pose growing security risks. Unlike traditional vulnerabilities, these packages are intentionally designed to deceive users, making detection challenging due to evolving attack methods and the lack of structured datasets. In this work, we empirically evaluate the effectiveness of Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), and few-shot learning for detecting malicious source code. We fine-tune LLMs on curated datasets and integrate YARA rules, GitHub Security Advisories, and malicious code snippets with the aim of enhancing classification accuracy. We came across a counterintuitive outcome: While RAG is expected to boost up the prediction performance, it fails in the performed evaluation, obtaining a mediocre accuracy. In contrast, few-shot learning is more effective as it significantly improves the detection of malicious code, achieving 97% accuracy and 95% balanced accuracy, outperforming traditional RAG approaches. Thus, future work should expand structured knowledge bases, refine retrieval models, and explore hybrid AI-driven cybersecurity solutions.

The manufacturing sector increasingly relies on artificial intelligence (AI) to automate fault detection processes. However, existing methods for anomaly detection often struggle with data imbalances and high computational demands, limiting their scalability in industrial environments. This paper implements a system for efficient gearbox anomaly detection in manufacturing processes. We propose a machine learning-based automated system for detecting anomalies in the sound produced by gearbox systems in functional testing facilities. An evaluation of four machine learning (ML) algorithms was performed. Gaussian Mixture Model (GMM), Isolation Forest, K-Means, and One-Class Support Vector Machine (OC-SVM). The OC-SVM algorithm demonstrated the highest level of performance, achieving 88% specificity. This paper also discusses the development and on-site deployment of a fully functional solution that enables real-time fault detection. Our approach reinforces the feasibility of using sound analysis with semi-supervised learning to enhance anomaly detection in industrial settings, bridging the gap between AI-driven quality control and practical deployment. Future work will focus on improving the defective dataset and increasing the interpretability of the prediction model. These objectives are designed to promote improved transparency and trust in the system.