https://repositorio.ufu.br/handle/123456789/17904 2026-04-03T18:46:21Z https://repositorio.ufu.br/handle/123456789/48151 Title: Extração automatizada de chaves criptográficas em Ransomwares: uma abordagem forense baseada em AOB Abstract: Ransomware threats have emerged as a primary vector for cyberattacks, with incidence and the complexity of their techniques continuing to grow. Recent reports highlight the significant financial impacts of these attacks, with losses reaching millions of dollars per incident. Ransomware typically employs cryptographic mechanisms to encrypt files on compromised systems, making data recovery contingent on the payment of a ransom, generally associated with the receipt of the cryptographic key or a decryption tool. In this context, this work proposes an automated mechanism for extracting cryptographic keys from ransomware by analyzing volatile memory during malware execution. To gain greater control over the malware’s internal behavior and to enable reproducible experi ments, a custom ransomware sample was developed, inspired by notable strains such as WannaCry and LockBit. The proposed approach leverages AOB (Array of Bytes) signa tures to identify, in real-time, routines related to the generation and temporary storage of cryptographic keys in volatile memory. The experimental process involved static analysis with IDA Free, dynamic memory inspection with Cheat Engine, and the development of an automated tool, AOBTool. The results indicate that the cryptographic key can be located and extracted from memory before it is disposed of, within a timeframe shorter than that required to complete the file encryption process. This highlights the potential of the proposed technique as a valuable support tool for digital forensics and for mitigating ransomware-related incidents. 2025-12-19T00:00:00Z https://repositorio.ufu.br/handle/123456789/48103 Title: Análise Comparativa de Tecnologias de IA para o Desenvolvimento de Chatbots Especializados 2025-09-23T00:00:00Z https://repositorio.ufu.br/handle/123456789/48094 Title: GPT Teacher: desenvolvimento de um agente de LLM para programação assistida em ambiente VSCode Abstract: The computer programming teaching and learning process presents significant challenges that often result in comprehension difficulties and student demotivation. This paper presents the development and evaluation of GPT Teacher, an LLM-based agent for assisted programming integrated into the Visual Studio Code (VSCode) environment. The primary objective was to design and implement an LLM-based tool focused on programming education, capable of supporting the acquisition of programming competencies more effectively than generic coding assistants. The methodology involved creating a functional prototype based on a dual-agent architecture: a Diagnostic Agent, responsible for technical code analysis, and a Guidance Agent, tasked with translating this analysis into a constructive and educational dialogue for the student. The results of the functional validation demonstrate that the proposed approach is robust and promising, confirming the hypothesis that LLM agents, when structured within a specialized system, can serve as powerful allies in the programming teaching-learning process by reconciling technical rigor with pedagogical effectiveness. 2025-09-26T00:00:00Z https://repositorio.ufu.br/handle/123456789/48093 Title: Geração automática de relatórios de confiabilidade: projeto e implementação da camada front-end para identificação de causas de falhas de software na plataforma X-RAT Abstract: Operating systems such as Windows 10 and 11 support critical activities in both personal and corporate environments, so failures in this software can compromise service availability and user experience. In this context, it becomes relevant to systematically analyze failure events and structure information that supports the diagnosis of root causes and the improvement of the reliability of the evaluated computers. This work presents the development and enhancement of the X-RAT platform (X-Reliability Analysis Tool), a tool designed for the analysis and automatic generation of software reliability reports in operating systems (OSs), focusing on Windows 10 and 11. The platform collects logs, extracts failure events, stores the data in a database, and processes them through failure categorization algorithms based on criteria such as kernel failures (OSₖₙₗ), service failures (OSₛᵥ𝒸), system application failures (OSₐₚₚ), and user application failures (USRₐₚₚ). In addition, several statistical metrics are computed, including the frequency of failure causes, the distribution of failures by category, and the occurrence of these categories throughout the day (Early Morning, Morning, Afternoon, and Evening) and the days of the week. Finally, based on the times between failures, reliability metrics such as MTBF (Mean Time Between Failures) and MTTF (Mean Time To Failure) are applied, as well as analyses based on probability distributions. Each analysis is formalized in a report automatically generated by the platform. The goal of this work is to fully refactor the previously developed platform, which is currently outdated and incomplete. The new version will feature additional functionalities, including automatic failure collection and more accurate diagnostics of failure causes, enabling deeper analysis and contributing to more informed decision-making in improving the reliability of the analyzed computers. 2025-11-18T00:00:00Z