Automating Bug Analysis in Regulatory Information Requirements Using AI Agents

• Wendell Marques Sidia
• Caina Oliveira Sidia
• Edluce Veras Sidia
• Eduardo Oliveira Sidia
• Thiago Fragata Sidia

DOI: https://doi.org/10.5753/sast.2026.31114

Resumo

This paper proposes an agent-based system for the automated validation of homologation compliance in Android software builds. To address the inefficiencies of manual verification, the proposed system employs an autonomous agent equipped with specialized tools to interface with internal systems, enabling the systematic analysis and cross-referencing of embedded assets and device parameters against regulatory specifications. Consequently, the system eliminates manual intervention in the verification of regulatory information requirements. The findings suggest that the delegation of compliance tasks to specialized AI agents enhances operational reliability and optimizes the quality assurance lifecycle in large-scale mobile development.

Palavras-chave: Homologation, Mobile Device Compliance, Autonomous AI Agents, React Paradigm, Langchain, Large Language Model Orchestration

Referências

Islem Bouzenia and Michael Pradel. 2025. Understanding software engineering agents: A study of thought-action-result trajectories. arXiv preprint arXiv:2506.18824 (2025).

Dongming Jin, Zhi Jin, Xiaohong Chen, and Chunhui Wang. 2024. MARE: Multi-Agents Collaboration Framework for Requirements Engineering. CoRR abs/2405.03256 (2024). arXiv preprint arXiv:2405.03256 10 (2024).

Haolin Jin, Linghan Huang, Haipeng Cai, Jun Yan, Bo Li, and Huaming Chen. 2024. From llms to llm-based agents for software engineering: A survey of current, challenges and future. arXiv preprint arXiv:2408.02479 (2024).

Wendell Marques, Andrezza De Melo Bonfim, Edluce Leitao Veras, Daniel Souza, Vinicius Gabriel da Silva, et al. 2025. Application of Computer Vision to image review process in the software development. In Simpósio Brasileiro de Testes de Software Sistemático e Automatizado (SAST). SBC, 150–152.

Malik Abdul Sami, Muhammad Waseem, Zheying Zhang, Zeeshan Rasheed, Kari Systä, and Pekka Abrahamsson. 2024. AI based multiagent approach for requirements elicitation and analysis. arXiv preprint arXiv:2409.00038 (2024).

Yongjian Tang and Thomas Runkler. 2026. LLM-Based Agentic Systems for Software Engineering: Challenges and Opportunities. arXiv preprint arXiv:2601.09822 (2026).