Desenvolvimento de software assistido por inteligência artificial: uma análise comparativa entre ambientes conversacionais e ferramentas agênticas.

The adoption of Large Language Models (LLMs) has transformed the software development cycle, offering significant gains in rapid prototyping. However, the use of purely conversational interfaces frequently results in a loss of architectural coherence and code degradation across multiple iterations. This paper critically analyzes the construction process of a sensor monitoring web application (PHP and XML persistence), comparing initial generation in a conversational environment (Claude Chat) with structured refinement in a code-oriented agentic tool (Claude Code). The results highlight the occurrence of the code drift phenomenon in chat interactions, in addition to integration failures, security vulnerabilities (Cross-Site Scripting), and version incompatibilities. In contrast, the transition to an agentic environment combined with context persistence in Markdown files (CLAUDE.md and README.md) and modular technical auditing reduced token consumption, eliminated critical failures, and ensured system stability. Furthermore, the infeasibility of local models (Gemma 4) in restricted computational infrastructures compared to the efficiency of cloud environments is discussed.

Authors

Publication Details

Journal
Revista fisio&terapia.
Published
2026-11-19
DOI
https://doi.org/10.69849/84apx086
Primary Topic
Software Engineering Research
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Desenvolvimento de software assistido por inteligência artificial: uma análise comparativa entre ambientes conversacionais e ferramentas agênticas.

Valber Sardi Lopes, Manoel Henrique Reis de Oliveira, Adriano Soares de Oliveira Bailão, Andrea Barboza Proto Sardi
Revista fisio&terapia.
Software Engineering Research
article

Desenvolvimento de software assistido por inteligência artificial: uma análise comparativa entre ambientes conversacionais e ferramentas agênticas.

Valber Sardi Lopes, Manoel Henrique Reis de Oliveira, Adriano Soares de Oliveira Bailão, Andrea Barboza Proto Sardi
article en

Abstract

The adoption of Large Language Models (LLMs) has transformed the software development cycle, offering significant gains in rapid prototyping. However, the use of purely conversational interfaces frequently results in a loss of architectural coherence and code degradation across multiple iterations. This paper critically analyzes the construction process of a sensor monitoring web application (PHP and XML persistence), comparing initial generation in a conversational environment (Claude Chat) with structured refinement in a code-oriented agentic tool (Claude Code). The results highlight the occurrence of the code drift phenomenon in chat interactions, in addition to integration failures, security vulnerabilities (Cross-Site Scripting), and version incompatibilities. In contrast, the transition to an agentic environment combined with context persistence in Markdown files (CLAUDE.md and README.md) and modular technical auditing reduced token consumption, eliminated critical failures, and ensured system stability. Furthermore, the infeasibility of local models (Gemma 4) in restricted computational infrastructures compared to the efficiency of cloud environments is discussed.

Revista fisio&terapia.Vol. 30(164)
Software Engineering Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon