Abstract

Abstract

Abstract
This study explores prompt engineering for automated white-box integration testing of RESTful APIs using Large Language Models (LLMs). Four versions of prompts were designed and tested across three OpenAI models (GPT-3.5 Turbo, GPT-4 Turbo, and GPT-4o) to assess their impact on code coverage, token consumption, execution time, and financial cost. The results indicate that different prompt versions, especially with more advanced models, achieved up to 90% coverage, although at higher costs. Additionally, combining test sets from different models increased coverage, reaching 96% in some cases. We also compared the results with EvoMaster, a specialized tool for generating tests for REST APIs, where LLM-generated tests achieved comparable or higher coverage in the benchmark projects. Despite higher execution costs, LLMs demonstrated superior adaptability and flexibility in test generation.

Abstract

Abstract
Recent verification tools aim to make formal verification more accessible for software engineers by automating most of the verification process. However, the manual work and expertise required to write verification helper code, such as loop invariants and auxiliary lemmas and assertions, remains a barrier. This paper explores the use of Large Language Models (LLMs) to automate the generation of loop invariants for programs in Dafny. We tested the approach on a curated dataset of 100 programs in Dafny involving arrays, strings, and numeric types. Using a multimodel approach that combines GPT-4o and Claude 3.5 Sonnet, correct loop invariants (passing the Dafny verifier) were generated at the first attempt for 92% of the programs, and in at most five attempts for 95% of the programs. Additionally, we developed an extension to the Dafny plugin for Visual Studio Code to incorporate automatic loop invariant generation into the IDE. Our work stands out from related approaches by handling a broader class of problems and offering IDE integration.

Abstract
Large language model (LLM)-powered assistants are increasingly used for generating program code and unit tests, but their application in acceptance testing remains underexplored. To help address this gap, this paper explores the use of LLMs for generating executable acceptance tests for web applications through a two-step process: (i) generating acceptance test scenarios in natural language (in Gherkin) from user stories, and (ii) converting these scenarios into executable test scripts (in Cypress), knowing the HTML code of the pages under test. This two-step approach supports acceptance test-driven development, enhances tester control, and improves test quality. The two steps were implemented in the AutoUAT and Test Flow tools, respectively, powered by GPT-4 Turbo, and integrated into a partner company's workflow and evaluated on real-world projects. The users found the acceptance test scenarios generated by AutoUAT helpful 95% of the time, even revealing previously overlooked cases. Regarding Test Flow, 92% of the acceptance test cases generated by Test Flow were considered helpful: 60% were usable as generated, 8% required minor fixes, and 24% needed to be regenerated with additional inputs; the remaining 8% were discarded due to major issues. These results suggest that LLMs can, in fact, help improve the acceptance test process, with appropriate tooling and supervision.