EvoMaster is an open-source AI-driven tool for automatically generating system-level test cases for web and enterprise applications, primarily targeting REST, GraphQL, and RPC APIs. It aims to improve software quality by generating effective test suites for regression testing and identifying application faults, benefiting developers and QA engineers.

How It Works

EvoMaster employs an Evolutionary Algorithm combined with Dynamic Program Analysis to generate test cases. It starts with an initial population of random tests and iteratively evolves them to maximize metrics like code coverage and fault detection. AI heuristics, rooted in Search-Based Software Testing, further enhance performance. For white-box testing, it analyzes JVM bytecode, utilizing techniques like testability transformations and taint analysis for more effective test generation.

Quick Start & Requirements

• Install/Run: docker run -v "$(pwd)/generated_tests":/generated_tests webfuzzing/evomaster --blackBox true --maxTime 30s --ratePerMinute 60 --bbSwaggerUrl https://petstore.swagger.io/v2/swagger.json
• Prerequisites: Docker is required for the example. White-box testing requires a JVM (JDK 8 supported, higher versions may require specific settings). REST APIs need an OpenAPI/Swagger schema.
• Resources: No major hardware requirements; works on older laptops. Execution time for good results can range from hours to a day.
• Docs: www.evomaster.org

Highlighted Details

• Independently studied as a top-performing fuzzer for RESTful APIs.
• Supports white-box testing for JVM applications, analyzing bytecode and SQL interactions.
• Generates tests in Python, JavaScript, Java JUnit, and Kotlin JUnit formats.
• Handles authentication via headers and cookies, with dynamic token acquisition support.

Maintenance & Community

EvoMaster is a research-driven project with funding from the European Research Council and the Research Council of Norway. Community interaction is encouraged via GitHub issues and discussions.

Licensing & Compatibility

• License: LGPL (v3).
• Compatibility: Generally compatible with commercial use, but LGPL's copyleft provisions apply to modifications of the library itself.

Limitations & Caveats

White-box testing requires manual driver creation. Support for JDKs above 8 requires specific configuration. RPC API schema definitions are not directly supported, requiring manual driver implementation. Mocking external services is a work in progress.