About MutahunterAI
Mutahunter is an open source command line tool for mutation testing that uses a large language model to generate the mutants. Mutation testing makes small deliberate changes to source code and checks whether the existing tests fail, which shows how well those tests actually protect the code. The project describes itself as language agnostic.
A user installs the Python package, which needs Python 3.11 or later, sets an API key for a model provider such as OpenAI, and runs the tool against a source file with the command that executes the project's tests. The README example runs a Maven test command against a Java class with the GPT 4o mini model, and prints each mutant as killed or survived. A final report gives the mutation coverage percentage, the total, survived, killed, timeout and compile error mutant counts, and the model cost of the run.
It suits developers and test engineers who want to measure the strength of a unit test suite without a language specific tool. The repository includes a Java Maven example and a test folder.
The code is published on GitHub under the AGPL 3.0 license and is free to use. Model usage is billed by whichever provider the user connects with their own key, and the project sets out no paid plans.