wumpus-hunter - run test suite repeatedly to find flaky tests

It is common that an automated test suite for some software mostly works, but that one or more tests sometimes fail. These are called "flaky tests".

The wumpus hunter is a tool that runs the test suite for some software repeatedly. At a very high level, the wumpus hunter clones the other software's git repository, and runs the test suite, and does this repeatedly, and logs and reports results. The details are slightly more complicated, but the basic idea is that.

See spec.yaml for a sample configuration file.

Run wumpus-hunter --help (or cargo run -- --help from the source tree) to get an overview of how to run.

Example:

wumpus-hunter spec.yaml --logs /tmp

License

The wumpus hunter is distributed under the terms of both the MIT license and the Apache License (Version 2.0).

See LICENSE-APACHE and LICENSE-MIT for details.

lib.rs:

Run test suite of a project repeatedly, to find flaky tests that only sometimes fail.

It is surprisingly easy to accidentally implement an automated test so that it sometimes fails. This might be, for example, when there is a race condition in the test, or in the code under test. The wumpus hunter helps find such tests, by running them repeatedly.

The wumpus hunter is configured to run a test command for a given git repository and branch in that repository. It clones the repository, and pulls updates, and runs the command, either for a given number of times, or indefinitely.

The wumpus hunter is primarily a command line program, but is exposed also as a library in the hope it can be useful for building other useful tools.