Bashunit
bashunit is an open-source testing library for Bash scripts. The project is available at github.com/TypedDevs/bashunit and the hosted documentation is at bashunit.typeddevs.com. It provides a structured test runner for shell scripts: describe/test function blocks, assertion helpers, mocking support, and a pass/fail report output. It runs as a single self-contained binary.
The library brings the standard unit-testing pattern to Bash: tests are written as functions named test_* inside *_test.sh files; the runner discovers and executes them; each test uses assertion functions (assert_equals, assert_contains, assert_exit_code, etc.) to validate behaviour. A --filter flag runs subsets of tests. Output follows a standard green/red report format compatible with CI pipelines.
Placed in Tools / Assess / inner.
Shell scripts are a common execution layer for deployment automation, backup jobs, and setup routines, but they are almost universally untested. When a script breaks, the failure surfaces in production rather than in a test run. bashunit applies the same feedback loop that unit testing provides for application code — run assertions against the script's behaviour, catch regressions before they reach production.
The absence of competitors at this niche is notable: shellcheck provides static linting but no execution-level assertions; bats-core is the closest alternative but carries more dependency weight. bashunit is distributed as a single binary, which minimises friction for projects that do not want to introduce a package manager dependency for test infrastructure.
Inner position reflects the zero-overhead install, direct applicability to any project with non-trivial shell automation (deploy scripts, backup pipelines, setup scripts), and complementarity with the just and runme blips already on this radar.
Trial gate: at least one existing shell script with non-trivial logic covered by a bashunit test suite, with the suite running in CI.
