What are the main components of the Cucumber ecosystem?

A working Cucumber setup is made up of a handful of distinct pieces that work together, each with a specific responsibility.

1. Gherkin - the parser and grammar defining the plain-language syntax feature files are written in.
2. Feature files - the .feature files containing scenarios written in Gherkin.
3. Step definitions - code implementing the behavior for each Gherkin step.
4. Hooks - setup/teardown code running around scenarios or steps, independent of specific step text.
5. A test runner - the JUnit/TestNG (or equivalent) integration that discovers and executes everything.
6. Formatters/plugins - components producing output like HTML or JSON reports from a run's results.

These pieces are deliberately separated so that the specification layer (feature files, readable by anyone) stays decoupled from the implementation layer (step definitions, written by developers), which is central to Cucumber's whole premise: business-readable specifications and developer-maintained automation coexisting without either one compromising the other.