agentleFS
rstackjsrstackjs/agent-skills/skills/rstest-best-practices/SKILL.md

rstest-best-practices

Skill97 starsChanged 3 months agoMITView on GitHub

Set up, write, or review Rstest tests and configuration, including environments, mocks, snapshots, coverage, and CI.

We split this skill into the rules it gives an agent

20 rules. About 0 are common, 4 are distinctive, and 16 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 16

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Add @rstest/core, a package-local test script, and a config file whose extension matches the package's module mode.

    line 13
  2. Very specific:

    Verify test discovery with rstest list, then run focused tests, the full suite, and the project's existing build or type-check command.

    line 16
  3. Very specific:

    Use defineConfig from @rstest/core and a config file extension compatible with the package's module mode.

    line 20
  4. Very specific:

    Prefer explicit imports from @rstest/core over globals: true.

    line 21
  5. Very specific:

    Use setupFiles for shared matchers and cleanup.

    line 22
  6. Very specific:

    Use node for server/SSR tests, happy-dom or jsdom for simulated DOM tests, and Browser Mode for real-browser behavior.

    line 24
  7. Distinctive:

    Use multiple test projects only when environments or configurations genuinely differ.

    line 26
  8. Very specific:

    Prefer small tests that verify one public behavior.

    line 30
  9. Distinctive:

    Keep test APIs and test files in the same module system as the project.

    line 31
  10. Very specific:

    Prefer await expect(promise).resolves...

    line 32
  11. Very specific:

    Use includeSource only for small utilities, guard tests with import.meta.rstest, and define it as false in production builds.

    line 33
  12. Very specific:

    Use rs.fn() for functions, rs.spyOn() for object methods, and rs.mock() factories for modules.

    line 37
  13. Distinctive:

    Keep snapshots small and deterministic: use inline snapshots for short output, file snapshots for large structured output, and serializers …

    line 44
  14. Very specific:

    Use rstest for a single run.

    line 51
  15. Very specific:

    Use rstest list to verify discovery, positional filters or -t for focused runs, -u for intentional snapshot updates, and -c for a …

    line 52
  16. Very specific:

    If passing-test logs are too noisy, consider silent: 'passed-only'.

    line 54
  17. Very specific:

    Start with a focused repro and --reporter=verbose.

    line 58
  18. Very specific:

    Use DEBUG=rstest and inspect dist/.rstest-temp/.rsbuild/ to verify final Rstest/Rsbuild/Rspack configuration and generated build output.

    line 59
  19. Distinctive:

    Use breakpoints through the Rstest VS Code extension or a JavaScript Debug Terminal for runtime failures.

    line 60
  20. Very specific:

    Use rstest-debugging for systematic startup, build, runtime, logging, memory, or performance diagnosis.

    line 61

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Rstest best practices

Apply these rules when setting up, writing, or reviewing Rstest projects.

Workflow

  • Inspect the existing package manager, dependency policy, module mode, build configuration, test layout, and framework conventions before editing.
  • Add @rstest/core, a package-local test script, and a config file whose extension matches the package's module mode.
  • Reuse existing build configuration only when doing so lowers maintenance cost and remains compatible. An adapter is optional; use an independent Rstest config when adapter integration is incompatible or disproportionately costly, and explicitly reproduce required aliases, plugins, defines, environments, and externals.
  • Configure test discovery, environment, and setup files intentionally. Do not enable passWithNoTests unless an empty test set is an expected and documented state.
  • Verify test discovery with rstest list, then run focused tests, the full suite, and the project's existing build or type-check command.

Configuration and environments

  • Use defineConfig from @rstest/core and a config file extension compatible with the package's module mode.
  • Prefer explicit imports from @rstest/core over globals: true.
  • Use setupFiles for shared matchers and cleanup; use plugins, resolve, or source for ordinary build integration and low-level tools only when necessary.
  • Set include explicitly when the package's test layout is non-standard; do not assume the default discovery pattern.
  • Use node for server/SSR tests, happy-dom or jsdom for simulated DOM tests, and Browser Mode for real-browser behavior; install the selected environment package explicitly.
  • For React or Vue component tests, register the matching build plugin, use the corresponding Testing Library, and centralize matcher extension and cleanup in a setup file.
  • Use multiple test projects only when environments or configurations genuinely differ; keep global options such as reporters, coverage, pool, isolation, and bail at the root.

Test writing

  • Prefer small tests that verify one public behavior; use .skip or .todo intentionally and never commit .only.
  • Keep test APIs and test files in the same module system as the project; do not mix ESM/CJS assumptions without verifying the runtime and bundler behavior.
  • Prefer await expect(promise).resolves... and .rejects... over try/catch patterns that can miss assertions.
  • Use includeSource only for small utilities, guard tests with import.meta.rstest, and define it as false in production builds.

Mocking

  • Use rs.fn() for functions, rs.spyOn() for object methods, and rs.mock() factories for modules; mock external boundaries rather than the subject under test.
  • Choose clearMocks, resetMocks, or restoreMocks according to whether calls, implementations, or original methods must be restored between tests.
  • Choose the module-mocking API that matches the module system; verify the exact request string and register hoisted/static mocks before the module under test is evaluated.
  • Module mocks replace runtime behavior but do not necessarily remove the real dependency from the Rspack build graph. Use aliases or narrowly scoped externals only when the build still traverses an unwanted or unavailable dependency, and verify the runtime externalization format.

Snapshots and coverage

  • Keep snapshots small and deterministic: use inline snapshots for short output, file snapshots for large structured output, and serializers for paths or unstable data. Review every update.
  • Enable coverage with --coverage or coverage.enabled.
  • Choose a coverage provider supported by the installed Rstest version and add its package explicitly; do not copy a V8 or Istanbul provider choice without checking project needs.
  • Set coverage.include deliberately before adding thresholds; choose reporters for humans and CI artifacts separately.

Running and CI

  • Use rstest for a single run; rstest run is an optional explicit equivalent. Use rstest --watch or rstest watch only for local development.
  • Use rstest list to verify discovery, positional filters or -t for focused runs, -u for intentional snapshot updates, and -c for a non-default config.
  • In CI, never use watch mode. Use sharding with blob reports and rstest merge-reports when distributing tests; use JUnit when the CI system requires machine-readable results.
  • If passing-test logs are too noisy, consider silent: 'passed-only'; disable it temporarily when diagnosing setup or runtime output.

First-line debugging

  • Start with a focused repro and --reporter=verbose; use --printConsoleTrace for noisy or unclear console output.
  • Use DEBUG=rstest and inspect dist/.rstest-temp/.rsbuild/ to verify final Rstest/Rsbuild/Rspack configuration and generated build output.
  • Use breakpoints through the Rstest VS Code extension or a JavaScript Debug Terminal for runtime failures.
  • Use rstest-debugging for systematic startup, build, runtime, logging, memory, or performance diagnosis.

Documentation

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.