| name | zapo-manager-test-runner |
| description | Runs Zapo Manager local test suites. Use when user says "rodar testes do manager", "validar endpoints do manager", "validar botoes do frontend manager", "smoke fake", "fake-server", or asks to run the local API/UI/fake-server manager tests. |
Zapo Manager Test Runner
This skill runs the local Zapo Manager validation suite as an operator-facing quality gate. Act as a pragmatic test engineer: preserve the project's offline-safe defaults, separate real WhatsApp sends from mocked/local checks, and return evidence a developer can act on without reading raw Playwright logs.
Resolution rules
- Bare paths and
{skill-root} (e.g. scripts/run-manager-tests.ps1) resolve from this skill's installed directory.
{project-root} resolves to the active Zapo Manager repository root.zapo-manager-test-runner resolves to this skill directory's basename.
Defaults
- Treat
{project-root} as D:/Projetos Dev/Outros/apis-whatsapp-doc-testes/zapo-manager unless the user gives another explicit checkout.
- Use
npm run test:manager for the full local gate.
- Use
npm run test:manager:api for backend endpoint coverage.
- Use
npm run test:manager:ui for frontend button/function coverage with mocked API.
- Use
npm run test:manager:smoke:fake for fake-server smoke coverage against @zapo-js/fake-server, fully offline and without real WhatsApp numbers.
- Use
npm run test:smoke:real only when the user explicitly asks for real smoke tests or confirms a connected test instance.
- Do not start real WhatsApp sends from the offline-safe API/UI gate.
Run
Use scripts/run-manager-tests.ps1 instead of retyping commands when PowerShell is available:
powershell -ExecutionPolicy Bypass -File "{skill-root}/scripts/run-manager-tests.ps1" -Mode all -ProjectRoot "{project-root}"
Choose -Mode api, -Mode ui, -Mode ui-real, -Mode all, -Mode fake, or -Mode real from the user's request. If the script is unavailable, run the matching npm command directly from {project-root}.
ui-real assumes a backend process visible in its own terminal window and validates the UI against the live Express/Prisma/Postgres/Redis stack without mocking HTTP. Do not use it for WhatsApp sends or pairing flows.
Report
Return a compact result:
- command(s) run
- pass/fail status and failing test names if any
- whether a backend or Vite server was reused or started when visible in output
- generated artifacts worth opening, such as
test-results/
- any risk that remains outside the selected mode, especially real WhatsApp delivery and fake-server protocol coverage
If a failure depends on local runtime state, say that concretely. For example: a disconnected or unregistered WaClient can make message endpoints return runtime 5xx after auth/body validation has already passed.
