Skip to main content

Testing

Husky includes automated coverage and a large manual scenario suite under testproject/.

Fast validation loop​

For normal development:

make build
make test
husky validate --strict

Automated coverage areas​

The repository includes tests for:

  • config parsing and validation
  • schedule evaluation
  • DAG cycle detection
  • retry state transitions
  • output capture modes
  • healthcheck behavior
  • REST API handlers
  • integration and crash-recovery scenarios

Manual scenario suite​

The main operator guide is:

  • testproject/MANUAL_TEST_GUIDE.md

It covers end-to-end scenarios such as:

  • basic scheduling
  • DAG pipelines and output passing
  • retry and concurrency controls
  • SLA budgets and healthchecks
  • notifications
  • tags and bulk operations
  • audit trail
  • timezones
  • crash recovery and catchup
  • validation errors
  • REST API spot checks
  • dashboard workflows
  • WebSocket log streaming

Useful scenario directories​

  • testproject/01-basic
  • testproject/02-pipeline
  • testproject/03-reliability
  • testproject/04-sla-healthcheck
  • testproject/05-notifications
  • testproject/06-tags
  • testproject/07-audit-trail
  • testproject/08-timezone
  • testproject/09-catchup
  • testproject/10-validation-errors

Suggested release checklist​

Before shipping a change:

  1. run build and automated tests
  2. validate config fixtures
  3. exercise at least one manual scenario for the changed subsystem
  4. verify dashboard and API behavior if operator-facing changes were made
  5. verify crash or reload behavior for scheduler and executor changes