A pipeline job is an executable series of steps paired with an execution environment. Jobs turn a repository change into a bounded unit of work, such as building, testing, or planning infrastructure. section 6.3.7
Use jobs to make pipeline behavior explicit and reusable:
- Attach filters when a job should run only for a branch or tag. section 6.3.7
- Declare dependencies when jobs must run in sequence; omit them when independent jobs can run concurrently. section 6.3.7
- Keep repository-local pipeline code focused on triggers and ordering, while shared versioned libraries hold reusable step details. section 6.3.7
Infrastructure plan job
A practical job checks out code, initializes Terraform, creates a plan, renders it as JSON, and optionally runs a Checkov scan. Parameters and extension points allow teams to reuse the job without duplicating its implementation.
