Browse documentation

Build with a Job

Run a Job, inspect its result, and understand artifact access.

Use a Job to run a repeatable build against project Git source. This guide assumes a Job is already configured and you have permission to run it; creating a Job and configuring build inputs are separate administrative steps. A Job is a definition, while a Run records one execution at an exact commit.

Select the Job and source

rp job list
rp job show <job-number>

Check the Job's stored source and execution settings before running. A manual Run can override its source without changing the Job definition. Use a full commit SHA when the result must correspond to one exact revision, for example when you plan to publish its artifacts.

Run and inspect

rp job run <job-number> --source <full-sha> --attach
rp job runs <job-number>
rp job log <job-number> <run-number>

--attach waits and streams progress; the Run is still identified by its project, Job number, and Run number. A Job with declared non-secret Containerfile build inputs also accepts repeatable --build-arg NAME=VALUE. Do not put secrets in build arguments. Confirm the Run succeeded and its source matches what you intended before relying on any artifact.

When the Job produces artifacts, rp job artifacts <job-number> <run-number> --output <directory> downloads them for authorized project members. Job artifacts are not anonymous public downloads. A successful Run can be retained only according to its Job's retention policy unless it is frozen or held by a published Release. A Run is not an App deployment; deploy an App separately if you intend to run a service.