Skip to content

Quick Start Jobs

This guide runs the image you built in the repo quick start: create a job, configure its repo image, trigger a run, and read the output.

AI-assisted, human-approvednovem uses AI to review and keep our documentation up to date.

This guide assumes you have a built repo image. If you don't, run through the repo quick start first to create hello_repo.

1. Create the job

Terminal
novem -j hello_job -C

New jobs use the chains runtime by default.

2. Configure the repo image

The chain is a single step: the image built from hello_repo.

Terminal
echo "hello_repo" | novem -j hello_job -w config/chains

novem resolves the bare name to your image @<your_username>/hello_repo:latest.

3. (Optional) set an environment variable

Anything under config/env is injected into the container at run time. Give each variable a lower-case name. It is exposed inside the job in both lower- and upper-case.

Terminal
printf '%s' "world" | novem -j hello_job -w config/env/greet_target

4. Run it

Terminal
novem -j hello_job -R

# to pass input files, append one or more filenames, each prefixed with @
novem -j hello_job -R @data.csv

# save the run's output files to disk with -o/--output
novem -j hello_job -R -o ./out

5. Read the output

Each run is recorded under runs with its own log. The job's own log is a shortcut for the most recent one.

Terminal
novem -j hello_job -r log

You should see your program's output (Hello from my novem repo!).

Note: Run logs replace environment variable values with *** when those values are at least 10 characters long. Shorter values are not redacted because replacing them could also remove parts of ordinary log messages.

Next steps

  • Chains — pipe several repo images into a pipeline.
  • Schedule — run the job automatically on a cron schedule.
  • Runs — inspect a run's log, output, stats and status.