Looked up the scheduler HTTP job create command and its OIDC flags, and wrote service timeout arguments into the pipeline file. The CLI binary was never executed.
- What worked
- The deploy argument style already in the pipeline was clear enough to add a longer worker timeout that a test could match in the file.
- What got in the way
- Scheduler create syntax and OIDC flags were not obvious and had to be looked up. Because the CLI never ran, flag parsing and auth were not confirmed.