Skip to main content
The agent of a schedule automation’s run reads and reports with these three commands. You do not run them yourself. The runner gives the agent a run key and sets OUTERLAYER_RUN to the run’s id, and OuterLayer refuses a call with any other key. The commands only read and report. OuterLayer files the issues after the run ends. All three commands take --json, --url and --app-id. A failed command prints the reason and exits 1.

outerlayer run topics

List the topics that the automation already filed issues for, so the agent can use a topic’s exact text when it finds the same problem. Run it before the run reports.
The command prints one topic on each line, newest first, with at most 50 topics. It prints nothing when the automation has no topics. The list holds open and closed topics. With --json, the output is {"topics": ["<topic>", ...]}, and --quiet prints nothing. A gateway that has no topics route counts as a gateway with no topics: the command prints nothing and exits 0.

outerlayer run found-issue

Report one issue that the run found. Run it once for each issue.
The command prints how many more issues the run can report. With --json, the output is {"issuesLeft": <number>}. When the automation’s daily limit leaves none, the command refuses and names the limit.

outerlayer run found-nothing

Report that the run found nothing to file. Run it instead of found-issue, never after it.
With --json, the output is {"recorded": true}.