Lateping ping

Find out when a scheduled GitHub Actions workflow stops running.

GitHub turns off scheduled workflows in public repos after 60 days without activity, and delays or drops runs under load. Nothing fails, so nothing tells you. This action reports each run to Lateping: if a run doesn't report on schedule, or reports a failure, you get an alert.

Use it

1. Create a check at app.lateping.com with the same cron expression as your workflow, timezone UTC.

2. Save its ping URL as a repository secret named LATEPING_URL:

   gh secret set LATEPING_URL --body "https://lateping.com/p/<check-id>"

3. Add the action as the last step of the job:

   on:
     schedule:
       - cron: "17 2 * * *"
   
   jobs:
     backup:
       runs-on: ubuntu-latest
       steps:
         - uses: actions/checkout@v4
         - run: ./scripts/backup.sh
   
         - uses: lateping/ping@v1
           if: always()
           with:
             url: ${{ secrets.LATEPING_URL }}
             status: ${{ job.status }}

if: always() makes the step run even when an earlier step failed. A successful job sends a success ping; a failed or cancelled job sends /fail, which alerts within a minute. If the workflow never runs at all, no ping arrives and the check alerts after its grace period.

Time each run

Add a start ping as the first step. Lateping records how long each run took (run analytics are on Pro and above).

      - uses: lateping/ping@v1
        with:
          url: ${{ secrets.LATEPING_URL }}
          status: start

Report a command's exit code

To report one command rather than the whole job, pass its exit code. 0 is a success; anything else is a failure, and the alert shows the code.

      - id: backup
        run: ./scripts/backup.sh; echo "code=$?" >> "$GITHUB_OUTPUT"
      - uses: lateping/ping@v1
        with:
          url: ${{ secrets.LATEPING_URL }}
          status: ${{ steps.backup.outputs.code }}

Inputs

Input	Default
url	required	The check's ping URL. Keep it in a secret.
status	success	start, success, ${{ job.status }} (success, failure, cancelled), or an exit code 0 to 255.
fail-on-error	false	Fail this step if the ping can't be sent. Off by default, so monitoring never breaks your job.

The step outputs signal: start, success or fail.

How it works

One curl call: POST to the ping URL, /start, /fail or /<exit code>, with a 10 second timeout and three retries. No dependencies, no data beyond the request itself, and the URL is never printed to the log. Guide: GitHub Actions scheduled workflow not running.

License

MIT. See LICENSE.