Docs › Vercel cron monitoring with RunVouch

Vercel cron monitoring with RunVouch

Vercel cron monitoring for the part Vercel leaves to you: a cron entry in vercel.json hits a route on a schedule and logs the status code. A route that returns 200 after an early return is a success; a function that hits its duration limit is a log line; a run that never started is nothing at all.

How it runs on Vercel Cron Jobs

Inside the route handler (Next.js route handler or serverless function), with two fetch() calls. Cron jobs run against the production deployment only.

Store the key

Project Settings, Environment Variables: RUNVOUCH_KEY, production scope. Keep CRON_SECRET as Vercel documents it so only Vercel can call the route.

Report the run (two HTTP calls)

// app/api/cron/report/route.ts
export async function GET(req: Request) {
  if (req.headers.get("authorization") !== `Bearer ${process.env.CRON_SECRET}`) return new Response("no", {status: 401});
  const h = {"X-API-Key": process.env.RUNVOUCH_KEY!, "content-type": "application/json"};
  const {run_id} = await (await fetch("https://api.runvouch.com/v1/runs/start", {method: "POST", headers: h,
    body: JSON.stringify({agent: "nightly-report", source: "vercel"})})).json();
  const rows = await buildReport();
  await fetch("https://api.runvouch.com/v1/runs/end", {method: "POST", headers: h,
    body: JSON.stringify({run_id, status: "ok", evidence: {rows: rows > 0}})});
  return Response.json({rows});
}
// vercel.json: {"crons": [{"path": "/api/cron/report", "schedule": "0 2 * * *"}]}

The start call returns run_id; the end call takes status (ok or fail), optional cost and tokens, and evidence as a JSON object of booleans. Full reference on the API page.

Register the cadence and caps

rv agent nightly-report --cadence 24h --grace 30m --max-runtime 1h --evidence --cap-run-cost 2

Register the agent once, from anywhere with the key. Cadence is what turns a schedule that stopped into an alert; --evidence makes a run without evidence a failure; the caps pause the agent when it overspends.

What goes silent on Vercel Cron Jobs

• Cron jobs only run on production; a preview deployment that looks fine never runs them. MISSED on the production agent is what tells you the schedule is not live.
• The Hobby plan runs crons once a day at most and may run them within the hour, not at the minute; set grace accordingly.
• Function duration limits end a run without a 5xx you will act on; STALLED covers a start without an end.

What Vercel Cron Jobs does not tell you

Vercel's documentation is explicit: Vercel will not retry an invocation if a cron job fails, and cron delivery is best effort, so a scheduled run can be skipped by a transient error, in which case no runtime log is created for it. There is no failure notification; errors are found through the View Logs button on the Cron Jobs settings page. On Hobby, a cron may run at any point within the scheduled hour. An Instant Rollback does not update the crons, so the schedule can keep calling the old routes.

What RunVouch detects

Set up in two minutes

1. Get a free key (3 agents, no card) and store it where this page says.
2. report the run (two http calls): copy the snippet above into the scheduled job.
3. Register the cadence once: rv agent nightly-report --cadence 24h --grace 30m --evidence, or let the first run create the agent and set the cadence on the dashboard.
4. Send one test alert: curl -X POST https://api.runvouch.com/v1/settings/test-alert -H "X-API-Key: $RUNVOUCH_KEY". The next missed, failed or empty run reaches the same channels.

Also running cloud functions?

Supabase (pg_cron + Edge Functions). AWS Lambda + EventBridge Scheduler. Google Cloud Scheduler. Azure Functions timer trigger. The full list of runtimes is on integrations, and the same agent can report from several of them: one job, one cadence, wherever it happens to run.

Need a key? Get a free key · Stuck? contact