Docs › Prefect deployment monitoring with RunVouch
Prefect deployment monitoring with RunVouch
Prefect deployment monitoring for the run that stays Late: a Prefect deployment with a schedule creates flow runs; a worker on the right work pool executes them. When the worker is gone, runs pile up as Late and Scheduled, and the flow's own failure hooks never fire because nothing failed.
How it runs on Prefect
Around the flow: wrap the process that runs the flow (rv run ... -- python flow.py or prefect deployment run), or use the Python client inside the flow function. Prefect's own retries and logging stay untouched.
Store the key
A Prefect Secret block (Secret.load("runvouch-key")) or an environment variable on the work pool's job template.
Wrap the job
from prefect import flow, task
from prefect.blocks.system import Secret
import os, runvouch
@flow(name="nightly-report")
def nightly_report():
os.environ["RUNVOUCH_KEY"] = Secret.load("runvouch-key").get()
with runvouch.vouch("nightly-report", evidence=lambda: {"report": os.path.exists("out/report.html")}) as run:
rows = extract()
run.tool("llm.summarise", {"rows": len(rows)}, cost=0.03)
write_report(rows)
# prefect deploy ... --cron "0 2 * * *"rv fails open: if RunVouch is unreachable the job still runs and you get one warning line.
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 Prefect
- Late runs are visible in the UI and in automations you have to create; RunVouch's MISSED needs neither.
- A flow that completes with an empty result is Completed; evidence separates it from a good run.
- Prefect task retries multiply LLM calls; the retry-storm detector sees identical tool inputs across retries.
What Prefect does not tell you
Prefect shows Late runs in the UI and can notify through an Automation, if you create one with the right trigger and a notification block. Without that, a work pool with no worker, a worker whose process died, or a deployment whose schedule was set inactive, produces a growing list of Scheduled runs and no message. A flow that completes with an empty result is Completed.
What RunVouch detects
| Alert | What it means here |
|---|---|
| MISSED | cadence plus grace passed and no run started |
| FAILED | non-zero exit or a reported failure, with the stderr excerpt |
| NO_EVIDENCE | the run said ok but the file, URL or assertion you required is missing |
| STALLED | a run started and never ended within max runtime |
| RETRY_STORM | the same tool called with identical input many times in one run |
| BUDGET_RUN / BUDGET_DAY | cost cap crossed; the agent is paused until you resume it |
| DRIFT | duration or output size far off its 7-run baseline |
Set up in two minutes
- Get a free key (3 agents, no card) and store it where this page says.
- wrap the job: copy the snippet above into the scheduled job.
- 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. - 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 orchestrators?
Dagster. Celery beat. APScheduler (in-process). Temporal Schedules. 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