Periodic Jobs & Health Notebooks
Scheduled Automation, Zero Drift
Define recurring tasks that run on a schedule — health checks, data syncs, cleanup scripts, compliance audits. Powered by Health Notebooks: multi-step jobs that Argos executes autonomously with full logging and alerting.
Health Notebooks
Multi-step runbooks that execute on a cron schedule. Each step uses your connected tools — no custom scripting required.
📋 Multi-Step Execution
Define jobs as a sequence of steps. Each step runs a tool action on a connector — query a database, check a Kubernetes deployment, call an API, verify a certificate. Steps chain together with full context.
⏰ Cron Scheduling
Standard cron expressions for flexible scheduling — every 5 minutes, hourly, daily at 3am, weekly on Monday. Each execution is logged with duration, status, and full output.
What You Can Schedule
Health Checks
Verify endpoint availability, certificate expiry, DNS resolution, database connectivity every few minutes.
Cleanup & Maintenance
Remove stale resources, rotate secrets, archive old logs, clean up orphaned cloud resources on a schedule.
Reports & Syncs
Generate daily cost reports, sync CMDB inventories, export compliance evidence to Slack or Confluence.
Security Audits
Daily CSPM scans, weekly IAM permission reviews, certificate expiry alerts, vulnerability scan triggers.
Drift Detection
Compare live infrastructure state against Terraform, detect config drift, alert on unauthorized changes.
Capacity Planning
Track resource usage trends, predict capacity needs, alert before you hit limits.
How It Works
Define Steps
Pick a connector + tool action for each step. Chain them together in a notebook.
Set a Schedule
Cron expression or interval. Argos runs it automatically at the right time.
Get Alerted
If a step fails or returns an unexpected result, Argos alerts you via Slack, email, or PagerDuty.
Frequently Asked Questions
What tools can I use in a periodic job?
Any tool from any connected connector — Kubernetes, AWS, GCP, databases, GitLab, Jira, Slack, and 30+ more. Each step picks a connector and an action.
How granular is the scheduling?
Standard cron expressions — from every minute to once a year. You can also trigger jobs manually or via API/webhook.
What happens when a job fails?
Failed steps are logged with full output. Alerts are sent to your configured channels (Slack, email, PagerDuty). You can set retry policies and failure thresholds.
Explore More
Automate the Boring Stuff
Health checks, cleanup, reports, audits — running on schedule so you don't have to.