How Actions work in OpsLevel

Why Self-Service Matters Now More Than Ever

Engineers move faster when they don’t need a ticket to deploy code, spin up infrastructure, or page on-call. But unchecked autonomy can break standards and introduce risk. OpsLevel Actions give teams the best of both worlds: one-click task automation with guardrails baked in.

See an end-to-end setup of Actions in this on-demand demo video

What You Can Automate in OpsLevel

Below are core tasks customers turn into Actions. Each one replaces a multi-step, ticket-driven flow with a single, auditable click.

• Run a CI/CD pipeline – trigger a deploy straight from a service page.
• Spin up AWS infrastructure – create an S3 bucket or EC2 instance pre-tagged to standards.
• Create a new service – scaffold repo, infra, and ownership via a Service Template.
• Kick off an incident – page on-call in PagerDuty or open a Sev-1 in ServiceNow.
• Send a Slack alert – notify #dev-prod when a feature flag flips.
• Rotate secrets or configs – invoke a HashiCorp Vault rotation workflow.

If a system can be reached by webhook or an integrated app, you can turn it into an Action.

How Actions Work: a Three-Step Model

1. Trigger – today a user clicks Run Action (event-based triggers are on the roadmap).
2. Inputs – dynamic form fields defined in YAML:
   • Catalog-aware dropdowns (services, teams, environments, domains)
   • Conditional visibility and default expressions
   • Multi-select support
3. Execution – OpsLevel formats the payload and calls the target system, then streams asynchronous status updates back to the UI.

Prefer code over clicks? Define an Action Template in YAML and commit it to Git. OpsLevel renders the same UI automatically.

New Capabilities for 2025

• Catalog-aware dropdowns pull live data—no hard-coding values.
• Conditional logic shows only the fields a user needs.
• Multi-select fields pass multiple environments, tags, or owners in one go.
• Embedded widgets surface Actions on service, team, or dashboard pages.
• Asynchronous status updates show progress without refreshing.
• Built-in approvals require a user or team sign-off before execution.

Service Templates: Specialized Actions for New Services

Service Templates run on the same Action framework but come pre-wired to:

1. Generate a repository from a starter kit.
2. Provision infrastructure (Terraform, CloudFormation, Pulumi, etc.).
3. Register the new service in OpsLevel with owners, tier, and metadata.

Result: a production-ready service in minutes, every time.

Example Walk-Through: “Lock Down Deploys in Three Minutes”

1. Create an Action → Settings > Manage Actions > New Action.
2. Define form inputs – Service (catalog dropdown), Environment (dev / staging / prod), Reason (required text).
3. Configure the webhook – point to your deployment API and map form values.
4. Add an approval – require anyone on Platform-Team when environment = prod.
5. Publish – the Action now appears as a button on every service page that meets the filter.

Total setup time: ~3 minutes. Dev friction removed: hours every week.

Use Cases → Outcomes

Below are six common automations and the value customers report after implementation:

1. Run a CI/CD Pipeline

• Pain Before: engineers bounced between Jenkins, GitHub, and Slack to deploy.
• Outcome After: one click from the service page. FM Global cut deploy prep from days to 20 minutes.

2. Spin Up AWS Infrastructure

• Pain Before: Terraform scripts lived in a private repo accessible only to platform engineers.
• Outcome After: developers select environment and tags; Action triggers Terraform Cloud. First-time infra requests dropped from ~2 hours of tickets to under five minutes.

3. Create a New Service

• Pain Before: ad-hoc cookie-cutter plus manual repo setup led to standards drift.
• Outcome After: Service Template scaffolds everything. BigPanda created production-ready services ten times faster.

4. Kick Off an Incident

• Pain Before: responders followed a six-step runbook to page on-call and update Slack.
• Outcome After: a single Action opens PagerDuty and posts to #incident, shaving precious minutes off MTTR.

5. Send a Slack Alert

• Pain Before: inconsistent wording and forgotten channels.
• Outcome After: pre-formatted Action ensures the right people see the right context, increasing acknowledged alerts by more than 30 percent.

6. Rotate Secrets or Configs

• Pain Before: quarterly vault rotation was a manual, 45-minute slog.
• Outcome After: Action rotates secrets in under two minutes with a full audit trail.

Takeaway: every Action shrinks lead time and slashes cognitive load—without compromising control.

Best Practices & Guardrails

• Use team-based approvals for production-critical Actions to combine speed with accountability.
• Scope visibility with catalog tags so engineers see only the Actions relevant to their components.
• Promote high-value Actions via widgets so discovery happens naturally during daily work.
• Version-control Action Templates to enable peer review and maintain history for every change.

Ready to Simplify Your Workflows?

If you aren't already a customer, book a call with our team to learn more and see Actions firsthand, or view our docs to see how it works in depth.