Fleet 2.9.1 is out.See what's new →
FleetFleet
Agent templateDevOps

Site Reliability Engineer AI Agent (Template)

An SRE agent monitors service health, responds to on-call signals, and works to reduce toil through automation. In a fleet, it receives incident signals from your alerting system, investigates the likely cause, and either applies the appropriate runbook steps or escalates with a clear summary.

SRE work requires deep familiarity with your specific service topology, SLO definitions, and runbooks. A role-specific prompt encodes these so the agent can navigate your observability tooling and apply the correct runbook rather than starting from scratch on each incident.

What this agent owns

  • Respond to incident signals with initial triage and impact assessment
  • Investigate error spikes, latency regressions, and availability drops using observability tooling
  • Apply runbook steps and document what was tried and what worked
  • Write post-incident summaries with root cause analysis and follow-up action items
  • Identify and automate repetitive operational tasks to reduce manual toil

Recommended model: Claude Opus

Incident diagnosis often requires reasoning across multiple correlated signals; Opus handles ambiguous multi-signal analysis more reliably.

Example tasks

  • Investigate a 5xx error spike and identify the upstream service causing it
  • Write a runbook for recovering from a database connection pool exhaustion event
  • Automate a manual log rotation task that runs weekly
  • Draft a post-incident report for a 45-minute API outage
# create an agent from this template, then start it
$ fleet agent create --name sre--vendor claude-code --template <template-name>
$ fleet agent start sre

Find the exact template name with fleet template list.

Run this agent in your fleet

Jail Fleet-launched Linux agents, then run saved workflows with review, approvals, and an audit trail. Prove it at /security/#containment.