sre-oncall skill

Overview

This skill is the SRE Oncall Agent that handles incident response, escalation, recovery, and postmortem documentation. It acts as an operational companion for on-call engineers, coordinating actions and producing required artifacts. The agent focuses on clear handoffs, traceable incident logs, and actionable postmortems to reduce recurrence.

How this skill works

The agent inspects incoming alerts and incident metadata, categorizes severity, and triggers escalation paths when needed. It coordinates recovery steps, records timeline events into the incident log, and compiles a structured postmortem after resolution. All artifacts are stored in a predictable location for review and follow-up.

When to use it

• When an alert indicates a production service outage or degradation
• During on-call shifts to coordinate triage and escalation
• To document recovery actions and create a reproducible timeline
• After incident resolution to create a postmortem and lessons learned
• When enforcing escalation policy and tracking accountability

Best practices

• Triage quickly: assign severity and initial owner within the first 5–10 minutes
• Log every action and decision in the incident timeline for traceability
• Follow predefined escalation policies and update them when gaps appear
• Keep postmortems factual, blameless, and focused on systemic fixes
• Store incident logs and postmortems in the designated directories for easy retrieval

Example use cases

• A sudden database failure where the agent triggers escalation and documents recovery steps
• A latency spike in an API where the agent coordinates mitigation and records timeline events
• After deploying a fix, the agent compiles a postmortem outlining root cause and preventative tasks
• On-call rotation handoffs where the agent provides a concise incident summary and status
• Periodic audits to review incident logs and identify recurring failure modes

FAQ