sre-engineer

Start by identifying measurable SLIs such as latency, error rate, and saturation that reflect real user experience, then set targets based on the service's actual traffic and dependency profile. This Skill walks through that workflow and flags any target not grounded in observable signals.

Define burn-rate thresholds that trigger concrete actions, such as a feature freeze or risk review, and document how the policy is enforced rather than just written down. The Skill recommends the smallest enforceable policy matched to your reliability goals.

An SLI is a measurable indicator like request latency, an SLO is the target threshold for that indicator, and the error budget is the tolerated amount of SLO violation over a period. The Skill uses all three together to frame reliability trade-offs against feature velocity.

No. Live incident response, containment, and postmortem authorship are explicitly deferred to incident-handling work. This Skill covers proactive reliability engineering that reduces the likelihood and blast radius of future incidents.

Identify manually repeated operational tasks, then build automation or self-service paths that remove the work rather than relocate it to another team. The Skill also audits alert quality so on-call engineers only get paged for actionable issues.

Claims about actual latency or burn rate cannot be validated without live telemetry or load testing. The Skill flags such claims as assumptions and lists what needs production measurement to confirm.