Generates SLOs, burn-rate alerts, chaos experiments, runbooks, and capacity plans for production systems.
npx skills add https://github.com/sundarshahi/drydock --skill sre-sundarshahi
Please help me install this Agent Skill. Skill: sre Source: https://github.com/sundarshahi/drydock/tree/main/skills/sre Command: npx skills add https://github.com/sundarshahi/drydock --skill sre-sundarshahi
make production-ready CI gate, and hand on-call engineers runbooks with exact kubectl and PromQL commands. ## Quick Start Ask the drydock orchestrator to run the sre skill to define SLOs, alerts, runbooks, and a production-readiness gate for the deployed services.High-intent search queries and answers about installing and using this skill.
The Skill writes sli-definitions.yaml and burn-rate-alerts.yaml using the Google SRE workbook multi-window method, with PromQL expressions restricted to metric names declared in the project's observability contract. Latency thresholds are read from performance-budget.yaml rather than hardcoded.
DevOps owns infrastructure provisioning, CI/CD pipelines, and base monitoring installation. SRE owns SLO definitions, error budgets, service-level alerting, incident response processes, chaos engineering, and capacity planning, and DevOps implements the thresholds SRE defines.
Yes. The emitted production-ready.sh script and GitHub Actions job exit non-zero on metric-name drift, invalid alert rules, SLO-versus-budget mismatches, unjustified factor-compliance failures, missing kill-switch keys, or failing tests. Overrides require a logged accepted-with-justification entry.
Each Chaos Mesh scenario defines a measurable steady-state hypothesis using contract metric names before injecting failures like pod kills, network partitions, or resource pressure. The game-day playbook includes explicit abort criteria and progresses from low-risk to high-risk experiments.
Runbooks contain exact kubectl and PromQL commands with real pod labels and metric names, decision trees instead of prose, and an emergency mitigation step that flips an ops kill-switch flag from config/feature-flags.yaml before attempting a rollback.
hainamchungAutomate multi-cloud infrastructure provisioning and DevOps workflows across AWS, Azure, and GCP.
Camilo8902Generate SLO-based alerts and dashboards with charts, thresholds, and runbook links.
saolalabDefine SLIs, SLOs, error budgets, alerts, and dashboards for service reliability.
nagisanzeninGenerates SLOs, chaos experiments, incident runbooks, and capacity plans for production systems.