hatch3r-reliability-verify

What problem does it solve?

It prevents agent-produced services from being marked “done” without production-grade reliability controls, so failures surface in time and rollouts don’t silently harm users.

Core Features & Use Cases

• Reliability gate with 9 machine-checkable “done” criteria: validates SLO presence, kill switch, outbound timeouts, retries with decorrelated jitter, health probes, graceful shutdown, runbook URLs, staged rollout configuration, and blast-radius documentation.
• Operational readiness for production: ensures alerts link to runbooks with required sections and that progressive delivery is configured to roll back on SLO burn-rate.
• Release-prep artifact for downstream use: writes a PR verdict plus a JSON audit output for release workflow automation.
• Use case: Before shipping a new microservice, verify that the repo’s SLO/alerts, Kubernetes probes, rollout strategy, and runbooks are complete so on-call can triage quickly and rollback confidently.

Quick Start

Run the reliability verification gate by invoking it on a PR that touches service code or Kubernetes/alert configuration, so it evaluates all nine production-readiness gates and blocks the merge on any failure.