NthLayer Reliability Check

Version updated for https://github.com/rsionnach/nthlayer to version v0.1.0a19.

  • This action is used across all versions by 0 repositories.

Action Type

This is a Composite action.

Go to the GitHub Marketplace to find the latest changes.

Action Summary

NthLayer is a reliability-focused GitHub Action that enables platform teams to define reliability requirements as code, validate SLOs against dependencies, and automate tasks like drift detection, deployment gating, and blast radius analysis. It helps teams proactively identify and address reliability issues, ensuring SLO feasibility, monitoring error budgets, and generating artifacts such as dashboards and alerts. By shifting reliability checks earlier in the development workflow, it reduces incidents and improves service resilience.

Release notes

What’s New

nthlayer simulate — Monte Carlo SLO Simulation

Predict the probability of meeting your SLA before you ship. Model dependency chains, run 10,000 simulated quarters, find your weakest link.

nthlayer simulate service.yaml --manifests-dir ./manifests/

Key features:

  • Models failure timelines using exponential MTBF and lognormal MTTR distributions
  • Cascades failures through dependency graphs (critical and non-critical)
  • Produces P(meeting SLA), percentile distributions, error budget forecasts
  • What-if scenarios: redundant, improve, remove, degrade — explore architectural changes before implementing them
  • JSON output for CI/CD integration with --min-p-sla gate
  • Pure transport (ZFC) — no model calls, no API costs

Also included

  • Configurable alert for duration — per-severity override for Prometheus alert for durations
  • Error budget policy DSL — declarative on_exhausted behaviors (freeze_deploys, require_approval, notify)

Stats

  • 46 new tests, 3,844 total passing
  • 11 new files, 2 modified
  • Full documentation

Full Changelog: https://github.com/rsionnach/nthlayer/compare/v0.1.0a18...v0.1.0a19