Select version

Explanation

Explanation
Activation and PackageInstance lifecycle
Availability levels and derived pod spread
Compliance evidence as build output
Cross-resource validation at cluster evaluation time
DAG-driven deployment: scheduler, readiness, pruning, drift, blast radius
Dependency resolution and why explicit deps matter
Deploy receipts and reproducible provenance
Deployment lifecycle as the value center: diff, health, reconcile, rollback, GC
Escape hatches and explicit unsafe grants
Flavors, roles, and substrate capabilities
Helm integration philosophy: normalization, patching, DAG reconstruction, native conversion path
How Kix compares with Helm, Kustomize, CDK8s, GitOps controllers, and raw YAML
How Kix maps package resources to Kubernetes manifests
How string context carries identity and typed facts
How typed endpoint facts drive ordering and network policy
Identity hashes and the content-addressing model
Imports vs managed packages
Kix as a strict Kubernetes superset
Kubernetes readiness and post-deploy verification solve different problems
Network policy derivation from the graph
Nix store as Kix IR
Package vs instance
Roles as linker symbols
Scorecard levels: manifest, package, namespace, cluster
Server-side apply, field ownership, and drift
The dependency graph as the model
The Kix compiler pipeline
The three personas: app maintainer, platform team, cluster operator
Why deploy inputs are committed
Why Kix uses Nix modules/options for package config

Select version

Explanation

The design decisions and mental models behind Kix.

On this page