Configuration Schema (`krewire.yaml`)

Krewire operates under the One Config rule: all settings across all eight workloads are declared in a single file—krewire.yaml.

There is no separate ssg.yaml or tool-specific configuration file. Configuration is typed, validated against schema rules (packages/validation), and loaded using packages/config.

Canonical Schema Overview

# 1. Project Identification (Required)
project:
  name: "my-service"            # Unique project name (DNS-safe identifier)
  kind: "service"               # Workload kind: app, cli, site, book, worker, service, infra, runtime
  version: "0.1.0"              # Semantic version (SemVer)
  author: "Krewire Team"        # Author or organization name

# 2. Source Content Input (For site and book workloads)
input: "content"                # Root directory for Markdown manuscripts or docs

# 3. Build & Compilation Pipeline
build:
  output: ".krewire/build"       # Output directory for compiled assets or binaries
  base: "/"                     # Base URL path for static asset resolution
  tags: ["netgo"]               # Custom Go build tags
  target: "binary"              # Build target override: binary, site, book, wasm, infra

# 4. Local Development Server
dev:
  port: "8080"                  # HTTP port for local development server
  host: "127.0.0.1"             # Host binding address
  watch:                        # Custom directories to monitor for file changes
    - "pages"
    - "components"
    - "layouts"
    - "assets"

# 5. Workload-Specific Settings (Optional depending on kind)

# Background Worker Settings (kind: worker)
worker:
  concurrency: 10               # Maximum concurrent workers
  queue: "memory"               # Queue connection URL (memory, redis://, nats://)
  dead_letter_queue: "failed"   # Queue for failed messages

# Microservice Settings (kind: service)
service:
  port: "8080"
  discovery:
    provider: "kubernetes"      # consul, etcd, kubernetes, static
    service_name: "my-service"
  rate_limit:
    requests_per_second: 500
    burst: 100

# Cloud Infrastructure Settings (kind: infra)
infra:
  provider: "aws"               # aws, gcp, kubernetes
  region: "ap-southeast-1"
  state:
    backend: "s3"
    bucket: "krewire-infra-state"

Schema Sections Reference

`project`

Key	Type	Required	Description
`name`	String	Yes	Name of the project or service.
`kind`	String	Yes	Workload kind (`app`, `cli`, `site`, `book`, `worker`, `service`, `infra`, `runtime`).
`version`	String	Yes	Current semantic release version string (`vX.Y.Z`).
`author`	String	No	Author, team, or maintainer name.

`build`

Key	Type	Default	Description
`output`	String	`.krewire/build`	Directory path where build artifacts are written.
`base`	String	`/`	Base URL prefix for static links and assets.
`tags`	Array	`[]`	Go build tags passed to the compiler.
`target`	String	auto	Explicit target compiler dispatch override.

`dev`

Key	Type	Default	Description
`port`	String / Int	`8080`	Local port for the dev server (can be overridden by `KIW_PORT`).
`host`	String	`127.0.0.1`	Network interface address to bind to.
`watch`	Array	auto	Additional filesystem paths to observe for live reload.

Environment Overlays & Dotenv (`.env`)

Krewire automatically loads .env files from the project root during local development:

# .env
KIW_PORT=3000
DATABASE_URL=postgres://user:pass@localhost:5432/db

Values declared in environment variables take precedence over settings in krewire.yaml.

Configuration Schema (krewire.yaml)