Configuration

A.R.C. uses a declarative configuration model. You define your desired state in arc.yaml, and the CLI generates the infrastructure.

arc.yaml Reference

# Version of the manifest schema
version: "1.0.0"

# Workspace tier (determines base services)
tier: super-saiyan  # Default tier

# Feature flags to enable/disable capabilities
features:
  voice: false         # Scarlett, Daredevil (Voice AI)
  security: false      # J.A.R.V.I.S., Nick Fury, RoboCop (Auth & Secrets)
  observability: false # Dr. House, Watson, Friday (Monitoring)
  chaos: false         # T-800 (Chaos Engineering)

# Service-specific overrides
services:
  arc-gateway:
    config:
      port: 8080
      enable_tls: false

# Environment variables for all services
environment:
  LOG_LEVEL: "info"
  ENVIRONMENT: "development"

Feature Flags

Feature flags control which optional services are included in your platform.

Voice (features.voice)

Enables real-time voice AI capabilities.

Services added: Daredevil (LiveKit), Scarlett (Voice Agent), Sentry (Ingress), Scribe (Egress) Resource impact: +2GB RAM, +1 CPU core When to enable: Building conversational AI, voice assistants, real-time collaboration

Security (features.security)

Enables production-grade identity and secrets management.

Services added: J.A.R.V.I.S. (Kratos), Nick Fury (Infisical), RoboCop (Guardrails) Resource impact: +1.5GB RAM, +0.5 CPU core When to enable: Multi-user applications, production deployments, compliance requirements

Observability (features.observability)

Enables the full monitoring stack.

Services added: Black Widow (OTEL), Dr. House (Prometheus), Watson (Loki), Columbo (Jaeger/Tempo), Friday (Grafana), Hermes (Promtail) Resource impact: +4GB RAM, +2 CPU cores When to enable: Production deployments, performance debugging, SLA monitoring

Chaos (features.chaos)

Enables infrastructure resilience testing.

Services added: T-800 (Chaos Mesh) Resource impact: +1GB RAM, +0.5 CPU core When to enable: Production readiness validation, resilience testing

Tiers

Tiers are pre-defined service collections that serve as starting configurations.

SuperSaiyan (Default)

The standard development tier. Includes 13 base infrastructure services plus any feature-flagged services.

Configuration	Services	RAM	CPU
Base only	13	4 GB	2 cores
+ Security	16	6 GB	3 cores
+ Voice	17	8 GB	4 cores
+ Observability	19	12 GB	6 cores
All features	26	16 GB	8 cores

SuperSaiyanBlue (Coming Soon)

Production environments with enhanced observability and Redis Cluster.

UltraInstinct (Coming Soon)

Large-scale deployments with chaos engineering and multi-region support.

Environment Variables (.env)

The .env file contains credentials and configuration for your services:

# Database
POSTGRES_USER=arc
POSTGRES_PASSWORD=your-secure-password
POSTGRES_DB=arc

# Redis
REDIS_PASSWORD=your-redis-password

# Security (if enabled)
SESSION_SECRET=your-session-secret
JWT_SECRET=your-jwt-secret

# Voice AI (if enabled)
OPENAI_API_KEY=sk-your-api-key

Example Configurations

Minimal (AI Agent Only)

version: "1.0.0"
features:
  voice: false
  security: false
  observability: false
  chaos: false

13 services, 4GB RAM minimum.

Voice AI Platform

version: "1.0.0"
features:
  voice: true
  security: false
  observability: false
environment:
  VOICE_MODEL: "whisper-large"

17 services, 8GB RAM minimum.

Production-Ready

version: "1.0.0"
features:
  voice: false
  security: true
  observability: true
services:
  arc-gateway:
    config:
      enable_tls: true
environment:
  LOG_LEVEL: "warn"
  ENVIRONMENT: "production"

22 services, 12GB RAM minimum.

Full Development Stack

version: "1.0.0"
features:
  voice: true
  security: true
  observability: true
  chaos: true
environment:
  LOG_LEVEL: "debug"
  TRACING_SAMPLE_RATE: "1.0"

27 services, 16GB RAM minimum.

Customizing After Initialization

Edit arc.yaml to add or remove features at any time:

# After editing arc.yaml
arc workspace run

To stop running services before changing configuration:

docker compose -f .arc/generated/docker-compose.yml down
arc workspace run

---

Next: Explore the Platform Services for detailed documentation on each service.