Vapora/provisioning/schemas/platform/configs/README.md

Platform Configs

Composed Nickel configurations ready for export.

Configuration Files

Composed configurations for VAPORA with different deployment modes:

• vapora-solo.ncl - Solo mode (development)
• vapora-multiuser.ncl - Multiuser mode (team)
• vapora-enterprise.ncl - Enterprise mode (production)
• main.ncl - Entry point for exporting all configs

Each file combines:

1. Schema - Structure definition from vapora/main.ncl
2. Common Defaults - Base values for all modes
3. Mode Defaults - Mode-specific overrides (solo/multiuser/enterprise)
4. User customizations - Optional overrides (commented examples)

Composition Pattern

VAPORA Schema (vapora/main.ncl)
    ↓
Platform Common Defaults (platform/defaults/common/)
    ↓
Mode-Specific Defaults (platform/defaults/deployment/{mode}.ncl)
    ↓
User Customizations (optional)
    ↓
Final Configuration

Configuration Details

Solo Mode (vapora-solo.ncl)

Best for: Local development, testing, PoCs

Defaults:

• Host: 127.0.0.1 (localhost only)
• Backend: 2 workers, file-based database
• Agents: 3 max instances, no NATS
• Router: Cost tracking disabled
• Security: JWT only (no TLS, no MFA)

Customization examples:

# Enable debugging
monitoring.log_level = "debug",

# Change port
backend.port = 9001,

# Enable Ollama
llm_router.providers.ollama_enabled = true,

Multiuser Mode (vapora-multiuser.ncl)

Best for: Team development, staging, internal deployments

Defaults:

• Host: 0.0.0.0 (network accessible)
• Backend: 4 workers, remote SurrealDB
• Agents: 10 max instances, NATS enabled
• Router: Cost tracking enabled (per-role budgets)
• Security: TLS + MFA + audit logging
• Knowledge graph: 30-day retention

Customization examples:

# Set external domain
frontend.api_url = "https://api.vapora.internal:8001",

# Adjust team budgets
llm_router.budget_enforcement.role_limits = {
  architect_cents = 750000,    # $7500/month
  developer_cents = 500000,    # $5000/month
},

# Enable additional providers
providers.openai_enabled = true,
providers.gemini_enabled = true,

Enterprise Mode (vapora-enterprise.ncl)

Best for: Production deployments, large organizations, HA

Defaults:

• Host: 0.0.0.0 (clustered)
• Backend: 8 workers, 2000 max connections
• Agents: 50 instances, NATS cluster
• Router: All providers enabled, cost optimization
• Database: SurrealDB cluster, 100 pool size
• Security: TLS enforced, MFA required, audit enabled
• Observability: Prometheus, OpenTelemetry, tracing
• Knowledge graph: 90-day retention
• Backup: Every 6 hours

Customization examples:

# Set production domain
frontend.api_url = "https://api.vapora.production.com",

# All providers with custom Ollama endpoint
ollama_url = "http://ollama-cluster.production:11434",

# Aggressive cost control
llm_router.budget_enforcement.near_threshold_percent = 70,

# Extended learning window
agents.learning.recency_window_days = 30,

Usage Patterns

1. Export Solo to JSON

nickel export schemas/platform/configs/vapora-solo.ncl > vapora-solo.json

2. Export Multiuser to JSON

nickel export schemas/platform/configs/vapora-multiuser.ncl > vapora-multiuser.json

3. Export Enterprise to JSON

nickel export schemas/platform/configs/vapora-enterprise.ncl > vapora-enterprise.json

4. Export with User Customizations

Create a custom config file that imports and customizes:

# custom-vapora.ncl
let helpers = import "schemas/platform/common/helpers.ncl" in
let schema = import "schemas/vapora/main.ncl" in
let defaults = import "schemas/platform/defaults/deployment/multiuser.ncl" in

helpers.compose_config schema defaults {
  backend.port = 9001,
  llm_router.providers.ollama_enabled = true,
  monitoring.log_level = "debug",
}

Export:

nickel export custom-vapora.ncl > vapora-custom.json

5. Generate TOML from JSON

# Export to JSON, then convert via template
nickel export schemas/platform/configs/vapora-solo.ncl | \
  jinja2 schemas/platform/templates/configs/toml.j2 > vapora.toml

6. Generate Docker Compose

# Generate from multiuser config
nickel export schemas/platform/configs/vapora-multiuser.ncl | \
  jinja2 schemas/platform/templates/docker-compose/docker-compose.yaml.j2 > docker-compose.yml

# Deploy
docker compose up -d

7. Generate Kubernetes ConfigMap

# Generate from enterprise config
nickel export schemas/platform/configs/vapora-enterprise.ncl | \
  jinja2 schemas/platform/templates/kubernetes/configmap.yaml.j2 > configmap.yaml

# Deploy
kubectl apply -f configmap.yaml

Exporting All Configurations

Use the main entry point to export all modes:

nickel export schemas/platform/configs/main.ncl > all-configs.json

This generates:

{
  "solo": { ... },
  "multiuser": { ... },
  "enterprise": { ... }
}

Composition Details

Each config uses the helper functions to compose:

let helpers = import "../common/helpers.ncl" in

helpers.compose_config schema defaults_mode user_customizations

This merges in order:

1. schema - Defines structure and types
2. defaults_mode - Overrides with mode-specific values
3. User customizations - Final overrides

References

• Parent: ../README.md
• VAPORA schema: ../../vapora/README.md
• Platform helpers: ../common/helpers.ncl
• Platform defaults: ../defaults/README.md
• Platform values: ../values/README.md
• Templates: ../templates/README.md
• Constraints: ../constraints/README.md
• Validators: ../validators/README.md