Principle IV: Hybrid Deployment
Source:
.specify/memory/constitution.md
Overview
Enterprise agents must be securely deployed across hybrid cloud and AI systems with resilience and compliance built-in. Agents must integrate seamlessly with existing IT ecosystems while respecting enterprise controls, data sovereignty, and regulatory constraints.
Hybrid deployment ensures agents operate where data and workloads reside, with security and compliance baked into infrastructure rather than bolted on afterward.
Non-Negotiable Rules
| Rule | Description |
|---|---|
| Multi-Environment | Support from air-gapped to hyperscaler deployments |
| Model Portfolio | Frontier, fine-tuned, and edge models supported |
| Compliance-Aware | Infrastructure aligned with HIPAA, SOX, GDPR requirements |
| Sandboxing | Lightweight virtualization mandatory for all agent execution |
| Progressive Rollout | Feature flags, canary deployments, and rollback capabilities |
| Kill-Switches | Circuit breakers with graceful degradation |
Deployment Target Selection
Enterprise Feature
CDK/Terraform deployment patterns, 3-tier testing infrastructure details, sandboxing configurations, and checkpoint evidence requirements are available to enterprise consumers. Contact us for access.
Related Agents
- cloud-architect — Primary agent for deployment design
- infrastructure-engineer — CDK/Terraform execution
- kubernetes-engineer — K3D/K3S cluster management