Role-Segmented Knowledge Architecture
Comprehensive technical specifications, machine-readable schemas, architectural decisions, and operational runbooks for every platform participant.
Core Architectural Guarantee: Decoupled Intelligence
VaultHost strictly isolates AI reasoning planes from the execution plane. AI models synthesize structured ActionIntents but never directly touch databases, hypervisors, or operating systems. All state changes pass through the Action Gateway.
Select Your Operational Persona
Documentation tailored precisely to your authorization context, tooling, and system plane.
AI Agents & LLM Runtimes
Zero-ambiguity schemas, ActionIntent specifications, MCP tool definitions, and risk-tier taxonomies designed for autonomous reasoning engines.
Developers & Platform Engineers
Architecture Decision Records (ADRs), Post-Quantum Cryptography integration, REST/JSON-RPC specs, and language SDKs.
Security & Staff Administrators
Operational procedures for managing Tier 3 Dual-Authorization Approval Queues, immutable audit verification, and emergency rollbacks.
Enterprise Customers
SLA commitments (99.999%), SOC 2 Type II and ISO 27001 compliance, zero-egress billing guarantees, and tenant isolation.
Partners & SaaS Vendors
Standards for integrating SaaS tools via Model Context Protocol (MCP) using delegated user OAuth tokens. Zero web-scraping policy.
API Reference & OpenAPI 3.1
Complete REST & JSON-RPC API reference, downloadable OpenAPI 3.1 specification, and multi-language client code generator.
Docs as Code Infrastructure
All VaultHost documentation lives directly in our repository under the /docs/ directory. We maintain Architecture Decision Records (ADRs), Runbooks, and auto-generated API references alongside our source code.
Model Context Protocol (MCP) Server Live
Autonomous agents can query POST /api/action-gateway/mcp using JSON-RPC 2.0 with tools/list to dynamically discover all cloud capabilities with embedded @risk_tier metadata.