phase-4-trust-boundaries

name	phase-4-trust-boundaries
description	Phase 4 Trust Boundary Analysis guide. Use when defining trust zones, crossing points, and security boundaries between architecture nodes.

Phase 4: Trust Boundary Analysis

Objective

Identify where trust levels change in the system. Every crossing point is a potential attack surface that needs security controls.

Trust-boundary state starts empty. Define zones, crossing points, and boundaries for the system under review before advancing.

Concepts

• Trust Zone: A region where components and data stores share the same trust level
• Crossing Point: Where data flows between zones (requires authentication/authorization)
• Trust Boundary: The security perimeter with specific controls

Tools Reference

Start with manage_trust_boundaries(action="describe", section=SECTION) for the exact contract.

manage_trust_boundaries(action="add", section="zones", values=ZONE)

Parameter	Values
trust_level	Untrusted, Low, Medium, High, Full

manage_trust_boundaries(action="link", section="zones", values=LINK)

Assign an architecture node to exactly one trust zone. The link fields are zone_id and node_id; a node may be either a component or a data store.

manage_trust_boundaries(action="add", section="crossing_points", values=CROSSING)

Parameter	Values
authentication_method	Password, Multi-factor, Certificate, Token, Biometric, API Key, IAM Role, OAuth, None, Other
authorization_method	Role-based, Attribute-based, Discretionary, Mandatory, Policy-based, Rule-based, None, Other

manage_trust_boundaries(action="link", section="crossing_points", values=LINK)

Map existing connections to crossing points.

manage_trust_boundaries(action="add", section="boundaries", values=BOUNDARY)

Parameter	Values
type	Network, Process, Physical, Container, Virtual Machine, Account, Other
crossing_point_ids	List of crossing point IDs
controls	List of security control names (strings)

Other Phase 4 Tools

• manage_trust_boundaries(action="list", section="all") -- Review the complete model
• manage_trust_boundaries(action="detection_plan", section="all") -- AI-powered boundary detection
• manage_trust_boundaries(action="analysis_plan", section="all") -- Security analysis guidance
• CRUD actions accept values for one record and items for a batch

Common Trust Zone Patterns

Web Application

Zone	Trust Level	Architecture Nodes
Internet	Untrusted	End users, external APIs
DMZ	Low	Load balancer, CDN, WAF
Application	Medium	App servers, API services
Data	High	Databases, caches, queues
Admin	Full	Admin consoles, CI/CD

Microservices

Zone	Trust Level	Architecture Nodes
Public	Untrusted	API Gateway, public endpoints
Service Mesh	Medium	Internal microservices
Data Layer	High	Databases, object stores
Secrets	Full	KMS, secret managers

Workflow

1. Call manage_workflow(action="guidance", phase="4")
2. Call manage_trust_boundaries(action="detection_plan", section="all") for AI-guided detection
3. Create trust zones based on security domains
4. Assign every component and data store to a zone with action="link", section="zones"
5. Define crossing points where data flows between zones
6. Map connections to crossing points with action="link", section="crossing_points"
7. Create trust boundaries with security controls

Completion Criteria

• All trust zones defined
• Every architecture node assigned to exactly one zone
• Every inter-zone connection mapped to exactly one matching crossing point
• Every crossing point assigned to a trust boundary
• No crossing points created when all communication remains within one zone
• Call manage_workflow(action="advance") to proceed to Phase 5

Common Pitfalls

• Putting everything in one trust zone
• Missing the boundary between internal services and external APIs
• Not specifying authentication/authorization at crossing points
• Forgetting admin/management plane boundaries