​

SpineWorkspace — Human Workbench Information Architecture

​

1. Purpose

• a departmental database;
• a second source of truth;
• an integration console;
• an AI chat shell.

​

The Tier-0 Principle

Business Context is the semantic foundation of SpineWorkspace. Without Business Context, the Workbench has nothing to present. With it, the Workbench already knows what matters, what changed, and what needs attention.

​

2. IA model

SPINEWORKSPACE HUMAN WORKBENCH
│
├── ORIENTATION
│   ├── Home
│   ├── Today
│   ├── My Work
│   ├── My Queue
│   ├── Tasks
│   └── Recent Work
│
├── COMMUNICATION
│   ├── Email
│   ├── Chat
│   ├── Calendar
│   ├── Meetings
│   ├── Calls
│   └── Shared Documents
│
├── ORGANIZATIONAL MEMORY
│   ├── Knowledge
│   ├── Decisions
│   ├── Documents
│   ├── Policies
│   ├── Processes
│   ├── Meeting Memory
│   └── Work History
│
├── WORK
│   ├── Active
│   ├── Waiting
│   ├── Needs Decision
│   ├── Needs Approval
│   ├── Delegated
│   ├── Blocked
│   ├── Escalated
│   └── Completed
│
├── BUSINESS DOMAINS
│   ├── Sales
│   ├── Marketing
│   ├── Account Success
│   ├── Technical Account Success
│   ├── Finance
│   ├── RevOps
│   └── Organization-specific domains
│
└── WORK SURFACE
    ├── Tables
    ├── Boards
    ├── Canvas
    ├── Timeline
    ├── Entity View
    ├── Workflow View
    └── Document / Knowledge View

​

3. Main interaction model

USER SEES WORK
 ↓
CURRENT STATE
 ↓
CONTEXT
 ↓
WHAT CHANGED
 ↓
WHAT NEEDS ATTENTION
 ↓
WHAT THE TWIN RECOMMENDS
 ↓
WHAT THE HUMAN MUST DECIDE
 ↓
WHAT HAPPENS NEXT

​

4. Adaptive projection

Organization
× Role
× Domain
× Responsibility
× Authority
× Permissions
× Current Work
× Business Context
× Policy
× User Preferences
        ↓
Projection
        ↓
Human Workbench

​

5. Workbench operating semantics

• Store — retain governed information in the appropriate context/memory scope.
• Ask — ask the Twin for context, analysis or recommendation.
• Decide — make a human decision.
• Approve — authorize a governed consequential action.
• Reject — refuse a proposal.
• Redirect — change the objective/constraints.
• Consult — communicate with another human.
• Delegate — transfer Work through the governed assignment layer.
• Verify — inspect the actual outcome.
• Continue — advance the Work after reconciliation.

​

6. Context Drawer / active context

Entity
Current State
Related Work
Relevant History
Communications
Decisions
Evidence
Policies
Risks
Dependencies
Twin Context

​

7. Adaptive boundary

• view order;
• density;
• prefetch;
• notification ranking;
• capability placement;
• personal shortcuts.

• canonical truth;
• shared schema;
• authority;
• permissions;
• policies;
• governance thresholds;
• organizational memory publication;
• external side effects.

​

8. Human/Twin surface relationship

                 SHARED WORK
                     │
          ┌──────────┴──────────┐
          │                     │
   HUMAN WORKBENCH        TWIN WORKBENCH
          │                     │
     Human judgment        AI coordination
     Human authority       Delegation
     Relationship          Execution
     Approval              Monitoring
     Accountability        Reporting

​

9. Acceptance criteria

1. it projects shared Spine state rather than creating a departmental copy;
2. active context is assembled around Work;
3. role/domain/responsibility/authority affect projection;
4. AI receives sandbox context rather than canonical access;
5. consequential actions require the governance path;
6. communications remain linked to Work;
7. source-system outcomes return through reconciliation;
8. published Organizational Memory is clearly distinct from pending candidates.