MWMS AI Employee Version And Canon Sync Framework

Document Type: Framework
Status: Active
Version: v1.0
Authority: HeadOffice
Applies To: All MWMS AI Employees, Brain-specific AI operators, automation agents, and future AI execution systems
Parent: HeadOffice Brain
Last Reviewed: 2026-04-26

Purpose

The MWMS AI Employee Version And Canon Sync Framework defines how AI Employees remain aligned with the latest approved MWMS Canon, frameworks, protocols, templates, and operating standards.

MWMS is an evolving ecosystem.

Brains, frameworks, workflows, and decision systems will continue to improve over time.

Because of this, AI Employees must not be treated as static workers.

They must operate as version-aware system participants.

This framework ensures AI Employees can:

  • identify the current canon source
  • operate from approved framework versions
  • detect when their instructions may be outdated
  • request refresh when source pages change
  • remain subordinate to HeadOffice authority

Core Principle

AI Employees must be designed for continuous canon alignment.

An AI Employee is not valid simply because it was correct when created.

An AI Employee remains valid only when it stays aligned with the current approved MWMS source pages.

Static employees create drift.

Version-aware employees protect system integrity.

Canon Source Rule

Every AI Employee must have declared source pages.

Source pages may include:

  • Brain Canon
  • Brain Architecture
  • relevant frameworks
  • relevant protocols
  • relevant templates
  • relevant registry pages
  • HeadOffice governance pages

Each AI Employee must know:

  • which pages govern its behaviour
  • which page versions it is operating from
  • which authority layer controls its scope

If an employee cannot identify its source pages:

→ the employee is not operationally safe

Version Awareness Rule

Each AI Employee must record the version of every major source page it depends on.

Minimum version awareness fields:

  • Employee Name
  • Owning Brain
  • Active Role
  • Source Page Name
  • Source Page Version
  • Last Sync Date
  • Sync Status
  • Drift Risk

Example:

Employee Name:
Affiliate Offer Evaluator

Owning Brain:
Affiliate Brain

Source Pages:
Affiliate Brain Offer Intelligence v1.1
Affiliate Brain Messaging Architecture Framework v1.0
Affiliate Brain Strategic Narrative Framework v1.0
Affiliate Brain Message Mapping System v1.0
Affiliate Brain Messaging Canvas Template v1.0

Last Sync Date:
2026-04-26

Sync Status:
Current

Drift Risk:
Low

Employee Prompt Dependency Rule

AI Employee prompts must not be treated as independent truth.

Employee prompts are operational summaries of canon.

Canon remains the source of truth.

If the employee prompt conflicts with canon:

→ canon overrides the employee prompt

If a canon page changes:

→ related employee prompts must be reviewed

Prompt Refresh Trigger Rule

An AI Employee requires prompt review when any of the following changes occur:

  • source framework updated
  • Brain Canon updated
  • Brain Architecture updated
  • related protocol updated
  • related template updated
  • authority boundary changed
  • decision logic changed
  • workflow sequence changed
  • registry placement changed
  • new required input added
  • new failure condition added

Minor formatting changes do not require prompt refresh unless they affect behaviour.

Drift Detection Rule

AI Employees must check for drift before performing major work.

Drift indicators include:

  • source page version unknown
  • source page changed since last sync
  • employee instructions reference outdated page names
  • employee output skips new required fields
  • employee produces logic no longer aligned with canon
  • employee acts outside its Brain authority
  • employee uses deprecated workflow steps

If drift is detected:

→ employee must pause or flag HeadOffice review

Canon Sync Status Levels

Each AI Employee must have one sync status.

Current

Employee is aligned with latest approved source pages.

Needs Review

Source pages changed or employee may require update.

Outdated

Employee is known to rely on superseded logic.

Paused

Employee must not operate until reviewed.

Retired

Employee is no longer active.

Sync status must be visible to HeadOffice.

HeadOffice Sync Authority Rule

Only HeadOffice may approve an AI Employee as Current after a meaningful canon or framework change.

Brains may recommend updates.

Employees may flag drift.

HeadOffice confirms alignment.

This prevents self-authorising AI drift.

Employee Update Loop

The standard update loop is:

Source Page Updated

Affected Employees Identified

Prompt Dependency Reviewed

Employee Instructions Updated

Version Record Updated

HeadOffice Confirms Current Status

No AI Employee should continue operating silently after its governing rules have changed.

Employee Impact Check Integration

Every structural update to MWMS must include an Employee Impact Check.

The check must ask:

  • Does this update affect existing employees?
  • Does this update require a new employee?
  • Does this update change an employee’s scope?
  • Does this update change an employee’s inputs?
  • Does this update change an employee’s outputs?
  • Does this update require prompt refresh?

If yes:

→ update the relevant employee sync record

Source Of Truth Hierarchy

AI Employees must follow this hierarchy:

  1. HeadOffice Canon
  2. MWMS Architecture Registry
  3. Brain Canon
  4. Brain Architecture
  5. Brain Frameworks
  6. Brain Protocols
  7. Templates
  8. Employee Prompt
  9. Task Instruction

Lower levels may not override higher levels.

Task Instruction Boundary Rule

A task instruction may guide what the employee does in a specific case.

A task instruction may not override:

  • canon
  • authority boundaries
  • safety rules
  • financial controls
  • testing governance
  • compliance rules

If task instruction conflicts with source rules:

→ employee must flag conflict

Automation Readiness Rule

No AI Employee should be automated until it has:

  • declared source pages
  • version record
  • defined authority
  • defined inputs
  • defined outputs
  • drift detection process
  • HeadOffice review pathway

Automation without sync control creates long-term system risk.

Living System Rule

MWMS is expected to evolve.

Employee design must assume:

  • frameworks will change
  • templates will change
  • decision systems will improve
  • new Brains will emerge
  • authority boundaries may be refined

Therefore, every AI Employee must be modular and updateable.

No employee should be hard-coded around temporary logic.

Relationship To Other Pages

This framework works with:

  • How To Start A Session MWMS Operating Guide
  • MWMS Architecture Registry
  • MWMS Brain Registry
  • MWMS System Improvement Log
  • MWMS Brain To Brain Request Protocol
  • AI Output Standard Full File Delivery Rule
  • Brain specific Employee Registries

Failure Modes Prevented

This framework prevents:

  • outdated AI Employees using old logic
  • employee prompts drifting from canon
  • automation acting on superseded frameworks
  • duplicated employee behaviours
  • employees exceeding Brain authority
  • hidden prompt decay
  • stale decision systems
  • unsafe automation scaling

Drift Protection

The system must prevent:

  • AI Employees operating without source pages
  • AI Employees operating without version awareness
  • employee prompts replacing canon
  • canon changes failing to trigger employee review
  • automation running on outdated rules
  • HeadOffice losing visibility over employee alignment

Architectural Intent

The MWMS AI Employee Version And Canon Sync Framework ensures AI Employees remain aligned with the evolving MWMS ecosystem.

It allows MWMS to build automation without freezing system growth.

It protects the system from the risk that employees become outdated as frameworks improve.

This framework makes AI Employees updateable, governable, and safe to scale.

Final Rule

AI Employees must never be treated as fixed assets.

They are living operators inside a living system.

If an AI Employee is not synced to current canon:

→ it must not be trusted for execution

Change Log

Version: v1.0
Date: 2026-04-26
Author: HeadOffice

Change:
Created framework defining version awareness, canon sync, drift detection, prompt dependency control, and HeadOffice review requirements for all MWMS AI Employees.

Change Impact Declaration

Pages Created:
MWMS AI Employee Version And Canon Sync Framework

Pages Updated:
None

Pages Deprecated:
None

Registries Requiring Update:
MWMS Architecture Registry
HeadOffice Page Registry

Canon Version Update Required:
No

Change Log Entry Required:
Yes

End of Framework