System: MWMS
Document Type: Canon / Parent Framework
Authority Level: MCR Source Of Truth
Status: Draft For MCR
Version: v1.0
Primary Location: MCR
Future Operational Destination: HeadOffice Brain, AI Employee Canon, Automation Brain, AIBS Brain, Content Brain, Ads Brain, Research Brain, Data Brain, Experimentation Brain, Compliance Brain, Risk Brain, Prompt Vault
Parent Page: HeadOffice
Owner: Martyn
Developer Boundary: Do Not Touch M’s Active Build Areas Unless Specifically Assigned
Source Of Truth: MCR
Last Reviewed: 2026-06-08
Source / Origin: MWMS Internal Architecture / AI Automations by Jack Master Prompting And Prompt System Design Block / Mindstream Prompting Framework / MWMS Prompt Vault Direction / AI Employee Development Direction
MWMS Classification: Prompting Canon / Prompt Governance Framework / AI Employee Prompt Standard / Prompt Architecture Parent Page / Prompt System Control Layer
Primary Brain: HeadOffice Brain
Supporting Brains: AI Employee Canon, Automation Brain, AIBS Brain, Content Brain, Ads Brain, Research Brain, Data Brain, Experimentation Brain, Compliance Brain, Risk Brain

Related Pages: MWMS Prompt Architecture And Automation Output Reliability Framework, MWMS AI Employee Evaluation Scorecard Standard, MWMS AI Observability Metadata Standard, MWMS AI Work Session Persistence Standard, MWMS Agent Loop Control Framework, MWMS Next Action Picker Standard, MWMS AI Usage And Cost Visibility Standard, MWMS Source Visibility And Evidence Display Standard, MWMS Buyer First Authority Content And Channel Growth Framework, MWMS AIBS Business Diagnostic And Opportunity Discovery Framework, HeadOffice Kaizen Continuous Improvement Loop

Purpose

The purpose of the MWMS Prompting Framework is to act as the parent canon page for all MWMS prompting standards, prompt architecture rules, prompt-chain design methods, AI Employee prompt requirements, prompt testing protocols, and future Prompt Vault governance.

This page exists because prompting is not a minor support skill inside MWMS.

Prompting is one of the core operating layers of the entire ecosystem.

MWMS depends on prompts for:

AI Employees
course absorption
newsletter intelligence
content creation
research extraction
affiliate offer evaluation
AIBS diagnostics
paid ad analysis
compliance review
client reports
automation workflows
task routing
decision support
internal Brain operations

The core purpose of this page is:

To provide the central home for all MWMS prompting knowledge, standards, rules, and future prompt infrastructure.

Core Doctrine

The MWMS doctrine is:

Prompts are not throwaway instructions. Prompts are reusable system assets.

A weak prompt creates weak output.

A vague prompt creates inconsistent output.

An untested prompt creates unreliable automation.

An undocumented prompt creates system drift.

A prompt hidden inside chat history creates memory loss.

MWMS must therefore treat important prompts as structured assets that can be:

written
tested
versioned
stored
reused
improved
scored
observed
governed
retired when obsolete

The Prompting Framework exists to stop MWMS from losing valuable prompt knowledge across random chats, scattered files, untracked automations, or undocumented AI Employee instructions.

Strategic Importance

This framework is strategically important because every Brain in MWMS will eventually depend on prompting quality.

If prompting quality is poor, then every Brain becomes weaker.

If prompting quality is strong, then every Brain improves.

Prompting quality affects:

output accuracy
consistency
reliability
speed
cost
compliance
client trust
internal productivity
AI Employee performance
automation stability
strategic decision quality
future consultant delivery
system scalability

The strategic shift is:

MWMS must move from casual prompting to governed prompt architecture.

This does not mean every prompt needs to be overbuilt.

It means important prompts must be treated according to their importance, risk, and reuse value.

A one-off brainstorming prompt may remain casual.

A prompt inside an AI Employee, automation, client report, compliance review, or AIBS diagnostic must be structured and governed.

Definition

Prompting Framework means the MWMS parent system for managing how prompts are written, structured, tested, stored, versioned, reviewed, and improved.

Prompt asset means a reusable prompt that has enough structure, purpose, testing, and documentation to be trusted inside an MWMS workflow.

Prompt architecture means the design of prompt structure, inputs, context, examples, output format, chain logic, model selection, testing, and observability.

Prompt governance means the rules that protect MWMS from unreliable, untested, duplicated, risky, or unmanaged prompts.

MWMS Definition

The MWMS Prompting Framework is:

The HeadOffice-controlled parent canon for all MWMS prompt standards, prompt architecture, prompt chains, AI Employee prompting rules, prompt quality testing, Prompt Vault direction, and prompt governance across the full MWMS ecosystem.

Scope

This framework applies to:

prompt standards
prompt architecture
prompt chains
AI Employee prompts
automation prompts
content prompts
research prompts
course absorption prompts
newsletter intelligence prompts
AIBS diagnostic prompts
sales prompts
ad prompts
compliance prompts
data extraction prompts
model selection rules
prompt testing protocols
prompt scoring
prompt versioning
prompt observability
prompt improvement logs
future Prompt Vault
future Prompt Saver upgrade
prompt reuse across Brains
prompt-based client deliverables

This framework does not replace every specialist prompt page.

It acts as the parent control layer that connects those pages together.

Core Principle

The core principle is:

The more important the output, the more disciplined the prompt must be.

Not every prompt needs full governance.

But every important prompt should have enough structure to protect the workflow.

Low Discipline Prompt

Allowed for:

quick brainstorming
personal idea generation
one-off chat exploration
rough drafting
temporary thinking

Medium Discipline Prompt

Required for:

repeated content workflows
internal research tasks
recurring summaries
offer reviews
draft reports
reusable templates

High Discipline Prompt

Required for:

AI Employees
automation workflows
client-facing reports
compliance-sensitive tasks
AIBS diagnostics
paid ad systems
data extraction
structured outputs
operational decisions
build handoffs

Rule

Prompt discipline must match output risk.

The MWMS Prompting Framework Model

The MWMS Prompting Framework is organized across ten control layers:

Prompt Purpose Layer
Prompt Asset Layer
Prompt Architecture Layer
Prompt Chain Layer
Prompt Testing Layer
Prompt Storage Layer
Prompt Versioning Layer
Prompt Observability Layer
Prompt Governance Layer
Prompt Improvement Layer

1. Prompt Purpose Layer

Every important prompt must have a clear purpose.

The purpose defines why the prompt exists and what output it must create.

Prompt Purpose Questions

Ask:

What is this prompt for?
Which Brain uses it?
Which AI Employee uses it?
What workflow does it support?
What output should it create?
Who consumes the output?
What happens if the output is wrong?
Is this a one-off prompt or reusable prompt?
Does this prompt support a business-critical process?

Rule

A prompt with no clear purpose should not become an MWMS asset.

2. Prompt Asset Layer

MWMS must distinguish prompt assets from prompt liabilities.

Prompt Asset

A prompt asset is:

reusable
tested
clear
structured
documented
versioned where needed
tied to a Brain or workflow
capable of producing consistent outputs

Prompt Liability

A prompt liability is:

vague
hidden in chat history
rewritten every time
untested
not versioned
unclear in purpose
inconsistent in output
not tied to a workflow
difficult to improve

Rule

Useful recurring prompts should be upgraded into prompt assets.

3. Prompt Architecture Layer

Prompt architecture defines how the prompt is built.

A strong MWMS prompt may include:

identity
task
context
input variables
constraints
examples
output format
quality criteria
source rules
compliance rules
failure handling
model notes

The main operational page for this layer is:

MWMS Prompt Architecture And Automation Output Reliability Framework

Rule

Prompt architecture should reduce guessing and increase repeatability.

4. Prompt Chain Layer

Many MWMS tasks are too complex for a single prompt.

Prompt chains allow MWMS to break complex work into smaller controlled steps.

Prompt Chain Use Cases

Use prompt chains for:

course absorption
newsletter intelligence
content production
AIBS diagnostics
offer evaluation
sales page analysis
ad creative testing
compliance review
research synthesis
client reports

Rule

Complex outputs should be broken into staged prompt chains when quality improves.

5. Prompt Testing Layer

Prompts must be tested before serious use.

Testing should include:

normal inputs
messy inputs
short inputs
long inputs
missing information
edge cases
high-risk cases
real-world examples
failure examples

Rule

A reusable prompt is not ready until it has been tested against realistic inputs.

6. Prompt Storage Layer

Prompt storage prevents loss.

MWMS should not rely on conversation history as the only storage location for important prompts.

Important prompts should be stored in a system such as:

MCR
Prompt Vault
Prompt Saver
Supabase prompt table in future
automation documentation
AI Employee configuration files
approved workflow documentation

Rule

If MWMS will need the prompt again, it should not live only in chat history.

7. Prompt Versioning Layer

Important prompts should be versioned.

Versioning helps MWMS know:

what changed
why it changed
when it changed
who changed it
what problem was fixed
whether the new version improved output
whether the old version should be restored

Versioning Fields

Use:

Prompt Name:
Version:
Date Updated:
Owner:
Change Made:
Reason:
Test Result:
Decision:

Rule

A production prompt should not be changed silently.

8. Prompt Observability Layer

Prompt observability means MWMS can trace prompt behavior.

For important automations, MWMS should know:

which prompt ran
which version ran
which model was used
what input was passed
what output was produced
whether validation passed
what the cost was
what the latency was
whether human review was needed
whether failure occurred

Rule

Important AI outputs should be traceable back to the prompt that created them.

9. Prompt Governance Layer

Prompt governance protects the system from prompt chaos.

Governance is required when prompts affect:

client-facing output
compliance-sensitive output
business decisions
paid advertising
financial assumptions
regulated claims
personal data
staff or customer data
AIBS diagnostics
automated publishing
data extraction

Rule

The higher the risk, the stronger the prompt governance.

10. Prompt Improvement Layer

Prompts should improve over time.

MWMS uses the Kaizen loop:

Reflect → Reduce → Refine → Record

For prompts this means:

Reflect on output quality.
Reduce unnecessary complexity.
Refine instructions and examples.
Record what changed.

Rule

Prompt improvement should compound instead of being rediscovered repeatedly.

Prompting Framework Page Family

This parent page should contain or govern the following page family.

Current Page Family

MWMS Prompt Architecture And Automation Output Reliability Framework

Purpose:

defines how prompts are designed for automation reliability
covers prompt assets, atomic prompts, compound prompts, deconstruction, chaining, examples, model testing, cost, latency, and observability

Future Page Candidates

These should only be created when needed.

MWMS Prompt Vault Governance Framework

Purpose:

defines how prompts are stored, named, categorized, searched, versioned, and reused across MWMS

MWMS AI Employee Prompt Standard

Purpose:

defines mandatory prompt requirements for every AI Employee

MWMS Prompt Testing And Evaluation Framework

Purpose:

defines structured testing methods for prompts before deployment

MWMS Prompt Chain Design Framework

Purpose:

defines how multi-step prompt chains should be designed and maintained

MWMS Prompt Cost And Latency Governance Framework

Purpose:

defines how prompt cost, model choice, and latency are managed at scale

MWMS Prompt Library Naming And Classification Standard

Purpose:

defines naming conventions for prompt assets across Brains and workflows

Rule

Do not create prompt pages unless they solve a real governance or operational need.

Prompt Asset Naming Standard

Prompt assets should be named clearly.

Naming Format

Use:

MWMS [Brain / Function] [Task] Prompt vX.X

Examples:

MWMS Content Brain Buyer Question Extraction Prompt v1.0
MWMS AIBS Diagnostic Opportunity Scoring Prompt v1.0
MWMS Compliance Brain Ad Claim Risk Review Prompt v1.0
MWMS Research Brain Source Extraction Prompt v1.0
MWMS Newsletter Intelligence 15 Layer Extraction Prompt v1.0

Rule

Prompt names should tell MWMS what the prompt does without opening the prompt.

Prompt Classification Standard

Every important prompt should be classified.

Classification Fields

Use:

Prompt Name:
Brain:
AI Employee:
Workflow:
Prompt Type:
Risk Level:
Output Type:
Model:
Status:
Owner:
Version:
Last Reviewed:

Prompt Type Options

Conversational
One Shot
Atomic
Compound
Chain Step
Evaluation
Extraction
Classification
Generation
Review
Routing
Compliance

Risk Level Options

Low
Medium
High
Critical

Status Options

Draft
Testing
Approved
Deprecated
Replaced
Parked

Rule

Prompt classification makes prompt reuse and governance possible.

AI Employee Prompt Requirements

Every future AI Employee should have a prompt set.

Required Prompt Set

Each AI Employee should eventually define:

role prompt
task prompt
input interpretation prompt
output formatting prompt
quality review prompt
failure handling prompt
escalation prompt
logging metadata
version history

AI Employee Prompt Questions

Ask:

What is this Employee allowed to do?
What is it not allowed to do?
What inputs can it receive?
What outputs must it produce?
What should it do when uncertain?
What should be escalated to Martyn?
What should be escalated to M?
What should be escalated to Compliance Brain?
What should be logged?

Rule

An AI Employee without prompt standards is not ready for serious operational work.

Prompt Vault Direction

The future MWMS Prompt Vault should become the central library for all approved prompts.

The Prompt Vault should eventually support:

prompt storage
search
categories
Brain assignment
Employee assignment
status
versioning
testing notes
examples
owner
last reviewed date
copy/use tracking
Supabase sync
Chrome extension capture
prompt improvement notes

Prompt Vault Rule

Prompt Vault should store only prompts worth reusing, not every random chat instruction.

Prompt Saver Direction

The future Prompt Saver upgrade should support MWMS by capturing useful prompts before they disappear.

Prompt Saver should eventually allow:

saving prompt from browser
tagging by Brain
tagging by workflow
assigning status
adding notes
syncing to Supabase
linking to MCR page
tracking prompt version
storing example outputs
recording improvement ideas

Prompt Saver Rule

Prompt Saver should help convert useful prompt discoveries into governed MWMS prompt assets.

Prompt Testing Standard Summary

Important prompts should be tested using:

multiple inputs
edge cases
bad inputs
missing information
real production samples
output scoring
formatting validation
model comparison
cost comparison
latency review

Prompt Testing Rule

A prompt is not approved because it looks good. It is approved because it performs reliably.

Prompt Governance Standard Summary

Prompt governance is required when a prompt is:

reused often
automated
client-facing
compliance-sensitive
cost-heavy
connected to paid ads
connected to financial assumptions
connected to customer data
connected to AIBS diagnostics
used by an AI Employee
used in business decision-making

Prompt Governance Rule

The higher the impact of the prompt, the more governance it needs.

Prompt Improvement Standard Summary

MWMS should improve prompts through:

output review
failure analysis
example improvement
clearer constraints
better input variables
model testing
chain redesign
cost reduction
latency reduction
human feedback
Kaizen logs

Prompt Improvement Rule

Prompt learning must be recorded so it compounds.

Relationship To AI Employee Canon

The Prompting Framework supports the AI Employee Canon.

AI Employees need:

stable role instructions
task boundaries
output expectations
escalation rules
memory rules
quality controls
observability metadata
cost awareness
compliance constraints

The Prompting Framework defines how those prompt components should be structured and governed.

Rule

AI Employee quality depends on prompt governance.

Relationship To Automation Brain

Automation Brain depends on prompt reliability.

Automation should not be built around unstable prompts.

Before a prompt is automated, Automation Brain should confirm:

prompt purpose
output format
model choice
validation method
failure handling
cost estimate
latency fit
human review requirement
version record

Rule

Do not automate a prompt that still needs constant manual correction.

Relationship To AIBS Brain

AIBS Brain will use prompts for diagnostics, client reports, opportunity scoring, AI readiness reviews, and proposal support.

Because AIBS may involve client data, its prompts must include:

privacy constraints
approved data boundaries
no unnecessary exposure of sensitive data
clear output structure
diagnostic logic
risk notes
human review for high-risk findings

Rule

AIBS prompts must be safer and more controlled than casual internal prompts.

Relationship To Content Brain

Content Brain will use prompting for research, ideation, scripting, rewriting, content repurposing, AI visibility, and authority content.

Content prompts must be guided by:

buyer questions
proof
examples
platform context
compliance constraints
tone rules
output format
performance feedback

Rule

Content prompts should be based on proven content logic, not generic “write me a post” instructions.

Relationship To Research Brain

Research Brain will use prompting to extract, classify, summarize, compare, and route knowledge.

Research prompts must preserve:

source context
uncertainty
evidence
citations where needed
relevance to MWMS
distinction between fact and inference

Rule

Research prompts must protect evidence quality.

Relationship To Compliance Brain

Compliance Brain must review prompts that could create risk.

High-risk prompt categories include:

ad claims
health claims
financial claims
income claims
legal claims
privacy-sensitive outputs
client-facing diagnostics
affiliate claims
PPL claims
synthetic media
testimonial generation

Rule

Compliance-sensitive prompts must include risk controls before use.

Relationship To Experimentation Brain

Prompt testing should be treated like experimentation.

Experimentation Brain can test:

prompt versions
model choices
example counts
output structures
prompt-chain designs
cost differences
latency differences
output quality scores
human acceptance rates

Rule

Important prompt changes should be testable and measurable.

Relationship To HeadOffice Brain

HeadOffice governs prompting direction across MWMS.

HeadOffice should prevent:

scattered prompt chaos
duplicate prompt work
weak automation prompts
undocumented AI Employee prompts
prompt drift
prompt bloat
uncontrolled cost
risky prompt use
losing strong prompts in chat history

Rule

HeadOffice owns the prompting governance layer until a dedicated Prompt Brain exists.

Prompting Framework Operating Rules

Rule 1: Do Not Treat Production Prompts As Casual Chat

Any prompt used in automation or AI Employee work must be structured.

Rule 2: Store Reusable Prompts

If MWMS will use the prompt again, it should be stored.

Rule 3: Version Important Prompts

Important prompts need version control.

Rule 4: Test Before Automating

Do not automate untested prompts.

Rule 5: Use Examples Where Output Quality Matters

Examples improve reliability.

Rule 6: Separate Inputs From Instructions

Dynamic variables must be clearly separated from fixed instructions.

Rule 7: Track Cost And Latency

High-volume prompts must be cost-aware.

Rule 8: Use Human Review For High-Risk Outputs

AI should not silently publish or decide high-risk outputs.

Rule 9: Improve Through Kaizen

Prompt failures should become prompt improvements.

Rule 10: Do Not Create Prompt Page Bloat

Only create new prompt pages when there is a real system need.

Prompt Governance Checklist

Use this checklist for important prompts.

Identity

prompt has a clear name
prompt has an owner
prompt has a Brain
prompt has a workflow

Purpose

task is clear
output is clear
user of output is clear
risk level is clear

Structure

input variables defined
context included
guidelines included
constraints included
examples included where needed
output format defined

Testing

tested with multiple inputs
edge cases tested
failure modes recorded
model selected deliberately
cost checked
latency checked

Governance

version recorded
status assigned
review date recorded
storage location known
change log started if needed

Prompt Page Creation Rules

Do not create a new prompt-related MCR page for every method, technique, or tactic.

Create A New Prompt Page When

Create when:

the topic affects multiple Brains
the topic needs a formal MWMS standard
the topic will be reused often
the topic affects AI Employees
the topic affects automation reliability
the topic affects cost or risk
the topic needs governance

Do Not Create A New Prompt Page When

Do not create when:

the idea is a small tactic
the idea belongs inside an existing framework
the idea is tool-specific only
the idea is not yet operationally needed
the idea would create page bloat
the idea is already covered elsewhere

Rule

Prompting knowledge should be consolidated until there is a clear reason to split it.

Future Prompting System Roadmap

Stage 1: Canon Stabilization

Create core MCR pages:

MWMS Prompting Framework
MWMS Prompt Architecture And Automation Output Reliability Framework

Stage 2: Prompt Asset Capture

Begin identifying reusable prompts from:

course absorption
newsletter intelligence
content workflows
AIBS diagnostics
AI Employee design
automation workflows

Stage 3: Prompt Vault Rebuild

Rebuild MWMS Prompt Vault as a structured prompt library.

Stage 4: Prompt Saver Upgrade

Upgrade Prompt Saver to capture prompts from daily work and sync to MWMS storage.

Stage 5: AI Employee Prompt Sets

Every future AI Employee receives documented prompt sets and version history.

Stage 6: Prompt Observability

Prompt metadata becomes visible in AI task logs and automation outputs.

Stage 7: Prompt Optimization

Experimentation Brain and HeadOffice review prompt performance, cost, and quality over time.

Future AI Employee Ideas

These are parked candidates only.

Prompt Architect

Primary Brain: HeadOffice Brain / Prompting Framework
Status: Parked Candidate
Purpose: Designs reusable prompts and prompt chains for MWMS Brains and AI Employees.

Prompt Librarian

Primary Brain: Prompting Framework / Data Brain
Status: Parked Candidate
Purpose: Organizes prompt assets, naming, tagging, storage, versioning, and retrieval.

Prompt Quality Reviewer

Primary Brain: Experimentation Brain / Prompting Framework
Status: Parked Candidate
Purpose: Tests prompt outputs, scores quality, detects failure modes, and recommends improvements.

Prompt Cost Auditor

Primary Brain: Finance Brain / Prompting Framework
Status: Parked Candidate
Purpose: Reviews prompt cost, model choice, token use, and latency for high-volume workflows.

AI Employee Prompt Steward

Primary Brain: AI Employee Canon / Prompting Framework
Status: Parked Candidate
Purpose: Maintains prompt sets for AI Employees and ensures prompt changes are versioned.

Prompt Vault Manager

Primary Brain: Prompting Framework / Data Brain
Status: Parked Candidate
Purpose: Manages Prompt Vault structure, categories, prompt records, examples, and search.

Deferred Update And Parking Lot Section

This page creates later update needs.

Later Update 1: MWMS Prompt Architecture And Automation Output Reliability Framework

Confirm this page sits under:

Parent Page: MWMS Prompting Framework

Add link back to this parent page if needed.

Later Update 2: MWMS AI Employee Evaluation Scorecard Standard

Add prompt quality as a formal AI Employee evaluation category.

Include:

prompt reliability
prompt versioning
output consistency
prompt chain stability
failure handling
prompt observability

Later Update 3: MWMS AI Observability Metadata Standard

Add prompt metadata fields.

Include:

prompt name
prompt version
prompt chain step
model used
input tokens
output tokens
cost
latency
validation status

Later Update 4: MWMS AI Usage And Cost Visibility Standard

Add prompt-specific cost visibility.

Include:

cost per prompt run
cost per prompt chain
cost per successful output
model comparison
high-volume prompt alerts

Later Update 5: MWMS Prompt Vault Future Build

When development resumes in the right project, create or rebuild Prompt Vault around this canon.

Include:

Supabase sync
Chrome extension capture
prompt categories
versioning
tags
Brain assignment
search
examples
status

Later Update 6: MWMS Course Absorption Decision Registry

Record that this parent page was created as a structural support page for the Master Prompting block if registry update is required.

Drift Protection

This framework protects MWMS from:

losing useful prompts in chat history
building AI Employees without prompt standards
using vague prompts in automations
duplicating prompts across Brains
not knowing which prompt created an output
changing prompts without versioning
using untested prompts for important work
ignoring prompt cost
ignoring prompt latency
ignoring prompt risk
storing prompts without classification
creating prompt page bloat
treating prompting as casual rather than architectural

Drift Signals

Watch for:

“Just use the prompt from the chat.”
“We can find it later.”
“No need to save this.”
“The prompt worked once.”
“The prompt is somewhere in Make.”
“The AI Employee keeps changing output.”
“We do not know which prompt version this used.”
“The prompt costs too much but nobody noticed.”
“The output is wrong and nobody knows why.”
“We created another prompt page for a tiny method.”
“The prompt lives only in someone’s browser.”

Rule

If prompting knowledge matters, it must be captured, organized, and governed.

Strategic Summary

The MWMS Prompting Framework establishes prompting as a core MWMS operating layer.

It gives MWMS a parent page for:

prompt architecture
prompt chains
prompt assets
prompt testing
prompt storage
prompt versioning
prompt observability
prompt governance
Prompt Vault direction
Prompt Saver direction
AI Employee prompt standards

This page exists because MWMS will not scale on random prompts.

MWMS will scale on reusable prompt systems.

The key lesson is:

Prompting is not just how MWMS talks to AI. Prompting is how MWMS turns intelligence into repeatable work.

The more MWMS grows, the more important prompt governance becomes.

This page gives that governance a proper home.

Final Standard

The MWMS final standard is:

All important MWMS prompts must be treated as reusable system assets when they are used inside AI Employees, automations, client-facing workflows, research systems, content systems, diagnostics, compliance reviews, or recurring business operations.

A valid MWMS prompt asset should define:

name
purpose
Brain
workflow
owner
prompt type
inputs
context
guidelines
examples where needed
output format
model
testing status
version
last reviewed date
storage location
risk level

That is the MWMS Prompting Framework standard.

Change Log

Version: v1.0

Date: 2026-06-08
Author: HeadOffice

Change:
Created the MWMS Prompting Framework as the parent canon page for MWMS prompt standards, prompt architecture, prompt governance, AI Employee prompt requirements, Prompt Vault direction, and Prompt Saver direction.

Created this page to provide a proper MCR parent for:

MWMS Prompt Architecture And Automation Output Reliability Framework

Defined the MWMS Prompting Framework Model with ten control layers:

Prompt Purpose Layer
Prompt Asset Layer
Prompt Architecture Layer
Prompt Chain Layer
Prompt Testing Layer
Prompt Storage Layer
Prompt Versioning Layer
Prompt Observability Layer
Prompt Governance Layer
Prompt Improvement Layer

Added key operating sections:

Prompting Framework Page Family
Prompt Asset Naming Standard
Prompt Classification Standard
AI Employee Prompt Requirements
Prompt Vault Direction
Prompt Saver Direction
Prompt Testing Standard Summary
Prompt Governance Standard Summary
Prompt Improvement Standard Summary
Prompt Page Creation Rules
Future Prompting System Roadmap
Future AI Employee Ideas
Deferred Update And Parking Lot Section

Mapped the framework across:

HeadOffice Brain
AI Employee Canon
Automation Brain
AIBS Brain
Content Brain
Ads Brain
Research Brain
Data Brain
Experimentation Brain
Compliance Brain
Risk Brain
Prompt Vault

Purpose of creation:
To establish a formal parent page and canon home for all MWMS prompt standards, prompt assets, prompt chains, prompt testing, prompt governance, AI Employee prompt requirements, Prompt Vault development, and future prompt infrastructure.

END — MWMS PROMPTING FRAMEWORK v1.0