System: MWMS
Document Type: Rule
Authority Level: MCR Source Of Truth
Status: Draft For MCR
Primary Location: MCR
Future Operational Destination: HeadOffice Brain, AI Business Systems Brain, AI Manager, AI Employee Router, Brain Room, Course Absorption System, Content Brain, Offer Brain, Sales Brain, Creative Brain, Future AIBS Client Systems
Parent Page: HeadOffice
Owner: Martyn
Developer Boundary: No Development Action Authorized By This Page
Source Of Truth: MCR
Purpose
The purpose of this document is to define the MWMS Manual Build Versus Skill Build Decision Rule.
This rule establishes how MWMS decides whether a piece of AI-assisted work should be handled as:
a one-off manual build
a repeatable AI skill
an update to an existing skill
a framework
a checklist
a context library update
a parked item
or no action
MWMS must not turn every useful process into a formal AI skill.
MWMS must also not repeatedly perform the same high-value work manually when it clearly needs a reusable procedure.
This rule exists to prevent two opposite forms of drift:
building too many skills too early
failing to build skills when repeated work clearly needs structure
Manual builds and skill builds serve different purposes.
A manual build is useful when the work is custom, exploratory, large, or not yet repeated.
A skill build is useful when the work is repeated, procedural, context-dependent, and likely to drift without guidance.
This rule protects MWMS from over-systemizing too early and under-systemizing repeated work.
Scope
This rule applies to all MWMS work where AI is used to create, process, analyze, review, route, or improve business outputs.
This includes:
course absorption
MCR page creation
context library creation
lead magnet creation
webinar creation
content planning
offer evaluation
ad script generation
VEO3 script generation
sales asset creation
client report drafting
newsletter intelligence
developer handoff preparation
proof review
voice review
client Brain onboarding
future AIBS client systems
This rule supports:
HeadOffice Brain
AI Business Systems Brain
AI Manager
AI Employee Router
Brain Room
Course Absorption System
Content Brain
Offer Brain
Creative Brain
Sales Brain
Conversion Brain
Research Brain
Affiliate Brain
Ads Brain
future AIBS client systems
This rule does not authorize development work, plugin changes, Supabase changes, WordPress changes, automation wiring, tool access, or M developer action.
Core Definition
Manual Build means AI assists with a specific one-off or semi-custom task using current instructions, selected context, and human review.
Skill Build means MWMS creates a reusable procedural skill for repeated use by an AI Employee or workflow.
A manual build is a work session.
A skill build is reusable operating procedure.
A manual build creates output.
A skill build creates capability.
Core Principle
The core principle of this rule is:
Use manual builds for custom or uncertain work. Use skills for repeated, defined, and drift-prone work.
Do not create a skill before the method is clear.
Do not keep manually repeating work once the method is stable and repeated.
Manual Build Use Cases
Use a manual build when the work is:
new
exploratory
large
custom
one-off
not yet repeated
not yet procedural
not yet validated
based on incomplete context
dependent on heavy human judgment
being tested before systemization
Examples:
first version of a new MCR framework
first lead magnet for a new offer
custom webinar outline
new client discovery report
first attempt at a new Brain concept
experimental campaign angle set
one-off strategy document
course closeout summary
Manual builds are useful because they allow MWMS to discover the real procedure before locking it into a skill.
Skill Build Use Cases
Use a skill build when the work is:
repeated
procedural
defined
valuable
context-dependent
output-specific
validation-ready
likely to drift without guidance
useful for an AI Employee
useful for future client systems
Examples:
course block absorption
MCR full page output
developer handoff creation
voice checking
proof review
content folder scanning
client intake interview
offer evaluation
context library construction
newsletter intelligence extraction
lead magnet outline generation after the method is proven
A skill should reduce repeated instruction and repeated correction.
Decision Questions
MWMS uses the following decision questions.
Question 1: Has This Task Happened Before?
If no, use a manual build.
If yes, continue.
Question 2: Will This Task Happen Again?
If no, use a manual build.
If yes, continue.
Question 3: Does MWMS Have A Specific Way To Do It?
If no, use manual builds until the method becomes clear.
If yes, continue.
Question 4: Can The Input Be Defined?
If no, do not create a skill yet.
If yes, continue.
Question 5: Can The Output Be Defined?
If no, do not create a skill yet.
If yes, continue.
Question 6: Can The Output Be Validated?
If no, do not create a skill yet.
If yes, continue.
Question 7: Does Generic AI Drift On This Task?
If yes, skill build may be needed.
If no, a simple prompt may be enough.
Question 8: Does This Skill Support A Real MWMS Workflow?
If no, park or reject.
If yes, continue.
Question 9: Does An Existing Skill Already Cover This?
If yes, update or merge.
If no, consider new skill.
Question 10: Is The Timing Right?
If the workflow is not active yet, park the skill idea.
If active, build or update the skill.
Decision Outcomes
MWMS uses the following outcomes.
Manual Build Now
Use when work is custom, new, or not yet procedural.
Manual Build And Observe
Use when the task may become repeated but needs more examples first.
Create New Skill
Use when task is repeated, defined, valuable, and drift-prone.
Update Existing Skill
Use when an existing skill covers the task but needs improvement.
Merge With Existing Skill
Use when the idea overlaps an existing skill.
Create Framework Instead
Use when the material defines a concept or model, not repeatable execution.
Create Checklist Instead
Use when the material defines review criteria.
Update Context Library
Use when the material adds source truth, not procedure.
Park For Later
Use when the idea may become useful later but is premature.
Reject
Use when the idea is weak, duplicate, low-value, or unnecessary.
Manual Build Rules
Rule 1: Manual Builds Must Still Use Context
Manual does not mean random.
Important manual builds still require context activation.
Rule 2: Manual Builds Remain Draft Until Reviewed
A manual build output should not become active source truth without review.
Rule 3: Manual Builds Should Reveal Procedure
If the same manual build happens repeatedly, capture the emerging method.
Rule 4: Manual Builds Should Not Become Endless Repetition
If Martyn keeps asking for the same thing, consider skill creation.
Rule 5: Manual Builds Must Have A Destination
Even one-off work should have an output destination, review path, or parking decision.
Skill Build Rules
Rule 1: Skills Must Have Clear Trigger Conditions
A skill must define when it should activate.
Rule 2: Skills Must Have Defined Input
A skill must define what it can process.
Rule 3: Skills Must Have Required Context
A skill must define what context it reads.
Rule 4: Skills Must Have A Procedure
A skill must define the steps.
Rule 5: Skills Must Have Forbidden Actions
A skill must define what it must not do.
Rule 6: Skills Must Have Output Format
A skill must define what it produces.
Rule 7: Skills Must Have Validation
A skill must define how output is checked.
Rule 8: Skills Must Have Handoff Destination
A skill must define where the output goes.
Rule 9: Skills Must Start Conservatively
New skills should begin as Draft, Manual Use, or Assisted Use, not automation.
Rule 10: Skills Must Be Audited
Installed skills require review over time.
Manual Build To Skill Conversion
A manual build may become a skill after repeated use.
Conversion signs include:
same task requested repeatedly
same structure reused
same corrections repeated
same context required
same output destination
same review path
same Brain ownership
same validation needs
same failure modes
When these signs appear, MWMS should consider converting the manual build into a formal skill.
Manual Build To Skill Workflow
Step 1: Identify Repetition
Confirm the task repeats.
Step 2: Capture The Working Method
Extract how the manual build is being done.
Step 3: Identify Stable Sections
Find what remains the same each time.
Step 4: Identify Variable Sections
Find what changes by task, offer, client, or Brain.
Step 5: Define Trigger
Define when the skill should activate.
Step 6: Define Input
Define required input.
Step 7: Define Context
Define required context.
Step 8: Define Output
Define required output.
Step 9: Define Validation
Define validation rules.
Step 10: Install As Draft Or Manual Use
Do not over-promote.
Skill To Manual Downgrade
A skill may be downgraded back to manual build if it proves unstable.
Downgrade signs include:
skill triggers too often
skill triggers for wrong tasks
input varies too much
output varies too much
human rewriting is frequent
procedure is not stable
context is often missing
review burden is too high
skill creates more confusion than value
Possible outcomes:
downgrade to manual build
split skill
merge skill
park skill
retire skill
Update Existing Skill Rule
Before creating a new skill, MWMS must check whether an existing skill can be updated.
Create a new skill only when:
the task has a different trigger
the input is different
the output is different
the owning Brain is different
the validation is different
the procedure is meaningfully different
If only examples, wording, or minor procedure steps differ, update the existing skill.
Framework Versus Skill Rule
Create a framework when the material explains a concept, model, philosophy, or operating logic.
Create a skill when the material defines repeatable execution.
Example:
Context-Driven Asset Builder Framework explains the system.
Lead Magnet Builder Skill executes one repeatable asset workflow.
Both may be needed, but they are not the same.
Checklist Versus Skill Rule
Create a checklist when the material defines review criteria.
Create a skill when the AI must perform a procedure.
Example:
AI Brain Readiness Review Checklist checks readiness.
Client Brain Intake Skill may run the intake process.
Context Update Versus Skill Rule
Update a context library when the source material adds truth.
Create a skill when the source material adds procedure.
Example:
A new customer phrase updates Customer Language Bank.
A repeatable process for extracting customer phrases becomes VOC Extraction Skill.
Automation Boundary Rule
A skill being useful does not mean it should be automated.
Automation requires separate readiness review.
A skill may move through:
Draft
Manual Use
Proven Manual Use
Assisted Use
Controlled Automation Candidate
Actual automation requires separate technical approval and developer work.
This rule does not authorize automation.
Risk-Based Decision Rules
High-risk tasks should stay manual or assisted longer.
High-risk areas include:
developer instructions
paid traffic
finance
compliance
client-facing outputs
public sales claims
health claims
income claims
privacy-sensitive work
MCR canon changes
For high-risk tasks, skills may assist but human review remains required.
Client System Decision Rules
Future AIBS client systems require stricter decisions.
Client workflows should become skills only when:
client process is clear
client context exists
client approval rules exist
client data boundary is clear
client output is defined
client risk is manageable
client review is required
client skill remains isolated
Do not create client skills from vague client requests.
Decision Template
Use the following template when deciding manual build versus skill build.
Task Name:
Current Request:
Has This Task Happened Before:
Will It Repeat:
Specific MWMS Method Exists:
Input Defined:
Output Defined:
Validation Defined:
Context Required:
Existing Skill Or Standard:
Risk Level:
Business Value:
Timing:
Decision:
Reason:
Next Action:
Review Required:
Notes:
Minimum Decision Template
For quick decisions, use:
Task:
Repeat:
Method Clear:
Output Clear:
Existing Skill:
Decision:
Next Action:
Common Failure Modes
MWMS must prevent:
turning every idea into a skill
repeating manual work forever
building skills before the method is clear
creating skills without defined output
creating skills without validation
creating duplicate skills
using manual builds without context
treating manual outputs as source truth
automating skills before manual proof
client skill creation before client workflow clarity
downgrading failed skills without recording why
Governance Role
HeadOffice owns the MWMS Manual Build Versus Skill Build Decision Rule.
HeadOffice is responsible for:
preventing skill sprawl
preventing repeated manual friction
requiring context for important manual builds
requiring skill readiness before installation
ensuring existing skills are checked before new skills are created
ensuring high-risk tasks retain review gates
ensuring client skills remain isolated
ensuring automation is not implied by skill creation
Individual Brains may recommend manual builds or skill builds, but HeadOffice governs cross-Brain, high-risk, MCR, client-facing, and automation-related decisions.
Relationship To Other MWMS Standards
This rule supports and must align with:
MWMS Document Structure Standard
MWMS AI Skill Brainstorm And Prioritization Framework
MWMS AI Skill Builder And Audit Protocol
MWMS AI Skill Installation And Usage Protocol
MWMS Source Material To AI Skill Conversion Framework
MWMS AI Context Pack Template Standard
MWMS AI Context Activation And Usage Protocol
MWMS Context-Driven Asset Builder Framework
MWMS AI Brain Readiness Review Checklist
MWMS AI Brain Audit And Decay Prevention Framework
MWMS AI Output Validation Standard
MWMS AI Agent Skill Library Framework
MWMS AI Agent Memory And Context Framework
MWMS Brain Routing Rule
MWMS MCR Promotion To Brain Protocol
MWMS Architecture Registry
AI Business Systems Brain Canon
This rule provides the decision gate between one-off work and reusable procedural skill creation.
Drift Protection
This rule protects MWMS from:
skill sprawl
premature skill creation
endless repeated manual work
duplicate skill creation
skills without clear input or output
manual builds without context
manual outputs becoming source truth too early
automation assumptions
client skills created before context is ready
Any proposed skill that has not passed the manual-versus-skill decision rule should not be treated as active or installable.
Architectural Intent
The architectural intent of the MWMS Manual Build Versus Skill Build Decision Rule is to keep MWMS practical.
MWMS needs reusable AI capability, but it does not need unnecessary skill clutter.
The long-term goal is that every repeated AI workflow can answer:
Should this stay manual?
Should this become a skill?
Should it update an existing skill?
Should it become a framework?
Should it become a checklist?
Should it update context?
Should it be parked?
Should it be rejected?
When MWMS can answer these questions consistently, the system becomes more efficient without becoming overcomplicated.
Change Log
v1.0 — Initial Draft
Created the MWMS Manual Build Versus Skill Build Decision Rule as the decision rule for choosing between one-off manual builds, repeatable AI skills, skill updates, frameworks, checklists, context updates, parked items, or rejection.
This rule defines manual build use cases, skill build use cases, decision questions, decision outcomes, manual build rules, skill build rules, manual-to-skill conversion, skill downgrades, update rules, framework-versus-skill rules, checklist-versus-skill rules, context-update-versus-skill rules, automation boundaries, risk-based decisions, client system decisions, templates, failure modes, governance role, drift protection, and architectural intent.
Change Impact Declaration
Pages Created:
MWMS Manual Build Versus Skill Build Decision Rule
Pages Updated:
None
Pages Deprecated:
None
Registries Requiring Update:
MWMS Architecture Registry
HeadOffice Page Registry
AI Business Systems Brain Page Registry
AI Skill Registry
Canon Version Update Required:
No
Change Log Entry Required:
Yes
Employee Impact Check
Employees impacted:
HeadOffice Manager Employee
AI Manager
AI Employee Router
Skill Auditor
Course Absorption Agent
Context Library Builder
Content Planner Employee
Offer Strategist Employee
Creative Strategist Employee
Sales Strategist Employee
Research Analyst Employee
AI Business Systems Architect Employee
Required behaviour updates:
AI Employees must decide whether a repeated task should remain a manual build, become a skill, update an existing skill, become a framework, become a checklist, update context, be parked, or be rejected.
AI Employees must not recommend skill creation for one-off, vague, unvalidated, low-value, or premature tasks.
AI Employees must not keep repeating manual work when the task is stable, repeated, valuable, and likely to drift without guidance.
AI Employees must not imply automation from skill creation.
AI Employees must preserve human review for high-risk manual and skill-assisted work.
END MWMS MANUAL BUILD VERSUS SKILL BUILD DECISION RULE v1.0