MWMS Programmatic Video And Dynamic Visual Content Framework

System: MWMS
Document Type: Framework
Authority Level: MCR Source Of Truth
Status: Draft For MCR
Version: v1.1
Primary Location: MCR
Future Operational Destination: Content Brain, Brand Brain, Data Brain, Automation Brain, AIBS Brain, HeadOffice Brain, AI Manager, AI Employee Router, Task Executor Systems, Content Production Systems, Client Delivery Systems
Parent Page: Content Brain
Owner: Martyn
Developer Boundary: Do Not Touch M’s Active Build Areas Unless Specifically Assigned
Source Of Truth: MCR
Last Reviewed: 2026-06-27
Source / Origin: AI Automations by Jack — Lesson 129 Remotion × AntiGravity and Image + Video Block covering VEO generation, image-to-video production, dynamic visual content, creative candidate selection, reroll and regeneration workflows, visual approval dashboards, caption composition and platform-specific output
MWMS Classification: Programmatic Video Framework / Dynamic Visual Asset System / Data-Driven Content Production Framework / Creative Candidate Production And Approval Framework / Image-To-Video Production Framework
Primary Brain: Content Brain
Supporting Brains: Brand Brain, Data Brain, Automation Brain, AIBS Brain, HeadOffice Brain
Related Pages: Content Brain Content Production System Framework, Content Brain Content Repurposing Framework, Content Brain Content Brief Template, Content Brain Editorial Consistency Framework, Content Brain AI Content Quality Governance Framework, Content Brain Publishing Readiness Checklist, Content Brain Conversion Support Framework, Content Brain Story Design Framework, MWMS Voice Architecture And Brand Language Standard, MWMS Brand Signal and Authority Framework, MWMS Content Repurposing And Social Automation Engine Framework, MWMS Buyer First Authority Content And Channel Growth Framework, MWMS Market Driven Social Content Production Framework, MWMS AI App Builder And Productized Interface Framework, MWMS KPI Dashboard And Insight Summary Framework, MWMS AI Workflow Pipeline Standard, MWMS AI Output Validation Standard, MWMS AI Tool Permission And Access Framework, MWMS AI Agent Outcome Measurement Framework, MWMS Prompt Architecture And Automation Output Reliability Framework
Source Evidence: The source material describes systems for generating motion graphics, product demonstrations, social video, data-driven videos, AI images, image-to-video outputs and platform-ready visual assets through reusable components, structured prompts, persistent data connections, brand rules, automated rendering and human approval. The later Image + Video block adds formal creative candidate sets, reviewer selection, whole-set regeneration, selected-candidate finalisation, prompt and source lineage, and separate content, asset and publishing approval. The durable value is not the named software combination. It is the conversion of video and dynamic visual media from one-off manual editing tasks into governed, reusable, reviewable and data-responsive content production systems.
MCR Duplication Check: Content Brain already contains production, repurposing, editorial consistency, publishing readiness, story design and conversion support frameworks. This page remains the correct unified home for programmatic video, reusable motion components, dynamic visual generation, image-to-video transformation, creative candidate selection, controlled regeneration and final visual approval. No separate candidate approval or image-to-video page is required.

Purpose

The purpose of this document is to define the MWMS Programmatic Video And Dynamic Visual Content Framework.

This framework establishes how MWMS may design, produce, validate, select, update, reuse and distribute video and motion-based content through structured components, approved data, reusable templates, governed generation and controlled human approval.

The framework exists because conventional video and visual production often requires repeated manual work for:

layout

animation

text placement

chart creation

brand styling

timing

transitions

data updates

resizing

versioning

export

quality review

candidate comparison

revision

final asset selection

Programmatic production changes this model.

Instead of creating every asset from the beginning, MWMS may define reusable visual systems that accept controlled inputs and generate approved outputs.

Where generative systems are used, MWMS may also create multiple creative candidates from one approved brief, compare those candidates, select one for finalisation or reject the full set and request controlled regeneration.

This can support:

product demonstrations

software walkthroughs

animated explainers

social video

image-to-video content

text-to-video content

data summaries

campaign updates

client reports

dashboard recaps

case-study videos

comparison content

promotional assets

educational content

short-form platform content

dynamic quote content

news and source-based visual content

visual lead magnets

sales-support media

client-specific variants

The goal is not to remove human judgement.

The goal is to remove repetitive production work while preserving:

business purpose

source accuracy

content quality

creative direction

brand consistency

visual quality

rights and permissions

approval authority

publishing control

measurable outcomes

Core Principle

Programmatic production should automate repeatable construction, not outsource responsibility.

A render may be technically successful and still be:

factually wrong

visually misleading

poorly composed

off-brand

legally restricted

commercially weak

unsuitable for the intended platform

unworthy of publication

No asset becomes approved merely because a system generated it.

Scope

This framework applies to MWMS video and dynamic visual content that may be generated or updated from structured inputs.

This includes:

short-form social video

product demonstration video

application walkthroughs

dashboard summary video

campaign performance recaps

animated charts

data visualisations

educational explainers

case-study videos

offer comparison videos

onboarding videos

client update videos

sales-support assets

lead-generation assets

presentation animations

image-to-video assets

text-to-video assets

AI-generated visual sequences

quote-based social content

article-to-visual content

podcast-to-visual content

source-driven visual content

client-specific visual variants

captioned social media video

visual sales letters

landing-page video assets

This framework may support internal MWMS production, owned media, affiliate content, client work and productised AIBS services.

This framework does not authorise:

unreviewed public publishing

unverified factual claims

unauthorised likeness use

unauthorised voice cloning

unlicensed music or media

deceptive synthetic representation

fabricated testimonials

false product demonstrations

misleading before-and-after content

uncontrolled high-volume content generation

unapproved political or sensitive-event synthetic media

Definitions

Programmatic Video

Programmatic Video is video produced through reusable structures, components, data fields, rules and rendering logic rather than rebuilt manually for every output.

Dynamic Visual Content

Dynamic Visual Content is visual media that can change when approved underlying inputs change.

Inputs may include:

text

images

brand assets

product screenshots

structured data

charts

metrics

audio

captions

timing instructions

customer segments

campaign results

client records

approved claims

source articles

quotes

podcast extracts

approved still images

motion instructions

platform requirements

Creative Candidate

A Creative Candidate is one possible visual interpretation generated from an approved content brief and prompt contract.

A creative candidate is not a format variant.

It competes with other candidates for selection before final rendering.

Candidate Set

A Candidate Set is a controlled group of creative candidates generated from the same approved brief, source package and prompt contract.

Selected Candidate

A Selected Candidate is the candidate chosen by an authorised reviewer for finalisation.

Regeneration

Regeneration is a controlled return to generation after the candidate set is rejected or found unsuitable.

Regeneration must preserve the approved brief and fixed creative contract unless an authorised revision changes them.

Format Variant

A Format Variant is a destination-specific version of an already selected creative direction.

Examples include:

16:9 landscape

9:16 vertical

1:1 square

captioned version

silent version

language version

client-branded version

Creative Candidate Versus Format Variant Rule

A creative candidate represents a different creative interpretation.

A format variant represents the same selected creative interpretation adapted for a different destination.

These must not be treated as the same workflow state.

Image-To-Video

Image-To-Video is a production route in which an approved still image becomes the visual source for controlled motion generation.

Text-To-Video

Text-To-Video is a production route in which an approved brief, script or structured prompt is used to generate moving visual content without a required source image.

Prompt Contract

A Prompt Contract is the controlled instruction structure governing generation.

It may define:

fixed creative rules

variable content fields

required output format

duration limits

word limits

aspect ratio

camera behaviour

motion rules

style restrictions

brand requirements

negative constraints

example outputs

machine-readable response requirements

A programmatic video system may separate:

content

data

brand rules

scene structure

creative candidate generation

candidate review

animation logic

rendering

validation

publishing

performance learning

When Programmatic Production Is Appropriate

Programmatic production is strongest when:

the content structure repeats

inputs change predictably

brand rules are stable

the business purpose is clear

data can be verified

multiple variants are required

production speed creates value

manual editing adds little unique judgement

outputs can be validated consistently

the production volume is justified

candidate comparison improves quality

Programmatic production is weaker when:

the story is highly original

emotional nuance is central

data is unreliable

claims require careful judgement

visual direction changes constantly

each asset has a fundamentally different purpose

rights are unclear

the subject is sensitive

the visual could misrepresent a real person or event

MWMS should not force every video into automation.

Why Programmatic Video Matters

A governed programmatic video system may allow MWMS to:

turn one approved template into many outputs

update videos when data changes

reduce repetitive editing

preserve brand consistency

produce multiple aspect ratios

create client-specific variants

convert reports into visual summaries

create recurring campaign recaps

reuse approved animation components

improve content production speed

connect Content Brain to Data Brain

create productised AIBS services

support premium client deliverables

generate several creative directions before commitment

compare creative options systematically

preserve the lineage of the chosen asset

separate creative selection from final rendering

Without governance, programmatic video may create:

inaccurate charts

stale information

broken layouts

low-quality visual output

repetitive content

brand inconsistency

unsupported claims

rendering errors

publishing mistakes

synthetic-looking content

excessive production volume without value

false or exaggerated visuals

unlicensed likeness or voice use

creative candidate confusion

uncontrolled regeneration cost

loss of source and prompt lineage

publication of the wrong candidate

Programmatic Video Operating Model

The MWMS Programmatic Video And Dynamic Visual Content Framework uses fourteen operating layers:

Business Purpose

Audience And Use Case

Source And Data Control

Content Architecture

Prompt Contract And Variable Input Control

Scene Architecture

Reusable Component System

Brand And Motion System

Creative Candidate Generation And Selection

Rendering And Variant Logic

Validation And Quality Review

Publishing And Distribution Control

Performance Measurement

Learning And Template Improvement

Business Purpose

Every programmatic video system must begin with a business purpose.

Possible purposes include:

explain a product

demonstrate a workflow

summarise performance

educate an audience

support a sales conversation

improve onboarding

present proof

generate qualified interest

increase content output

reduce editing time

create client reporting value

support campaign testing

increase conversion

improve outreach response

support a landing page

Required Questions

What business problem does the video solve?

What action should the viewer take?

Why is video the correct format?

Why should this content be programmatic?

What part repeats?

What part changes?

What outcome will be measured?

Does the asset require multiple creative candidates?

What level of human judgement is required?

Business Purpose Rule

No dynamic video system should be created solely because the technology is impressive.

Audience And Use Case

The framework must define who the video serves.

Possible audiences include:

prospects

existing customers

affiliate buyers

clients

internal team members

business owners

campaign managers

sales teams

executives

platform users

newsletter subscribers

social audiences

Audience requirements may affect:

language

pace

visual complexity

video length

aspect ratio

captioning

proof requirements

call to action

data detail

tone

branding

accessibility

synthetic-media disclosure

Use Case Rule

A dashboard recap for an executive is not the same asset as a social video for a cold audience.

The same underlying data may require different visual treatments.

Source And Data Control

Programmatic video depends on controlled inputs.

Possible sources include:

approved content briefs

verified research

Content Brain records

Supabase data

campaign metrics

analytics

client-approved data

product information

approved screenshots

case-study evidence

proof records

brand assets

script records

offer records

dashboard outputs

approved articles

podcast transcripts

meeting transcripts

approved quotes

approved still images

approved voice assets

Each material input should define:

Source Name:

Source Type:

Source Authority:

Source URL Or Record ID:

Source Date:

Event Date Where Relevant:

Date Retrieved:

Owner:

Approved Use:

Freshness Requirement:

Validation Status:

Client Boundary:

Rights Status:

Source Control Rule

No unverified data should enter a production video as though it were fact.

Dynamic generation increases the importance of source control because one bad input may create many incorrect outputs.

For current events, the publication date and event date must not be confused.

Content Architecture

Programmatic video must separate the content structure from the rendering system.

Content architecture may include:

title

hook

problem

context

key message

supporting points

evidence

demonstration

comparison

proof

call to action

disclaimer

closing message

Content should be approved independently from visual rendering where risk or complexity requires it.

The content record should remain accessible throughout candidate generation, rendering, validation and publication.

Content Architecture Rule

The rendering system may present approved content.

It must not silently rewrite approved claims, meaning or commercial promises.

Prompt Contract And Variable Input Control

Every generative visual workflow must separate stable instructions from changing inputs.

Fixed creative contract elements may include:

visual style

brand rules

tone

camera behaviour

motion behaviour

duration

aspect ratio

resolution

caption style

text placement

safe zones

negative constraints

disclosure rules

output schema

quality threshold

Variable inputs may include:

subject

topic

source material

quote

business name

pain point

avatar

location

product

campaign

audience

call to action

approved image

approved data period

Prompt Contract Requirements

Each prompt contract should define:

Prompt Contract ID:

Prompt Contract Version:

Purpose:

Approved Models Or Model Class:

Fixed Instructions:

Variable Fields:

Required Fields:

Optional Fields:

Maximum Text Length:

Maximum Spoken Words:

Duration Limit:

Aspect Ratio:

Camera And Motion Rules:

Brand Rules:

Negative Instructions:

Example Input:

Example Output:

Output Format:

Validation Requirements:

Owner:

Status:

Prompt Contract Rule

Fixed rules must not be mixed casually with uncontrolled variable text.

A variable input must not be allowed to override:

brand constraints

rights restrictions

disclosure requirements

safety rules

duration limits

output format

approval boundaries

Scene Architecture

A programmatic video should be organised into defined scenes.

Possible scene types include:

opening hook

title card

problem statement

talking-point scene

screenshot demonstration

screen recording

animated text

chart

comparison

before-and-after

testimonial

proof point

feature highlight

process step

call to action

disclaimer

closing brand card

AI-generated environment

image-to-video scene

talking-character scene

Each scene should define:

Scene ID:

Scene Type:

Purpose:

Input Fields:

Duration:

Visual Component:

Animation Rule:

Audio Rule:

Caption Rule:

Transition:

Validation Requirement:

Scene Architecture Rule

Scenes should be reusable without becoming generic.

A scene template must still support the communication goal of the specific asset.

Reusable Component System

Reusable visual components may include:

title cards

lower thirds

metric cards

charts

progress bars

data tables

device frames

screenshot frames

quote cards

comparison layouts

call-to-action cards

logos

captions

transitions

background systems

end screens

candidate review cards

approval controls

Each component should define:

Component Name:

Purpose:

Approved Inputs:

Required Inputs:

Optional Inputs:

Maximum Text Length:

Brand Rules:

Responsive Rules:

Failure State:

Approved Uses:

Component Owner:

Component Status:

Reusable Component Rule

A component should become reusable only after it has demonstrated:

visual stability

brand alignment

responsive behaviour

input tolerance

rendering reliability

practical value

Brand And Motion System

Programmatic visual output must align with MWMS or approved client branding.

Brand controls may include:

logo treatment

typefaces

colour system

spacing

icon style

image style

background style

chart style

visual hierarchy

border radius

shadows

glow effects

motion speed

transition style

caption style

call-to-action treatment

Motion controls may include:

entrance animation

exit animation

timing

easing

duration

transition behaviour

sequencing

emphasis

loop behaviour

camera movement

subject movement

background movement

reduced-motion alternative

Brand And Motion Rule

Animation should support comprehension and emphasis.

It should not be used merely to make the asset look busy or technically impressive.

Creative Candidate Generation And Selection

A generative visual workflow may create several creative candidates before final rendering.

Candidate generation is appropriate when:

creative interpretation materially affects quality

one output is unlikely to be reliable

a reviewer needs meaningful choice

the cost of a poor final asset is high

brand or client judgement is required

different visual approaches may communicate the same approved message

Candidate generation is not required when:

the output is a deterministic template render

the approved creative direction is already fixed

multiple candidates would add cost without useful choice

the content is low-risk and visually standardised

Candidate Set Requirements

Each candidate set should define:

Candidate Set ID:

Content Record ID:

Source Package ID:

Prompt Contract ID:

Prompt Contract Version:

Generation Model:

Model Version Where Available:

Generation Date:

Generation Number:

Candidate Count:

Candidate IDs:

Candidate Preview Locations:

Generation Reference Or Seed Where Available:

Cost:

Status:

Candidate Statuses

Candidate Generation

Candidate Set Ready

Candidate Review

Candidate Selected

Candidate Rejected

Regeneration Requested

Finalisation Pending

Candidate Selection Requirements

The reviewer must be able to:

view every candidate

identify each candidate clearly

select one candidate

reject individual candidates

reject the entire candidate set

record a selection reason where appropriate

request regeneration

identify a brief or prompt defect

Candidate Selection Rule

Only an authorised reviewer may select the candidate that proceeds to finalisation.

Selection must not occur automatically merely because a candidate was returned first or assigned the first position.

Whole-Set Rejection And Regeneration

A reviewer may reject the full candidate set when:

no candidate meets the brief

the visual direction is wrong

the subject is inaccurate

brand rules were not followed

the result is misleading

quality is inadequate

text or logo rendering is unusable

rights or safety concerns arise

Regeneration must preserve:

original content record

approved source package

fixed prompt contract

brand rules

rights restrictions

previous candidate set

review decision

rejection reason

generation count

Regeneration may change:

approved variable inputs

creative emphasis

allowed camera direction

allowed motion direction

approved descriptive detail

Regeneration may not change approved claims or source facts without returning to content review.

Regeneration Rule

Regeneration is a controlled workflow state, not an uncontrolled retry loop.

A maximum regeneration count or escalation threshold should be defined.

When repeated regeneration fails, the work unit should be:

returned for brief revision

routed to manual production

parked

cancelled

Selected-Candidate Finalisation

Only the selected candidate may proceed to finalisation.

Finalisation may include:

upscaling

resolution improvement

image-to-video transformation

motion generation

caption composition

text overlay

audio addition

brand finishing

aspect-ratio adaptation

export preparation

Selected Candidate Lineage

The system must preserve:

candidate_set_id

selected_candidate_id

selection_reason

reviewer

selected_at

regeneration_count

prompt_contract_version

generation_model

generation_reference

source_image_id where relevant

transformation_type

final_render_id

Creative Candidate Rule

Generate options where options improve judgement.

Do not generate options merely to create volume.

Image-To-Video Production Route

Image-to-video production begins with an approved still image.

The standard route is:

Approved Content Brief
→ Approved Source Image Or Selected Candidate
→ Motion Instructions
→ Image-To-Video Generation
→ Draft Motion Review
→ Caption And Overlay Composition
→ Visual Validation
→ Final Approval
→ Format Variants
→ Publishing Handoff

Image-To-Video Requirements

Define:

Source Image ID:

Source Image Rights Status:

Source Image Approval Status:

Motion Purpose:

Subject Motion:

Background Motion:

Camera Motion:

Duration:

Aspect Ratio:

Loop Behaviour:

Audio Requirement:

Caption Requirement:

Transformation Model:

Disclosure Requirement:

Image-To-Video Rule

The motion system must not transform the approved image in a way that creates:

a false event

a false action

a false endorsement

a misleading product demonstration

an unauthorised likeness performance

a materially different claim

Text-To-Video Production Route

Text-to-video production begins with an approved brief, script or structured prompt.

The standard route is:

Approved Content Brief
→ Approved Script Or Structured Prompt
→ Prompt Contract Application
→ Candidate Generation
→ Candidate Selection
→ Draft Video Generation
→ Validation
→ Final Approval
→ Format Variants
→ Publishing Handoff

Text-To-Video Rule

Text-to-video generation must remain constrained by the approved message, visual contract, duration, rights and disclosure requirements.

Rendering And Variant Logic

The rendering layer converts approved inputs, selected candidates and components into final media.

Possible output variants include:

16:9 landscape

9:16 vertical

1:1 square

short-form version

long-form version

captioned version

silent version

client-branded version

language variant

campaign variant

personalised version

data-period variant

Variant logic should define:

which elements remain fixed

which elements may change

text-length limits

image-size requirements

chart limits

duration limits

resolution

frame rate

audio requirements

safe zones

export naming

output destination

selected creative reference

Rendering Rule

Different output formats should not be created through uncontrolled resizing.

Each format must preserve readability, hierarchy, timing, meaning and brand integrity.

Format variants may be created only after the creative direction or candidate has been approved.

Validation And Quality Review

Programmatic output requires content validation, creative candidate review, visual validation, audio validation and data validation where applicable.

Content Validation

Check:

script accuracy

source grounding

claim support

metric accuracy

spelling

grammar

call-to-action accuracy

disclosure requirements

legal or compliance requirements

client approval

Creative Candidate Validation

Check:

candidate aligns with brief

candidate depicts the correct subject

candidate does not invent unsupported events

candidate does not create misleading realism

candidate follows brand direction

candidate preserves approved message

candidate is suitable for finalisation

candidate rights are clear

Visual Validation

Check:

no clipped text

no overlapping elements

readable captions

correct logo

correct colours

correct aspect ratio

correct animation timing

smooth transitions

chart readability

sufficient contrast

correct image crop

no broken assets

no blank frames

no rendering artefacts

no malformed hands, faces or objects where relevant

no unapproved visual claims

no misleading before-and-after treatment

Audio Validation

Check:

correct narration

correct synchronisation

acceptable audio level

no clipped audio

music licensing

caption alignment

pronunciation

required silence or pauses

voice permission

synthetic voice disclosure where required

Data Validation

Check:

metric source

date period

calculation

units

labels

chart scale

comparison basis

update timing

Validation Rule

A successful render is not proof of a correct video.

A selected candidate is not automatically a final approved asset.

Content Approval, Asset Approval And Publishing Approval

MWMS recognises three separate approval decisions:

Content Approval

Confirms that the script, claims, source use, offer and call to action are acceptable.

Asset Approval

Confirms that the selected and rendered visual asset is accurate, usable, on-brand and suitable.

Publishing Approval

Confirms that the approved asset may be released to the specified destination at the specified time.

Approval Separation Rule

Content approval does not automatically approve the visual asset.

Asset approval does not automatically authorise publication.

Publishing And Distribution Control

Publishing is a separate authority from production.

Possible destinations include:

website

landing page

YouTube

Instagram

Facebook

LinkedIn

TikTok

email

client portal

internal dashboard

presentation

advertising platform

Each publishing handoff should define:

final asset

destination

format

title

description

thumbnail

captions

call to action

tracking

disclosure

rights

publish date

owner

approval status

Publishing Rule

Rendering permission does not equal publishing permission.

No programmatic workflow may automatically publish public or client-facing content without the required approval.

Synthetic Media And Rights Governance

Every workflow involving generated people, voices, brands, characters or events must confirm:

likeness permission

voice permission

logo permission

brand permission

music licence

image rights

commercial-use rights

model-use restrictions

disclosure requirements

retention requirements

Rights Rule

No rights, no use.

No permission, no cloned voice or likeness.

No disclosure where disclosure is required, no release.

Performance Measurement

Programmatic video should be measured by business outcome.

Possible metrics include:

Production Metrics

time to produce

manual editing time saved

rendering success rate

revision rate

output cost

component reuse rate

failed render rate

candidate generation cost

candidate selection rate

regeneration rate

average generations before approval

Content Metrics

view rate

completion rate

watch time

engagement

click-through rate

conversion

lead quality

sales use

client acceptance

report comprehension

Business Metrics

revenue influenced

production cost reduced

client value created

campaign performance improved

content volume increased without quality loss

time to update reduced

renewal support

productised service margin

Outcome Rule

More rendered videos do not automatically mean more value.

More candidates do not automatically mean better creative work.

Programmatic production should continue only when it creates useful, proportionate outcomes.

Learning And Template Improvement

Each production cycle should create learning.

Possible learning includes:

which hooks hold attention

which scene lengths work

which chart types are understood

which components fail

which text lengths cause layout problems

which aspect ratios perform best

which calls to action convert

which inputs are unreliable

which variants create unnecessary work

which candidates are consistently selected

which prompt contracts create strong candidates

which regeneration reasons repeat

which models perform reliably

which templates should be improved

which templates should be retired

Template Improvement States

Keep

Improve

Restrict

Split

Merge

Park

Retire

Learning Rule

Templates and prompt contracts should improve through evidence.

They should not accumulate indefinitely.

Programmatic Video Workflow

The standard MWMS workflow is:

Business Need Identified
→ Video Use Case Defined
→ Audience Defined
→ Source Inputs Identified
→ Data Authority Checked
→ Content Brief Created
→ Script Or Data Record Created
→ Evidence Validated
→ Prompt Contract Selected
→ Scene Plan Created
→ Components Selected
→ Brand Rules Applied
→ Candidate Generation Where Required
→ Candidate Set Ready
→ Candidate Review
→ Candidate Selected Or Regeneration Requested
→ Selected Candidate Finalised
→ Draft Render Created
→ Content Validation
→ Visual Validation
→ Audio And Data Validation Where Required
→ Human Review
→ Revisions
→ Final Asset Approval
→ Export Variants Created
→ Publishing Approval
→ Publishing Handoff
→ Performance Measured
→ Learning Recorded
→ Template Or Prompt Contract Updated Or Retired

Programmatic Video Types

MWMS recognises the following major programmatic video types.

Product Demonstration Video

Purpose:

Show how a product, application, tool or workflow operates.

Possible inputs:

product screenshots

screen recordings

feature data

approved script

interface labels

benefit statements

Data Summary Video

Purpose:

Convert metrics or reports into an understandable visual summary.

Possible inputs:

revenue data

campaign data

progress data

experiment results

operational metrics

client reporting data

Dynamic Dashboard Video

Purpose:

Create a recurring visual recap of dashboard information.

Possible inputs:

current metrics

previous-period metrics

alerts

notable changes

recommended actions

Educational Explainer

Purpose:

Teach a concept through reusable scenes and visual structures.

Possible inputs:

lesson content

steps

examples

diagrams

definitions

proof

Social Short-Form Video

Purpose:

Turn approved content into repeatable platform-specific videos.

Possible inputs:

hook

short script

captions

visual evidence

call to action

approved quote

approved source article

selected image candidate

Case-Study Video

Purpose:

Present a verified problem, intervention and result.

Possible inputs:

client permission

baseline

solution

result

proof

testimonial

disclaimer

Comparison Video

Purpose:

Compare two or more options, states or results.

Possible inputs:

comparison criteria

metrics

evidence

visual assets

conclusion

Personalised Video

Purpose:

Create approved variants for a client, segment, account or use case.

Possible inputs:

recipient or segment name

relevant context

approved offer

account data

customised call to action

Personalisation Rule

Personalisation must remain relevant, respectful, evidence-backed and permissioned.

It must not expose private data or create deceptive one-to-one communication.

Image-To-Video Asset

Purpose:

Turn an approved still image or selected image candidate into a controlled motion asset.

Possible inputs:

approved image

motion instructions

caption

audio

platform format

brand rules

Text-To-Video Asset

Purpose:

Turn an approved script or structured prompt into generated video.

Possible inputs:

approved script

prompt contract

audience

duration

style

camera rules

brand rules

Source-To-Visual Asset

Purpose:

Turn approved articles, reports, podcasts, quotes or news into governed visual content.

Possible inputs:

source record

source date

event date

extracted insight

approved post copy

candidate-generation prompt

platform requirements

Source-To-Visual Rule

A source-driven asset must not imply that the generated visual is documentary evidence unless it is genuine source media.

Content Brief Requirements

A Programmatic Video Content Brief should define:

Video Title:

Video ID:

Business Purpose:

Audience:

Use Case:

Primary Message:

Viewer Action:

Source Records:

Approved Claims:

Required Proof:

Format:

Duration:

Aspect Ratio:

Platform:

Brand:

Tone:

Voice:

Caption Requirement:

Audio Requirement:

Data Requirement:

Personalisation Requirement:

Candidate Generation Required:

Candidate Count Range:

Selection Authority:

Regeneration Limit:

Disclosure Requirement:

Rights Requirement:

Publishing Owner:

Success Metric:

Content Brief Rule

No content brief, no production.

No approved source, no trusted content.

Production Work Unit

Each programmatic video should have a work unit containing:

work_unit_id

content_record_id

video_id

video_title

business_purpose

audience

use_case

source_ids

source_validation_status

script_id

script_version

prompt_contract_id

prompt_contract_version

scene_plan_id

template_id

template_version

brand_profile_id

candidate_generation_required

candidate_set_id

candidate_count

selected_candidate_id

selection_reason

candidate_reviewer

candidate_selected_at

regeneration_count

generation_model

generation_model_version

generation_reference

source_image_id

transformation_type

final_render_id

aspect_ratio

duration

resolution

audio_status

caption_status

content_validation_status

visual_validation_status

audio_validation_status

data_validation_status

client_review_status

human_approval_status

publishing_owner

publishing_destination

publishing_status

render_status

render_error

performance_metrics

outcome_status

created_at

updated_at

No technical build is authorised by this framework alone.

Failure Handling

Possible failure types include:

missing source

unverified source

stale data

unsupported claim

prompt failure

candidate generation failure

no acceptable candidate

repeated regeneration

render failure

broken asset

layout failure

caption failure

audio failure

rights failure

approval failure

publishing failure

Failure States

Retry Generation

Regeneration Requested

Return To Brief

Return To Content Review

Manual Production Required

Awaiting Rights Confirmation

Awaiting Client Approval

Parked

Cancelled

Escalated

Failure Rule

A repeated automated failure should not create an endless loop.

The work unit must move to a human-readable failure state with the reason preserved.

Cost And Volume Control

Dynamic visual systems may create meaningful API, generation, storage and review costs.

Each workflow should consider:

cost per candidate

cost per candidate set

cost per render

cost per regeneration

storage cost

review time

publishing cost

expected business value

Cost Rule

Generation volume must be proportionate to expected value.

High-cost models should not be used automatically where lower-cost methods can meet the approved quality threshold.

Governance Role

Content Brain owns the MWMS Programmatic Video And Dynamic Visual Content Framework.

Content Brain is responsible for:

business purpose

audience

message

content brief

story

script

source use

content approval

candidate-review requirements

visual production requirements

Content Brain does not automatically own:

brand authority

data authority

automation infrastructure

client permission

publishing authority

commercial approval

Supporting Brain Responsibilities

Brand Brain

Owns:

brand identity

visual standards

logo treatment

motion style

voice and appearance rules

client brand boundaries

Data Brain

Owns:

data definitions

metric authority

calculation integrity

source records

freshness

lineage

Automation Brain

Owns:

workflow reliability

triggers

queues

retries

failure handling

monitoring

tool connections

AIBS Brain

Owns:

client packaging

client boundaries

service scope

support expectations

commercial delivery design

HeadOffice Brain

Owns:

cross-brain priority

strategic fit

risk escalation

governance conflict

investment and scale decisions

Human Authority

Human approval remains required when:

claims are high risk

the content is client-facing

the content uses likeness or voice

the subject is sensitive

a creative candidate must be selected

the visual may be mistaken for real evidence

the asset will be publicly released

the commercial impact is material

Client And AIBS Use

This framework may support client services such as:

programmatic social video

dynamic reporting video

product demonstration production

dashboard recap services

AI visual content production

image-to-video services

campaign creative variation

client-specific video assets

visual sales letters

landing-page video production

Client work must define:

scope

input responsibilities

approval responsibilities

brand requirements

rights and permissions

candidate count

revision and regeneration limits

publishing authority

support boundary

data boundary

retention

pricing basis

Client Boundary Rule

MWMS must not absorb unlimited creative regeneration into a fixed scope without defined limits.

Relationship To Content Production System

Content Brain Content Production System Framework governs the broader content lifecycle.

This framework defines the specialised production logic for programmatic video and dynamic visual media.

The relationship is:

Content Need
→ Content Brief
→ Programmatic Video Work Unit
→ Script And Scene Plan
→ Candidate Generation Where Required
→ Candidate Selection
→ Render
→ Validation
→ Publishing Readiness
→ Distribution
→ Performance Learning

Relationship To Content Repurposing

Content Brain Content Repurposing Framework determines when existing source material should become a new format.

This framework governs how approved source material becomes controlled motion and dynamic visual output.

Relationship To Market Driven Social Content

MWMS Market Driven Social Content Production Framework governs:

market relevance

buyer relevance

platform purpose

content idea

content angle

This framework governs:

visual generation

candidate selection

motion production

caption composition

format variants

final asset control

Relationship To Prompt Architecture

MWMS Prompt Architecture And Automation Output Reliability Framework governs the broader prompt contract, output schema and reliability standards.

This framework applies those standards specifically to dynamic visual and video production.

Relationship To Publishing Readiness

Content Brain Publishing Readiness Checklist determines whether the complete content asset is ready for release.

This framework does not replace publishing readiness.

It supplies the validated visual work unit and final render evidence required by that decision.

Relationship To Personalised Visual Sales Assets

Personalised sales assets intended for prospect outreach remain governed by the MWMS Personalised Visual Sales Asset Production And Governance Framework.

This page governs broader Content Brain production, editorial visual media, social video and dynamic content systems.

Minimum Viable Implementation

A minimum viable programmatic video system should include:

one clear business purpose

one approved audience

one content brief

one approved source package

one prompt contract or deterministic template

one controlled production route

candidate generation only where justified

one candidate review decision

one final render

content validation

visual validation

human approval

publishing handoff

performance measurement

The minimum viable system does not require:

many tools

many templates

fully autonomous publishing

large candidate sets

complex multi-agent orchestration

unlimited variants

Implementation Order

Recommended order:

Define Business Purpose

Define Audience

Define Source Authority

Create Content Brief

Define Prompt Contract Or Template

Define Scene Architecture

Define Brand And Motion Rules

Define Candidate Requirement

Define Selection And Regeneration Authority

Create Draft Output

Validate Content

Validate Visuals

Approve Final Asset

Create Necessary Format Variants

Approve Publishing

Measure Outcome

Improve Or Retire The System

Decision Checklist

Before approving a programmatic video or dynamic visual system, ask:

What business purpose does it serve?

What action should it create?

What information does it use?

Is the information verified?

Which elements repeat?

Which elements vary?

Which brand rules apply?

Which components are approved?

Which data period is shown?

Does the asset require creative candidates?

Who selects the candidate?

What happens when all candidates fail?

How many regeneration cycles are allowed?

Is prompt and source lineage preserved?

Is the final render linked to the selected candidate?

Who approves the content?

Who approves the asset?

Who approves publishing?

What rights and permissions apply?

How will performance be measured?

Architectural Intent

The architectural intent of the MWMS Programmatic Video And Dynamic Visual Content Framework is to create a controlled bridge between Content Brain, Brand Brain, Data Brain and Automation Brain.

The long-term system should allow MWMS to move from:

manual one-off production

to

reusable, source-controlled, brand-governed and outcome-measured visual production

without losing:

human judgement

creative accountability

source evidence

candidate lineage

rights control

publishing authority

Final Position

The MWMS Programmatic Video And Dynamic Visual Content Framework converts video from a purely manual editing activity into a governed content-production system.

It establishes:

business-purpose control

audience and use-case alignment

source and data governance

content architecture

prompt contract and variable-input control

scene architecture

reusable visual components

brand and motion rules

creative candidate generation

candidate selection

controlled regeneration

selected-candidate lineage

image-to-video and text-to-video routes

rendering and format-variant logic

content, data, visual and audio validation

separate content, asset and publishing approval

publishing separation

performance measurement

template and prompt improvement

client and AIBS use boundaries

The key shift is:

MWMS should not automate video merely to produce more media.

MWMS should use programmatic production where reusable structure, changing data, controlled visual systems and governed creative selection create measurable business value.

Final Rule

Video should become reusable where repetition creates value, but remain human-controlled where judgement, claims, emotion, rights or brand risk matter.

No business purpose, no production system.

No approved source, no trusted content.

No verified data, no dynamic chart.

No content brief, no render.

No prompt contract, no controlled generation.

No brand rules, no scalable template.

No acceptable candidate, no finalisation.

No selected-candidate lineage, no trusted final asset.

No validation, no publishing readiness.

No rights, no use.

No human approval, no high-risk release.

No publishing approval, no public release.

No measured outcome, no justified scale.

Change Log

Version: v1.1
Date: 2026-06-27
Author: HeadOffice

Change:

Updated the MWMS Programmatic Video And Dynamic Visual Content Framework using the AI Automations by Jack Image + Video block covering:

VEO text-to-video generation

image-to-video production

AI-generated social video

fixed creative contracts and variable prompt inputs

creative candidate sets

candidate review dashboards

candidate selection

whole-set rejection

controlled regeneration

selected-candidate upscaling and finalisation

caption and overlay composition

platform-specific visual output

source, prompt, candidate and final-render lineage

The update adds:

Creative Candidate definitions

Candidate Set definitions

Creative Candidate Versus Format Variant Rule

Prompt Contract And Variable Input Control

Creative Candidate Generation And Selection

formal Candidate Statuses

Whole-Set Rejection And Regeneration

Selected-Candidate Finalisation

Selected Candidate Lineage

Image-To-Video Production Route

Text-To-Video Production Route

Creative Candidate Validation

separate Content Approval, Asset Approval And Publishing Approval

expanded work-unit metadata

generation and regeneration cost controls

expanded workflow and decision checklist

The existing programmatic video, reusable component, data governance, rendering, validation, publishing, performance and client-delivery architecture was preserved and expanded.

Purpose of update:

To ensure Content Brain can govern generative visual production as a controlled creative process rather than treating the first generated output as final, while preserving source accuracy, prompt control, candidate lineage, human selection, rights, validation and publishing authority.

Change Impact Declaration

This v1.1 update expands the existing programmatic video framework.

It does not create a separate visual candidate approval page.

It does not authorise a technical build.

It does not authorise autonomous publishing.

It does not replace the broader Content Brain production, repurposing, social content, brand, prompt or publishing frameworks.

Pages Created

None

Pages Updated

MWMS Programmatic Video And Dynamic Visual Content Framework

Pages Deprecated

None

Standalone Pages Not Created

MWMS AI Visual Content Production And Approval Framework

MWMS Creative Candidate Selection Framework

MWMS Image-To-Video Production Framework

MWMS Text-To-Video Production Framework

MWMS Midjourney Approval Framework

MWMS Visual Reroll And Regeneration Framework

These concepts were absorbed into the unified tool-agnostic programmatic video and dynamic visual content framework rather than created as separate pages.

Registries Requiring Update

Content Brain Page Registry

Required Registry Change

Update the existing MWMS Programmatic Video And Dynamic Visual Content Framework entry from v1.0 to v1.1 and record the addition of creative candidate generation, candidate selection, controlled regeneration, selected-candidate lineage, image-to-video and text-to-video production routes.

Canon Version Update Required

No

Change Log Entry Required

Yes

Strategic Absorption Result

MWMS gains a governed framework for turning approved content, data, images and brand assets into reusable video and dynamic visual media while generating creative options where useful, preserving candidate and prompt lineage, requiring authorised human selection, controlling regeneration, separating content, asset and publishing approval, and measuring business value before scale.

END OF FULL FILE OUTPUT