Back to the AI Operating System
SEE IT BUILT
What an AI Operating System looks like once it’s built.
Three examples, spotlighted one layer at a time. The first one is ours: the same system we teach, filled in for The Landing Pad.
THE WHOLE SYSTEM
How The Landing Pad runs on AI
This is our actual AI Operating System. Take it in as a whole first, then scroll and we’ll spotlight each piece.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
SPOTLIGHT 1 OF 5 · START HERE
Everything starts with the outcome.
Before any tool or prompt, we wrote down what we’re actually trying to achieve. It sits in the middle of the system, right next to the AI tool we use every day, and every layer gets checked against it.
WHAT TO NOTICE
One sentence, in plain language
Every chat starts from the same question
It decides how much is worth building
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
SPOTLIGHT 2 OF 5 · LAYER 1
Governance: what are the rules?
Our AI Stance fits on one page. It says which tools we use, what never goes into them, and where the human stops and the AI starts.
WHAT TO NOTICE
Every tool is approved, demo only, or never
CUI never goes into any AI tool
Garrett reviews everything before it reaches a client
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
SPOTLIGHT 3 OF 5 · LAYER 2
Context: what does it need to know?
AI knows everything except your business, until you tell it. This is where we told it. Our Source Profiles, the data it can use and how, and the systems it can reach.
WHAT TO NOTICE
Every data source has a rule for how it’s used
Our Source Profiles load into every chat
When sources disagree, a set order decides
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
SPOTLIGHT 4 OF 5 · LAYER 3
AI in our work: where the time comes back.
Most of our AI work already lives at Level 2. Repeat work gets mapped once, with a trigger, an AI step, and a human check, then saved so it runs the same way every time.
WHAT TO NOTICE
Level 1 is for quick, one-off asks
Saved workflows, from Quick Start follow-ups to website updates
Every workflow ends with a human check
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
SPOTLIGHT 5 OF 5 · LAYER 3
No agents yet. On purpose.
Level 2 covers what we need right now. When a real Level 3 need shows up, we bring in a technical partner, same as we do for clients. And every output, at every level, gets checked before it ships.
WHAT TO NOTICE
Draft, accuracy check, human review, ship
Right compared to what? The outcome we named.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
TLP’s AI Operating System
How The Landing Pad runs on AI
The same system we teach, filled in for us · October 2026
1
Governance
“What are the rules?”
The TLP AI Stance. Where the human stops and the AI starts.
Use Policy
APPROVED
Claude · primary tool
Our existing GPTs
Framer AI for the website
DEMO ONLY
ChatGPT
Gemini
Copilot
NEVER
CUI in any AI tool
Pricing or pipeline in anything for outside eyes
Past performance beyond the approved list
Unclear? Ask Garrett. One seat, one decision-maker.
Privacy Guidelines
CUI never goes into a prompt.
GOES IN AS-IS, ON THE TLP ACCOUNT
Navy work
Quick Start transcripts & intake
Design schedules
University consult
The Quick Start intake form tells participants TLP uses AI
The Revenue Tracker stays confidential
Data Security
BEFORE
Personal account
Old TLP chats stay 12 months, then get deleted
→
NOW
TLP Team account
All new TLP work happens here, and only here
One seat, Garrett’s. No shared info@ seat until someone else needs access.
Zoom and other tools are moving from personal email to the TLP identity.
Notion is the system of record. Working files live in iCloud.
Human Review
PROOF-GATE
Garrett reviews everything before it reaches a client.
STAYS A HUMAN DECISION
Pricing
Approved past performance
Client relationships
Final sign-off
Every draft is checked against the outcome, the Voice guide, and TLP copy rules.
2
Context
“What does it need to know?”
AI knows everything except TLP, until we tell it.
Knowledge
What we know, written down.
COMPANY
TLP AI Context Reference v4
SELF
Garrett’s Voice guide
CUSTOMER
Personas and the five Doors
BRAND
TLP Design System
GOVERNANCE
TLP AI Stance
Our Source Profiles
Data
What we have, and how it’s used.
History & Milestones
Approved only
Session Feedback Log
Proof points
Story Bank
Pull freely
Services & Methods
Pull freely
TLP Network (CRM)
Internal only
Pipeline
Never in writing
Revenue Tracker
Confidential
Kept in Notion, our system of record
Connections
What it can reach, with our OK.
Notion
Live records and databases
Framer
thelandingpad.com
Claude Project
Context Reference and Voice guide, always loaded
iCloud
Working files, attached as needed
Only where the AI Stance allows
EXAMPLES OF GOOD
Navy workshop series
BTB Cohort 1 deliverables
Quick Start Source Profile & Launch Pack
THE OUTCOME WE’RE TRYING TO ACHIEVE
A one-person consultancy that delivers fast, keeps client information safe, and makes sure everything that leaves TLP reflects our standards.
Every layer gets checked against this.
+ New chat
Chats
Projects
Files
PROJECT
The Landing Pad
TLP_AI_Context_Reference_v4.md
Garrett’s Voice.txt
How can I help you today?
What’s the outcome you’re trying to achieve?
Claude · TLP Team account · one seat
WHEN SOURCES DISAGREE
The cascade wins, in this order
1
AI Context Reference
2
Capabilities Statement
3
TLP Design System
4
Classification Codes
3
AI in Our Work
“What should it do?”
Most of our AI work already lives at Level 2. The goal: keep moving repeat work up from one-off prompts into saved workflows.
LEVEL 1
Prompts
One ask, done well.
WE USE IT FOR
Follow-up texts and emails while the window is warm
Social captions and LinkedIn posts
Tidying Garrett’s own words, not translating them
Quick research and thinking through a decision
Every prompt carries
Outcome
Context
Examples of good
LEVEL 2 · WHERE WE LIVE
Projects & Automations
Repeatable work, saved and mapped.
WORKFLOW
TRIGGER
AI STEP
HUMAN CHECK
Quick Start follow-up
Transcript or intake form arrives
Source Profile, Workflow Map, debrief, follow-up email
Garrett reviews before it reaches the participant
Module development
New or refreshed training module
Objectives, workshop flow, scenarios, slides, job aids
Garrett owns the content and the room
Utility outreach
New utility target
Research filings, two-page brief, first-touch email
Garrett checks every claim and sends it himself
Website updates
A page needs a refresh
Mockup first, then the Framer build
Garrett edits the mockup and publishes
LEVEL 3
Agents
None yet, on purpose.
Level 2 already covers what TLP needs: persistence and repeatability.
A real Level 3 need?
We refer to a technical partner, same as we do for clients.
Evaluate the outputs
Every level. Right compared to what? The outcome we named.
1 · Draft
Built from our Source Profiles
2 · Accuracy check
Facts against the Context Reference. Approved past performance only. No pricing.
3 · Human review
The Voice guide and copy rules
4 · Ship
Garrett sends it
Human First. AI Second.™
© 2026 The Landing Pad LLC. All rights reserved.
WANT TO SEE WHAT YOURS LOOKS LIKE?
Start with one task. We’ll build the first piece together.
The AI Quick Start is free, 45 minutes, and virtual. Bring a real task and leave with an AI Source Profile and a Workflow Launch Pack.
© 2026 The Landing Pad LLC. All rights reserved.
Human First. AI Second.™
The Building On Method™
© 2026 The Landing Pad LLC.
All rights reserved.
Human First. AI Second.™
The Building On Method™