Your path through the academy
Agentic AI Automation Academy
A beginner-friendly academy for building safe agentic AI automation service offers with Claude Code, Codex, OpenClaw lab demos, templates, and a final project.
Start with the open lessons
Use the two free lessons to understand the service map and safe demo shape.
Work module by module
Open one lesson, finish the action panel, save the proof note, then continue.
Package one offer
Turn the saved proof into a scoped first paid-service conversation.
Each module states what the student should be able to explain, build, or decide.
Lessons point toward a concrete demo, workbook entry, package, sales asset, or delivery step.
Checkpoints and final-project artifacts make progress visible before the student moves on.
Learning path
Modules, proof points, and lesson access
Module 1
Free Preview
Use two complete lessons to map a buyer workflow and sketch a safe demo before deciding whether to enroll.
Module 1
Free Preview
Use two complete lessons to map a buyer workflow and sketch a safe demo before deciding whether to enroll.
You will build a first service map: one business type, one repeated workflow pain, one safer assistant output, one human approval boundary, and one plain-English offer sentence.
Your fake inquiry, demo output shape, safety note, and one-sentence business explanation.
Without rereading, write the main decision, artifact, or boundary you can now explain from Free Preview.
The Agentic AI Service MapDo not try to understand every AI-agent tool feature yet. Your goal is to see how a small workflow demo can become a responsible beginner service without making risky promises.freeBuild Your First Sellable DemoBuild a demo preview that shows the business value without connecting real accounts, exposing customer data, or promising automatic results.free
Module 2
Start Here
Choose a first-offer goal, learn the course rhythm, and start the artifact trail that becomes your final project.
Module 2
Start Here
Choose a first-offer goal, learn the course rhythm, and start the artifact trail that becomes your final project.
You will understand the course path, the workbook, and the final project before you start.
Your rough Final Project Snapshot.
Without rereading, write the main decision, artifact, or boundary you can now explain from Start Here.
Module 3
Agentic AI Service Foundations
Understand what a buyer is paying for, how the tool stack supports the work, and where safety and client-trust boundaries begin.
Module 3
Agentic AI Service Foundations
Understand what a buyer is paying for, how the tool stack supports the work, and where safety and client-trust boundaries begin.
You will describe your agentic AI automation service in business language.
Your safety boundary and client trust sentence.
Without rereading, write the main decision, artifact, or boundary you can now explain from Agentic AI Service Foundations.
What You Are Really SellingTurn 'I build AI agents' into a specific workflow outcome.LockedClaude, Codex, And OpenClaw Setup BasicsBe able to point to each part of the system and say what it does.LockedSafety, Privacy, And Client TrustKnow what the assistant is allowed to do and what it must never do.Locked
Module 4
Claude/Codex Setup And OpenClaw Lab
Prepare the tools, verify a local lab, build a fake-data demo, troubleshoot it, and document a client-safe boundary.
Module 4
Claude/Codex Setup And OpenClaw Lab
Prepare the tools, verify a local lab, build a fake-data demo, troubleshoot it, and document a client-safe boundary.
You will explain OpenClaw in beginner language and understand the setup path.
A short checkpoint note naming the completed proof, any blocker, and the artifact you will carry into the final project.
Without rereading, write the main decision, artifact, or boundary you can now explain from Claude/Codex Setup And OpenClaw Lab.
OpenClaw Lab: What It Is And How It WorksReplace mystery with a simple mental model.LockedPrerequisites And AccountsKnow what you need before copying install commands.LockedInstall The OpenClaw Lab Step By StepMake the `openclaw` command work in your terminal.LockedRun OnboardingConnect one model provider and create a local Gateway setup.LockedVerify The GatewayKnow whether the engine of the system is on.LockedOpen The Dashboard And Send First MessageProve that the setup can respond before building the lead demo.LockedSetup Rescue: Top 10 Beginner Setup ErrorsKnow what to do when install, onboarding, Gateway, dashboard, or auth does not work.LockedBuild The First Local Agentic AI DemoTurn a fake lead inquiry into a structured response an owner can approve.LockedConnect One Beginner Lab ChannelUnderstand that channels are optional for the first client-ready offer.LockedTesting And TroubleshootingFind weak spots before a client finds them.LockedFrom Demo To Client-Safe Agentic SystemSeparate a practice demo from a deliverable service.LockedModule 2 Checkpoint: Setup And First SystemPause, inspect your proof, document blockers honestly, and make sure the next module starts from evidence instead of vague progress.Locked
Module 5
Choose A Sellable AI Service
Compare beginner-friendly offers, choose a reachable niche, and identify one repeated workflow worth improving.
Module 5
Choose A Sellable AI Service
Compare beginner-friendly offers, choose a reachable niche, and identify one repeated workflow worth improving.
You will choose a realistic beginner offer.
Workflow pain map.
Without rereading, write the main decision, artifact, or boundary you can now explain from Choose A Sellable AI Service.
Module 6
Build Your First Agentic AI Systems
Build lead, inbox, FAQ, and reporting assistants, then turn the strongest result into a reviewable demo portfolio.
Module 6
Build Your First Agentic AI Systems
Build lead, inbox, FAQ, and reporting assistants, then turn the strongest result into a reviewable demo portfolio.
You will adapt the lead follow-up demo to your niche.
A short checkpoint note naming the completed proof, any blocker, and the artifact you will carry into the final project.
Without rereading, write the main decision, artifact, or boundary you can now explain from Build Your First Agentic AI Systems.
System 1 - Agentic Lead Follow-Up AssistantCreate the main system for your final project.LockedSystem 2 - Agentic Inbox Triage AssistantShow a second useful workflow without connecting a real inbox.LockedSystem 3 - Agentic FAQ Support AssistantPractice support drafting without hallucinated policies.LockedSystem 4 - Agentic Weekly Report AssistantShow a useful internal reporting workflow.LockedModule 4 Checkpoint: Demo PortfolioPause, inspect your proof, document blockers honestly, and make sure the next module starts from evidence instead of vague progress.Locked
Module 7
Package And Price The Service
Define deliverables and exclusions, practice a starter price, write a simple proposal, and pass the package checkpoint.
Module 7
Package And Price The Service
Define deliverables and exclusions, practice a starter price, write a simple proposal, and pass the package checkpoint.
You will package your demo as a bounded beginner service.
A short checkpoint note naming the completed proof, any blocker, and the artifact you will carry into the final project.
Without rereading, write the main decision, artifact, or boundary you can now explain from Package And Price The Service.
Turn A Demo Into A Service PackageMake the offer clear enough to sell responsibly.LockedPricing Your First ProjectsPrice the work without pretending you are an expert agency.LockedCreate A Simple ProposalMake the next step clear without overselling.LockedModule 5 Checkpoint: Package, Price, ProposalPause, inspect your proof, document blockers honestly, and make sure the next module starts from evidence instead of vague progress.Locked
Module 8
Sell The Agentic AI Offer
Build a focused prospect list, write honest outreach, run discovery, and handle objections without risky promises.
Module 8
Sell The Agentic AI Offer
Build a focused prospect list, write honest outreach, run discovery, and handle objections without risky promises.
You will build the first outreach list.
Objection response sheet.
Without rereading, write the main decision, artifact, or boundary you can now explain from Sell The Agentic AI Offer.
Find Your First 25 Prospects, Then Build Toward 100Find enough real prospects to test demand safely.LockedOutreach ScriptsAsk for interest without sounding like hype.LockedDiscovery CallsLearn the workflow before selling the build.LockedHandling ObjectionsSell responsibly and know when to walk away.Locked
Module 9
Delivery, Handoff, And Next Offers
Onboard, test, hand off, support, identify a sensible next offer, and assemble the final buyer-readable project.
Module 9
Delivery, Handoff, And Next Offers
Onboard, test, hand off, support, identify a sensible next offer, and assemble the final buyer-readable project.
You will onboard a client safely.
Final project packet.
Without rereading, write the main decision, artifact, or boundary you can now explain from Delivery, Handoff, And Next Offers.
Client OnboardingCollect the right information without grabbing unnecessary access.LockedTesting And HandoffFinish professionally instead of just saying 'it works.'LockedRetainers And Next OffersAvoid selling recurring fees before value exists.LockedModule 7 Checkpoint: Delivery ReadyPause, inspect your proof, document blockers honestly, and make sure the next module starts from evidence instead of vague progress.LockedFinal Project: Your First Agentic AI Automation OfferProve you can explain, build or simulate, package, and safely sell one beginner agentic AI automation service.Locked
Module 10
Optional: Client Apps, Hosting, And Domains
An optional path for turning a proven workflow into a small client app. Students describe the buyer result while Codex handles Supabase, the app build, Vercel hosting, and later domain setup under human approval.
Module 10
Optional: Client Apps, Hosting, And Domains
An optional path for turning a proven workflow into a small client app. Students describe the buyer result while Codex handles Supabase, the app build, Vercel hosting, and later domain setup under human approval.
You will use Codex to turn one buyer problem into a small app you could demonstrate and sell as a fixed-scope pilot.
Save the domain and handoff plan as optional later-stage delivery proof.
Without rereading, write the main decision, artifact, or boundary you can now explain from Optional: Client Apps, Hosting, And Domains.
Plan A Small Client App With CodexDescribe the buyer and result; let Codex plan the screens, data, access, testing, hosting, and handoff.LockedLet Your Agent Set Up Safe App DataTell the agent who should see what. You do not need to write SQL or memorize Row Level Security rules.LockedLet Your Agent Build The Client App ScreenChoose what the buyer should see and do. Let the agent handle Next.js, forms, validation, and protected data access.LockedLet Your Agent Publish A Vercel Test LinkGet a safe review link without learning deployment commands or treating the preview as a finished launch.LockedConnect The Domain Later And Hand Off The AppUnderstand what you must approve while the agent handles DNS details, checks, documentation, and rollback notes.Locked
Module 11
Optional: Faceless YouTube Service Workflow
An optional money path for using Codex to research original topics, create reviewed production packets, organize a client dashboard, and sell a bounded service without promising views or monetization.
Module 11
Optional: Faceless YouTube Service Workflow
An optional money path for using Codex to research original topics, create reviewed production packets, organize a client dashboard, and sell a bounded service without promising views or monetization.
You will define a safe faceless YouTube automation lane that avoids reused-content, inauthentic-content, spam, misleading synthetic media, and no-review publishing traps.
Save the YouTube automation service package as an optional portfolio example beside your core agentic AI automation offer.
Without rereading, write the main decision, artifact, or boundary you can now explain from Optional: Faceless YouTube Service Workflow.
Faceless YouTube Automation Rules And Risk BoundariesBefore building anything, set the rules for what your system will not copy, fake, exaggerate, automate, or publish without human approval.LockedResearch A Video Topic And Build An Original Angle In CodexBuild a topic system that helps you find useful video ideas without copying competitor videos or chasing empty trend spam.LockedBuild The Faceless Production Workflow In CodexMake Codex the operating system for the work: it organizes files, drafts scripts, reviews checklists, catches missing proof, and helps you build tracker or dashboard code when the workflow needs an app.LockedPackage A YouTube Automation Service With A Client DashboardTurn the workflow into something a buyer can understand: what you create, what they approve, what is not included, and how success is reviewed without income guarantees.Locked
Keep it simple
One lesson, one artifact, one next step.
The course contains a lot of support material, but the path stays simple: open the next lesson, follow the start panel, save one buyer-readable proof note, and keep moving.
Go to next lessonOptional guidanceBuyer path, tool guide, and starter offersOpen this after you choose a module or when you need help packaging the work.
Learning experience
Built like a guided apprenticeship, not a video dump.
The academy uses practice-first lesson design, visible progress, artifact checkpoints, and screenshot-backed agent workflows so beginners always know what to do, what to save, and what not to overpromise.
Start with small practice loops
The first lessons ask for short, concrete work: a service map, fake-data demo sketch, setup note, and first evidence artifact.
Every lesson produces proof
Students save a decision, checklist, test note, demo output, proposal section, or handoff artifact that feeds the final project.
The app always shows the next move
Dashboard progress, module proof points, lesson checkpoints, and completion states reduce uncertainty while the student works.
Agents are taught as operating skills
Claude Code and Codex sit at the center: explore, plan, implement, review, test, document, and keep the human in control.
Visual proof beats tool hype
OpenClaw remains the practical lab for selected demos, supported by owner-created screenshots, safe captures, and official source notes.
Readable, repeatable lesson structure
Each lesson uses the same brief, mission, workbench, resource kit, reading, and completion flow so students can scan instead of hunt.
Confusion gets captured before completion
Lessons end with a final clarity check that asks students to explain the move, verify the proof, and route precise blockers to support.
Commercial course map
How the modules become a first paid offer.
The academy is organized as a project-based path: students choose a buyer pain, use Claude Code and Codex to build reviewable proof, use OpenClaw for safe lab evidence where useful, and turn the work into one small scoped offer.
Every module names the operational result a buyer can understand before tools are discussed.
Claude Code plans context and Codex inspects artifacts, tests, docs, and handoff quality.
Proof stays scoped, fake-data friendly, human-approved, and clear about what is not promised.
Free Preview
See the service category and test whether agentic AI automation is worth learning before paying.
- Claude/Codex move
- Use Claude Code for the service map and Codex for a quick review of what the demo must prove.
- Proof to save
- Service map and first sellable demo sketch
- Money action
- Name one workflow a small business might pay to have inspected, cleaned up, or automated.
Start Here
Set the working rhythm: one lesson, one artifact, one buyer-readable proof point.
- Claude/Codex move
- Use Claude Code to plan the lesson outcome and Codex to check the saved artifact for gaps.
- Proof to save
- Course-use plan and artifact habit
- Money action
- Choose the weekly cadence that gets a real offer artifact finished instead of collecting notes.
Agentic AI Service Foundations
Understand what buyers actually buy: clearer workflows, faster follow-up, safer handoff, and usable proof.
- Claude/Codex move
- Use Claude Code to compare service angles and Codex to pressure-test the operational steps.
- Proof to save
- Buyer pain, trust boundary, and service map
- Money action
- Pick the smallest credible service result you can explain without income promises.
Claude/Codex Setup And OpenClaw Lab
Create a working agentic AI lab so future demos can be shown safely and repeatably.
- Claude/Codex move
- Use Claude Code for setup planning and troubleshooting notes; use Codex to inspect commands, docs, and setup proof.
- Proof to save
- Working lab, setup receipts, and first safe demo
- Money action
- Turn the setup into proof that you can configure, test, and explain a client-safe workflow.
Choose A Sellable AI Service
Select a niche, painful workflow, and service lane that can be sold as a small scoped pilot.
- Claude/Codex move
- Use Claude Code to compare niches and Codex to inspect the workflow for missing data, permissions, and handoff steps.
- Proof to save
- Niche choice, painful workflow, and offer angle
- Money action
- Write the first buyer-safe offer sentence with a narrow result and clear exclusions.
Build Your First Agentic AI Systems
Create small demos for lead follow-up, inbox triage, FAQ support, reporting, and portfolio proof.
- Claude/Codex move
- Use Claude Code to design each workflow and Codex to review outputs, edge cases, test notes, and reusable assets.
- Proof to save
- Demo portfolio with tests and captions
- Money action
- Pick the strongest demo and translate it into a paid pilot conversation.
Package And Price The Service
Turn the strongest demo into deliverables, scope, price logic, review boundaries, and a simple proposal.
- Claude/Codex move
- Use Claude Code to shape the package and Codex to inspect proposal clarity, missing risks, and handoff steps.
- Proof to save
- Package, price notes, and simple proposal
- Money action
- Prepare the first scoped offer without promising revenue, savings, or fully autonomous decisions.
Sell The Agentic AI Offer
Start honest conversations with prospects using proof, discovery questions, and a low-pressure next step.
- Claude/Codex move
- Use Claude Code to adapt outreach to buyer context and Codex to inspect scripts for hype, vagueness, and unsafe claims.
- Proof to save
- Prospect list, outreach script, discovery notes, and objection answers
- Money action
- Ask for one review call or paid pilot conversation, not a broad automation transformation.
Delivery, Handoff, And Next Offers
Onboard, test, hand off, and identify the next safe offer after the first project is delivered.
- Claude/Codex move
- Use Claude Code to prepare handoff docs and Codex to inspect test coverage, risks, and acceptance notes.
- Proof to save
- Client onboarding, test notes, handoff, and final project
- Money action
- Use delivery proof to propose the next small improvement only after the first scope is reviewed.
Optional: Client Apps, Hosting, And Domains
An optional path for turning a proven workflow into a small client app. Students describe the buyer result while Codex handles Supabase, the app build, Vercel hosting, and later domain setup under human approval.
- Claude/Codex move
- Use Claude Code for planning and Codex for inspection, testing, or reusable assets before saving proof.
- Proof to save
- Saved module artifact
- Money action
- Translate the module output into one buyer-safe next step.
Optional: Faceless YouTube Service Workflow
An optional money path for using Codex to research original topics, create reviewed production packets, organize a client dashboard, and sell a bounded service without promising views or monetization.
- Claude/Codex move
- Use Claude Code for planning and Codex for inspection, testing, or reusable assets before saving proof.
- Proof to save
- Saved module artifact
- Money action
- Translate the module output into one buyer-safe next step.
This commercial map is a work path, not an income claim: students build scoped proof, use fake data where needed, keep humans in approval, and make careful buyer-safe offers.
Market proof board
Students learn the tools buyers are already curious about.
The course should feel like an investment in marketable agentic AI skills: Claude Code for planning and buyer language, Codex for build/review/test evidence, and OpenClaw for selected fake-data lab proof.
$49.99 into a proof stackNo income promise. The point is to leave with assets a buyer can inspect.
Official Anthropic product imageClaude Code
Claude Code turns messy buyer problems into a plan.
Use Claude Code to read the workflow, clarify the buyer pain, define the safe scope, and turn the idea into language a prospect can understand.
Proof students build
Buyer workflow map, safe scope note, and first offer sentence.
Official OpenAI Codex app imageCodex
Codex turns the plan into reviewable delivery evidence.
Use Codex to create, inspect, test, and document the implementation so a buyer sees more than a prompt: they see a deliverable with checks.
Proof students build
Prototype note, test evidence, review summary, and handoff checklist.
Academy-owned lab capture based on OpenClaw docsOpenClaw
OpenClaw makes selected demos visible in a fake-data lab.
Use OpenClaw when a visible dashboard or gateway run makes the Claude/Codex plan easier to trust without exposing client data.
Proof students build
Sanitized lab screenshot, source note, approval boundary, and no-live-data disclaimer.
Official screenshots are references, not endorsements. Paid course screenshots should be owner-captured, source-noted, scrubbed, and kept private unless explicitly approved for public use.
Proof standards and screenshot sourcesOptional buyer trust review
Visual evidence roadmap
Screenshots should become buyer proof, not tool hype.
Every lesson has a planned screenshot, owned visual, or diagram target so students can show what Claude Code, Codex, and OpenClaw helped them inspect, build, verify, or safely demo.
48/48lessons mapped to visual proof
Start from a trusted source
Use official Claude Code, Codex, or OpenClaw references as context.
Capture one lesson action
Show one plan, review, demo output, or decision students can repeat.
Scrub private context
Hide keys, emails, customer names, paths, billing, repositories, and tokens.
Caption it for a buyer
Name the workflow pain, proof, approval point, and boundary.
high
Build Your First Sellable Demo
Students understand what a safe demo output looks like before connecting any real account.
Agentic AI Automation Academyacademy-ownedAcademy generated or owner-created democourse diagram
Open visualshigh
The Agentic AI Service Map
A beginner can see exactly how a vague AI-agent idea becomes one narrow service offer.
Agentic AI Automation Academyacademy-ownedAcademy generated diagramcourse diagram
Open visualshigh
Build The Faceless Production Workflow In Codex
Students understand Codex as the production command center for documents, checklists, review, and optional app/dashboard scaffolding.
Academy generated diagramcourse diagramOpenAIofficial reference
Preview lessonhigh
Build The First Local Agentic AI Demo
Students see the target output before building their own version.
OpenClawofficial referenceAcademy generated or owner-created democourse diagram
Preview lessonUse official web sources as references, owner-created captures for course proof, and generated diagrams when a real screenshot would leak private data or distract from the buyer-facing workflow.
The full source trail is maintained in the visual catalog and lesson-level proof panels. Official references are not endorsements; production lesson media should be academy-owned, source-noted, and scrubbed before paid-course use.
Tool tracks
Choose the path that matches why you came here.
Claude Code and Codex are the primary operating skills. OpenClaw stays visible as the applied lab where selected automation proof becomes concrete.
Buyer category
Agentic AI service path
Map one painful workflow into one safe buyer-readable service offer.
Primary operator skill
Claude Code path
Use Claude Code for orientation, setup reasoning, planning, checks, and implementation notes.
Primary build/review skill
Codex path
Use Codex for implementation, review, tests, docs, and reviewable demo evidence.
Applied lab proof
OpenClaw lab path
Use OpenClaw as the visible gateway and dashboard lab for selected fake-data demos.
Agentic stack
Market the whole agentic AI stack, not just one tool.
The academy leads with Claude Code and Codex as the coding-agent operating skills students can market, then uses OpenClaw as a practical automation lab where selected workflows become visible proof.
Agentic AI services
Students learn to sell workflow improvement: one painful process, one useful assistant output, one human approval point.
Claude Code
Claude Code is taught as a core operating skill for reading codebases, planning changes, fixing errors, running checks, and reviewing diffs under user control.
Codex
Codex is taught as a core build, review, testing, documentation, and multi-agent workflow skill for turning ideas into inspected software changes.
OpenClaw
OpenClaw stays in the course as a practical lab surface for selected workflow demos, gateway checks, dashboard evidence, and channel decisions.