No-install starter tier · v0.1 · workflow clarity before setup In stock

I spent two months rebuilding
my AI-agent system.
So you don't have to.

A 12-document ZIP for building one manual-first AI workflow in 7 days.

Choose this $49 Blueprint if you only need workflow clarity and do not want to install or configure tools yet. Choose the $297 Setup Kit if you are ready to build the actual Mac operator workspace and use first-install issue tracking and create workflow guardian handoffs for workflows that prove themselves.

Request $49 Blueprint checkout → View v0.2 setup kit

$49 no-install starter tier. Manual email checkout. ZIP delivered after payment. No Mac setup kit, NotebookLM Learning Pack, local coordinator layer, or workflow guardian handoff plus first-install issue tracking.

Selection Worksheet
02 · MD · 4 PAGES
Red-Lines Template
04 · MD · 3 PAGES
Sprint Tracker
07 · MD · 8 PAGES
Verification
06 · MD · 2 PAGES

The buyer pack

blueprint.zip

12 documents · PDF + Markdown · 1.4 MB · delivered manually after payment

$49

Preview · 12-document no-install starter pack · manual delivery after payment

$49No-install starter tier
12Documents in pack
7 daysRefund window
0Subscriptions

Package choice · why this is cheaper than the setup kit

Choose the $49 Blueprint for clarity. Choose the $297 Setup Kit for the actual Mac workspace.

$49

No-install Blueprint

Use this when you only need to pick one workflow, define red lines, build a source-of-truth folder, and run a 7-day manual test.

No tools required
$297

Mac AI Operator Setup Kit

Use this when you are ready to build the Mac workspace with Claude Code, Codex, VS Code, GitHub, NotebookLM, local operator guidance, local reports, review queues, update-cascade docs, QA gates, a first-install issue tracker, and a workflow guardian handoff for the established workflow.

View the $297 Setup Kit → Primary build path

This $49 tier does not include the Mac setup path, NotebookLM Learning Pack, local coordinator layer, workflow guardian handoff, local report/review-queue system, update-cascade docs, or operator workspace buildout.

Inside the pack · 12 documents · 1 sitting

No filler, no fake case studies, no "as featured in" badges. Just the pack.

01

Blueprint Guide

The agent reality check. Stack ladder. Safety layer. Seven-day plan. Dylan's mistake map.

PDF · 24 pages
02

Selection Worksheet

10-task table. Score on repetition, clarity, data safety, value, reviewability. Pick one workflow before anything else.

MD · 4 pages
04

Red-Lines Template

No secrets, no posting without approval, no deletes, no spending, no certainty where it is guessing.

MD · 3 pages
06

Verification Checklist

Eight pass/fail checks. Six safety-fail conditions. Approval-owner field. Each output is a draft until a human approves it.

MD · 2 pages
07

7-Day Sprint Tracker

Day-by-day deliverables, score sheet, end-of-week verdict. Keep manual, improve, automate later, abandon, or get expert help.

MD · 8 pages
12

Source-of-Truth Folder

GOAL · CONTEXT · TASKS · PROMPTS · VALIDATION · RED-LINES · EXAMPLES · HANDOFF. The eight files every workflow needs.

Folder template · 8 files

+ prompt pack · troubleshooting tree · completed example workflow · glossary · refund + expectations

The 7-day sprint · boring on purpose

Seven days. One verdict. No magic.

Start boring. Start safe. Start manual. The first win is repeatability, not autonomy.

Day 01
Pick one workflow
Day 02
Source-of-truth folder
Day 03
Run one real task
Day 04
Improve instructions
Day 05
Add verification
Day 06
Test three examples
Day 07
Decide

Why this exists

Cape Town · 2026 DW
"I spent two months trying to build a real AI-agent system on my Mac Mini. I broke it, rebuilt it, overbuilt it, locked it down, and finally got the map. This starter Blueprint packages the workflow clarity layer; the v0.2 Setup Kit is the fuller Mac operator path."

— Dylan Wichmann, founder of DWAI

Mistake map

  • memory without runtime surfaces fails
  • crons without action routing become noise
  • old docs poison new routes
  • credentials need strict gates
  • local services must bind to loopback
  • the first product should be clarity, not infrastructure

Safety layer · the draft-only rule

Your first agent must not do any of these.

Draft first. Human approves.

Every output is a draft until a human reviews and approves it. The AI may suggest. The human decides.

"If your AI setup has no owner, no red lines, and no verification, it is not an agent — it is chaos."

No secrets in chat.
API keys, passwords, cookies, recovery codes, auth files.
No public posting without approval.
Emails, DMs, posts, invoices, customer messages stay drafts.
No deletes, spending, deploys.
The agent suggests. The human acts.
No private scraping or cookie automation.
No bypassing platform limits.
No exposing local services.
Loopback only unless deliberately exposed and reviewed.
No certainty where it is guessing.
Mark assumptions. Flag missing context.

How delivery works · manual, by me

No subscription. No upsell sequence. Just the pack.

Step 01

Email BLUEPRINT

Request access from admin@dylanwichmannai.com. Mention the workflow you have in mind if you want a fit-check.

Step 02

Manual payment link

I send a payment link or invoice for $49. Pay in your own time. No subscription. No upsell sequence.

Step 03

blueprint.zip

PDF + 12 markdown templates. Delivered by hand. 7-day refund if it isn't useful.

7-day refund. Manual delivery. No outcome promises. This no-install tier stays separate from the $297 Mac AI Operator Setup Kit v0.2.

Stop rebuilding.
Start repeating.

$49 no-install starter tier. Manual checkout. ZIP delivered after payment. 7-day refund. No outcome promises.

Request $49 Blueprint checkout → View v0.2 setup kit

$49 no-install starter tier · separate from the $297 setup kit · 7-day refund · manual ZIP delivery

A validated workflow can become a workflow guardian only through owner-approved promotion.