SP-START-7 / printable worksheet pack

Seven days. Seven artifacts. No chaos.

A field-manual route for a beginner building an AI-assisted solo business safely: buyer, offer, stack, agent, content, customer, delivery.

Time7 short sessions
DifficultyBeginner-safe
OutputOne operating folder
AI levelDraft + organize only

Safety lock

AI may draft, organize, summarize, score, route, remind, compare, and create local files. Humans approve publish, send, buy, contact, deploy, delete, accounts, credentials, private data, and proof claims.

Day 01 / Guide 01

Buyer map

Who are you helping, and what annoying job do they want fixed?

Fill this card

Buyer type:
Specific person/business example:
Painful workflow:
Why it matters now:
What they already try instead:
One sentence version:

Safe AI help

Brainstorm examples, summarize notes, or make messy thoughts clearer.

Human approval lock

No fake proof, revenue, demand, testimonials, logos, or case studies.

Done means

A stranger understands who you help and what painful workflow you are studying.

Day 02 / Guide 02

Offer one-liner

What do you sell, and what gets better after someone buys it?

Formula

I help ____ fix ____ by delivering ____ so they can ____.

Before state:
After state:
Deliverable:
Boundary / not included:
Final one-liner:

Safe AI help

Draft ten plain-English offer sentences; you choose the clearest one.

Human approval lock

No guarantees, pricing promises, testimonials, public page, checkout, or outreach without review.

Day 03 / Guide 03

Smallest stack map

Which tools catch people, talk to them, sell the thing, do the work, and remember what happened?

LaneWhat must happenTool/manual methodKeep/remove/later
Catch
Talk
Sell
Do
Remember

Tool-bloat check

Approval lock

No accounts, subscriptions, purchases, API keys, live integrations, or credentials without approval.

Day 04 / Guide 04

Agent job card

What boring prep job can AI do while a human checks the result?

Agent job name:
Trigger:
Input it may read:
Output it must create:
Allowed actions:
Forbidden actions:
Source-of-truth file/folder:
Business rule AI must not guess:
Edge case:
Acceptance criteria:
Human review step:
Proof-loop check:

No sending, posting, deploying, contacting, buying, deleting, account changes, credentials, or private identity/payment/tax/KYC handling.

Day 05 / Guide 05

Proof / content route

Show one useful thing you learned without pretending it made money.

Reviewed work product:
Private details to remove:
Useful lesson:
Best format:
Draft headline:

Route

☐ Guide read
☐ Diagnostic
☐ Worksheet download
☐ Email draft
☐ Audit offer

Approval lock

Publish only after truth, privacy, proof, source, claim, and customer-permission review.

Day 06 / Guide 06

Customer tracker

When someone asks a question, write it down and decide the next helpful step.

DatePerson/sourceQuestion or signalObjectionNext stepFollow-up

Objection library

Approval lock

No automatic outreach, replies, proposals, CRM updates, customer promises, discounts, legal claims, or private customer data use without approval.

Day 07 / Guide 07

Delivery checklist

Do what you promised, check it, explain it, and save what you learned.

Delivery fields

Customer/result promised:
Inputs needed:
Work blocks:
Quality check:
Handoff note:

Finish checks

☐ Revision boundary clear
☐ Proof captured
☐ SOP updated
☐ Customer-facing items reviewed

Approval lock

Customer-facing work, proof, publishing, live changes, credentials, customer data, invoices, legal/payment claims, and final delivery stay human-approved.

Final page / approval gates

Humans approve the risk.

AI may do locally

Draft, organize, summarize, score, route, remind, compare, create local files, prepare checklists, and suggest options for review.

Approval required before

Publish, send, contact, buy, create accounts, accept terms, deploy, delete, change live systems, enter credentials, use private details, or make proof/revenue claims.

Manual proof rule

Do the workflow manually once. Observe what worked and what failed. Automate only the repeated prep after the process is visible.

Finish-line folder

stackpilot-start-here/
  01-buyer-map.md
  02-offer-one-liner.md
  03-smallest-stack-map.md
  04-agent-job-card.md
  05-proof-content-route.md
  06-customer-tracker.md
  07-delivery-checklist.md
  approval-gates.md

Sample filled artifacts / fictional

Tiny Bookkeeping Cleanup

This is a fictional teaching example. It is not proof, a case study, a testimonial, customer data, accounting advice, tax advice, or a revenue claim.

Buyer + offer

Buyer: solo home-service owner buried in receipts.

Pain: receipts, deposits, payments, and supply runs are scattered across apps and paper.

Offer: turn one messy month of receipts and payments into a clean owner report in seven days.

Safe AI role

AI may summarize approved local notes, draft category lists, flag missing information, and draft an owner-summary outline.

AI may not: log into accounts, contact customers, send emails, file taxes, move money, accept terms, or make final financial claims.

Seven artifact examples

DayFilled exampleApproval lock
01Owner-operator who dreads month-end receipts.No fake demand proof.
02One messy month → clean owner report.No guarantees or accounting claims.
03Local spreadsheet, folder, checklist, owner-summary doc.No account creation or credentials.
04Receipt-note organizer agent job card.Human checks every number and note.
05Receipt cleanup checklist content route.Publish only after truth/privacy review.
06Track “supplies disappear into the month” signal.No cold automation.
07Cleaned folder, missing-info list, summary, checklist.Customer-facing delivery is human-approved.