UIPrompt Library
Start Pro $14Account
Back to Income Lab
goalDraftReady now

Client onboarding and delivery system

Maps the current process, builds reusable artifacts, and proves the flow with fictional data before client use.

Setup capital
$0
Monthly cost
$0–100
Setup time
6–12 hours
Weekly effort
2–5 hours
First dollar
Unknown
Sales burden
None

User-selected scenario, not an expected outcome.

Every consequential external action pauses for explicit human approval.

Autonomy by business phase

Current level 3/5. Potential level 4/5 in 1–3 months.

1

Validate

Human-led

The owner defines the real service and constraints.

2

Build

AI-assisted

AI maps the process and drafts artifacts.

3

Acquire customers

Human-led

This system does not acquire customers.

4

Deliver

AI-executed with approval

AI can prepare approved communications and checklists.

5

Collect and operate

Agent-run with exception handling

Approved reminders and state updates can run with an exception queue.

AI responsibilities

  • Map steps
  • Draft forms and templates
  • Check completeness
  • Surface exceptions

Human responsibilities

  • Approve process and messages
  • Control client data and credentials
  • Resolve exceptions
  • Approve contracts and invoices

Approval gates

Approval applies to the exact action once. It is not blanket permission.

Explicit approval

Outreach

Approve every client-facing message before sending.

Deliver
Explicit approval

Use client data

Approve data fields, purpose, access, and retention before real use.

Deliver
Explicit approval

Connect credentials

The owner connects any external account; AI never receives raw secrets.

Build
Explicit approval

Contract

The human approves all terms and scope changes.

Collect and operate
Explicit approval

Invoice

The human reviews and sends every invoice.

Collect and operate
Runnable goal

Copy the complete asset

/goal

Goal: create a self-contained onboarding and delivery system for one existing service. Interview the owner about the signed scope, roles, inputs, outputs, deadlines, dependencies, communication rhythm, data sensitivity, invoice points, and known exceptions. Map the current path before proposing automation.

Design the minimum intake form, status model, file structure, kickoff agenda, client messages, delivery checklist, approval points, exception queue, and closeout handoff. Collect only data required for delivery. Never request or store raw credentials; the owner must connect any external account. Dry-run the full system with fictional data, including missing input, late approval, scope change, delivery failure, and cancellation. Revise until every state has an owner and next action. Wait for explicit approval before real client messages, data use, account connection, contract change, or invoice. Stop on missing authorization, unsafe credential handling, conflicting terms, or no exception owner.

MILESTONES
1. Map current service: Scope, roles, data, dependencies, and exceptions are explicit.
2. Build minimum system: Intake, states, templates, checklist, and ownership exist.
3. Dry-run failures: Fictional happy path and five exception states complete.
4. Owner handoff: Permissions, approvals, and operating guide pass review.

COMPLETION CRITERIA
1. Only necessary data is collected
2. No raw credential handling
3. Every state has owner and next action
4. Fictional dry run completed
5. Real external actions remain gated

BUDGET CEILING
$100 USD

EXTERNAL-ACTION GATES
1. Outreach — Explicit approval: Approve every client-facing message before sending.
2. Use client data — Explicit approval: Approve data fields, purpose, access, and retention before real use.
3. Connect credentials — Explicit approval: The owner connects any external account; AI never receives raw secrets.
4. Contract — Explicit approval: The human approves all terms and scope changes.
5. Invoice — Explicit approval: The human reviews and sends every invoice.

STOP CONDITIONS
1. Missing authorization
2. Unsafe credential design
3. Contract conflict
4. No exception owner
5. $100 ceiling reached

Evidence record

Projected scenarios never appear here as observed outcomes.

Evidence stateDraft

Goal has not completed a recorded fictional-data dry run.

Experiment count0 recorded

No model test date recorded

Supported modelOpenAI GPT-5

Catalog baseline; exact version must be recorded at first test

Stop and escalate

  • Data authorization is missing
  • Credential handling is unsafe
  • Contract and workflow conflict
  • No exception owner exists
  • Security incident
  • Sensitive data
  • Contract or billing dispute
  • Material scope change

Known failure modes

  • Automating a broken process
  • Collecting unnecessary data
  • Sending generic client messages
  • No exception path

Prerequisites and changelog

A service owner standardizing a repeated client workflow. · Create a clear intake, delivery, exception, and handoff system without exposing client data or credentials.

Before you start

  • Existing service scope
  • Owner who can approve client communications and data handling
  • Service operations
  • Client communication
  • Data handling basics

Changelog

  • 1.0.0 · 2026-07-12 — Initial launch workflow.