A project workbook for generative AI

AI Studio Projects

From Prompt to Finished Work

Take a brief through drafting, editing, production, and review. Work with ChatGPT or Claude, prepare campaign assets, direct images and sound, and test an automation with sample data.

Get the workbook · $10 Read the free sample

Digital PDF and practice files

500workbook pages

48original projects

4learning tracks

$10one-time price in USD

The learning path

The four project tracks

PROJECTS 01–12

Prompting, learning and client work

Compare ChatGPT and Claude, turn notes into decisions, practice client conversations, and build a working plan.

PROJECTS 13–24

Brand, campaigns and spreadsheets

Turn audience evidence into ads, newsletters, a visual identity, and a spreadsheet you can audit.

PROJECTS 25–36

Image, voice, video and music

Test image parameters, plan edits and motion, and make an avatar script, narration, music, and captions work together.

PROJECTS 37–48

Coding, assistants and automation

Practice API contracts, coding, and draft-only automation in Zapier or Make, then build a local review packet in Python.

Search the curriculum

Find a project

Search projects Learning track

48 projects

Give a Prompt a Testable Brief (pages 9–18)

Plan the Day Around the Bottleneck (pages 19–28)

Build a Status Brief From Three Notes (pages 29–38)

Turn Reading Into Recall Practice (pages 39–48)

Use an Analogy and Find Its Limit (pages 49–58)

Rehearse a Boundary Conversation (pages 59–68)

Facilitate a Short Decision Meeting (pages 69–78)

Write a Handoff Someone Can Run (pages 79–88)

Run a Freelance Discovery Call (pages 89–98)

Price a Small Job With a Scope Fence (pages 99–108)

Convert Mixed Feedback Into a Revision Plan (pages 109–118)

Write a Portfolio Case With Evidence (pages 119–128)

Scope a paid creative pilot (pages 129–138)

Turn buyer interviews into a testable message (pages 139–148)

Build a product page from a claim ledger (pages 149–158)

Design an ad test with one moving part (pages 159–168)

Draft a newsletter with an eligibility check (pages 169–178)

Brief a logo that survives everyday use (pages 179–188)

Adapt a trend into a campaign you can deliver (pages 189–198)

Turn spreadsheet rows into traceable copy (pages 199–208)

Check a campaign dashboard before sharing it (pages 209–218)

Estimate whether an AI service pays for its time (pages 219–228)

Edit a case study around evidence and permissions (pages 229–238)

Deliver a launch kit with one source of truth (pages 239–248)

Direct an image around the layout (pages 249–258)

Test Midjourney parameters with a comparison sheet (pages 259–268)

Inpaint an object without drifting the scene (pages 269–278)

Extend a frame to fit a wide header (pages 279–288)

Produce a batch of photos with locked product details (pages 289–298)

Give a still image one controlled motion (pages 299–308)

Build a short avatar demonstration around the action (pages 309–318)

Scope permission before using a cloned voice (pages 319–328)

Direct text-to-speech with a pronunciation sheet (pages 329–338)

Shape an original music bed around speech (pages 339–348)

Time captions and describe an important visual cue (pages 349–358)

Assemble a multimodal launch kit (pages 359–368)

Turn a client request into a studio intake record (pages 369–378)

Build a reusable assistant instruction card (pages 379–388)

Extract a checked JSON brief (pages 389–398)

Review a Python patch against real examples (pages 399–408)

Sketch a safe OpenAI API adapter (pages 409–418)

Map a Zapier automation with test records (pages 419–428)

Route a Make scenario through two test branches (pages 429–438)

Reconcile duplicate triggers before they become tasks (pages 439–448)

Evaluate the assistant on costly mistakes (pages 449–458)

Send useful diagnostics without copying the brief (pages 459–468)

Prove that a handoff refers to the current version (pages 469–478)

Capstone: Build a local studio release packet (pages 479–488)

PROJECT 01

Give a Prompt a Testable Brief

A vague request often produces a polished answer with the wrong facts or a fuzzy next step. Build a brief that specifies who will read the reply, what facts they need, and how you will judge it. Use a fictional repair shop so you can compare ChatGPT and Claude with the same input. Evaluate the observable result. The better answer is the one a customer can act on without guessing.

Workbook pages 9–18

Look inside

Read a workbook page

Select a track to read a page from a worked project. The 12-page sample shows the opening project and retains its original page numbers.

Preview a track

Open the 12-page sample

Workbook preview from Project 1, Give a Prompt a Testable Brief — Project 01 · Give a Prompt a Testable Brief · Page 12

12-page worked portfolio

The Signal Grove launch

Follow one fictional launch from its approved facts and audience brief through copy, media, and a final handoff. Use the completed packet as a reference when assembling your own.

Before you start

Common questions

Where should a beginner start?

Start with the first track. It covers writing prompts, checking answers against sources, and comparing ChatGPT with Claude using the same exercise.

What do I need to run the exercises?

Use a chat assistant and a text editor for the prompt projects. Media projects can use your preferred image, video, or audio tool. The Python capstone runs offline and needs no API key. Some optional tools may require their own accounts.

What does the Python capstone do?

It turns fictional studio requests into a local review packet. It flags missing source material, unapproved channels, and conflicting requests before someone reviews the work.

What format do I receive?

You receive the 500-page digital PDF and a ZIP with copyable project inputs, a curriculum index, and the offline Python capstone.

The complete workbook

AI Studio Projects · $10

Get the full workbook, the 12-page Signal Grove case, and copyable practice files.

Open the Gumroad listing

USD base price. Applicable tax is shown at checkout.