Seven skills that take a product from idea to a rendered PDF with a buy button.
The Digital Product Factory is the toolchain behind a real, finished product: ideation, competitor teardown, content build, branded images, sales copy, and a sanitiser-safe Gumroad landing page. It ships with the complete product it built, so you can read the actual output before you trust the method.
Seven skills, a pipeline, an image engine, and one finished product.
Copy the skill files into your Claude Code commands directory. Keep the three folders together - the skills reference them by relative path.
- setup - run this first. Writes your business identity and keys into the right files.
- product-ideation - generate ranked product ideas for your niche, with a distribution-fit-first discipline.
- product-teardown - tear down a competitor product into a differentiate-not-clone build brief.
- product-build - build a complete product's content: vault, AI dispatcher prompt, workflow, and more.
- product-image - generate branded cover, thumbnail, and ad images. Canva, Gemini, or free Cloudflare lanes.
- sales-copy - write landing-page and ad copy, in two audience registers.
- gumroad-landing - render a sanitiser-safe custom Gumroad landing page with a working buy button.
The renderer product-build hands off to.
- Jinja2 templates, WeasyPrint, and pandoc
- Markdown content plus a YAML config in, PDF and DOCX out
- Hard-fails on missing identity rather than shipping placeholder branding
The Python package behind product-image.
- Surfaces, styles, naming, lint gate, cost guard, and manifest writing
- Three backends, so you can pick a paid lane or the free Cloudflare one
- Ships with its own test suite
The pack ships the product it built.
Most skill packs describe an output. This one includes a complete finished product made with these exact skills, so you can open the real artefacts and judge the quality bar before you build anything.
The actual folder you get, including the rendered PDF and DOCX the pipeline produced.
What that lets you check.
- Whether the writing is good enough to put your name on
- What a real config looks like when it is filled in properly
- How the sales copy and the product content line up
- What the renderer actually outputs, not a description of it
The worked example carries the author's branding on purpose. It is a sample of what a fully configured setup produces, not a template to copy.
Operators who already live in Claude Code.
Product builders
You ship digital products and your pipeline currently lives in pasted prompts and half-remembered steps. This makes it repeatable.
Agency and consultancy operators
You want a productised offer alongside client work, and the bottleneck is packaging, not expertise.
Skill-pack collectors
You buy workflows to read them. The worked example is the reason to buy this one, and it is included in full.
What it needs, and what it does not do.
Read this bit properly. It is cheaper for both of us than a refund.
Requirements
- Claude Code, with custom command support
- Python 3.12+ for the pipeline, with jinja2, pyyaml, markdown2 and weasyprint in your own virtual environment
- pandoc, plus WeasyPrint's native Pango and Cairo dependencies, via your OS package manager
- For images: a Canva MCP connection, or a free Cloudflare Workers AI key, or a billed Gemini key. Pick the one lane you will use
- playwright and Chromium only if you use the marketing-image renders
Limitations, stated plainly
- Tested on macOS and Linux. Windows via WSL is untested
- No image generator is bundled. You connect a lane, and the free Cloudflare one costs nothing
- The Gumroad skill expects Gumroad's own CLI or its dashboard. Neither is included here
- Built against Claude Code's skill format as of mid-2026
- Updates are at the author's discretion. There is no guaranteed release cadence
- It does not find you customers. It builds and packages the product
Questions worth asking first.
Is this just a folder of prompts?
Partly, and the rest matters. Seven of the files are Claude Code skills, which are structured instruction files rather than one-shot prompts. The pipeline and image engine are real Python you run, with their own dependencies and a test suite. The worked example is a finished product with rendered PDF and DOCX files in it.
Will it work with something other than Claude Code?
The skills are markdown instruction files, so they port to any agent that reads instruction files, but they are written and tested against Claude Code's command format. The pipeline and image engine are plain Python and do not care what agent calls them.
Do I need to pay for image generation?
No. Three lanes are supported and one of them is Cloudflare Workers AI on its free tier, which needs an account ID and API token but no payment. The Gemini lane needs billing enabled. The Canva lane authenticates through an MCP connection.
What exactly is the worked example?
A complete digital product built with these skills, including its content, render config, sales copy, Gumroad landing page, and the rendered PDF and DOCX the pipeline produced. It is included read-only so you can inspect real output rather than a description of it.
How is this priced?
One-time, no subscription, no upsell sequence. The honest anchor is your own time: if you would assemble an equivalent toolchain yourself, the question is whether that is a better use of a weekend than reading someone else's finished one.
Can I use it for products I sell?
Yes. That is the entire point. You build and sell your own products with it. The worked example is a demonstration, not something to resell as-is.
Seven skills, one pipeline, and the product they built.
Instant download. One-time purchase, USD. No subscription, no upsell page, no drip sequence.
Get the pack - US$39