BlueCLBusiness brief
The generation workflow sends the submitted brief to the configured model provider.
Web Vending Machine / Generative websites
Turn a business brief into an editable site.
Web Vending Machine connects business descriptions, structured website generation, visual editing and site storage in one completed application.
Development complete
Three steps
From the first step to a useful result.
- 01
Describe the business
Prepare the business content and the direction for the page.
- 02
Generate a draft
The backend maps structured model output into page fields.
- 03
Make it your own
Use the Puck editor to review, change and save the page.
PRODUCT / IN VIEW
See the product.
Explore the interface, then open the service, source code or demo using the links above. Select an image to see it in full.
Getting started
Start with the right setup.
Access & setup
- Read the repository’s frontend and backend setup instructions.
- Configure the model provider and site storage in your own environment.
- Generate a small draft, then refine it in the visual editor.
Permissions & data
Understand what the workflow uses.
Editing and storage
The application stores editable site data so generated drafts can be reviewed and changed.
This is product usage guidance. The BlueCL website privacy notice covers this company website; consult the service’s own documentation for its deployed data handling. Website privacy ↗
Behind the product
How the product works.
Starting a website combines several jobs: writing copy, choosing a visual direction, assembling sections and editing the result. Generated text becomes useful when it can be mapped into a page that the user can keep refining.
01Business briefStructured content requirements
02Generation backendCurrent OpenAI Responses API
03Editable templateStructured fields → Puck components
04Saved websiteVisual edits → stored site data
Model integration
The FastAPI generation path uses the OpenAI Responses API to request structured content. Validation maps the result into fields consumed by the React Puck editor.
Implementation details
The Python shopping service builds prompts, uses a response schema, caches template results and tries configured model fallbacks for eligible errors. The editor can rebuild its component data from the saved AI response or load existing Puck data. A future Claude provider would sit in the generation backend, before the same validation and editor mapping.
Further improvements
Introduce a Claude provider adapter and compare schema adherence and editable-content quality using the same business briefs.
When you need a hand
Frequently asked questions.
Can I change the generated draft?
Yes. The Puck editor is part of the implemented workflow.
What does the generation brief include?
Describe the business, desired content and visual direction. The workflow maps the generated result to editable page fields.
Is there a public hosted version?
The source repository is linked. A public service URL has not been provided.
Product and source information updated: October 8, 2026