Showcase
Introduction¶
A domain-agnostic catalogue server: ingest ZIP items into a self-hosted app, organise them into collections, and share public portfolios.
What it is¶
CatalogueCanvas is a self-hosted catalogue server. You ingest your digital art work as a ZIP file, organise them into collections, enrich them with metadata and optional AI-generated descriptions, and publish curated portfolios as shareable slide-deck pages. It runs from a single Docker container.
What it does¶
- Ingest — one ZIP becomes one catalogue item; the main image is auto-converted to a WebP preview.
- Organise — group items into collections and a built-in Favourites set.
- Exploration — simple quick exploration of the work uploaded, searchable.
- Enrich — edit titles, tags, and Markdown notes; optionally describe items with a vision LLM1.
- Publish — assemble portfolios and share them publicly at a
/p/<slug>link. - Store & back up — keep assets across multiple libraries; export the database and all files.
Key concepts¶
| Concept | In one line |
|---|---|
| Item | A single catalogued work, created from one uploaded ZIP. |
| Collection | A named grouping of items. |
| Portfolio | A curated, shareable slide-deck of items. |
| Library | A storage location for item assets. |
→ Full definitions in the Glossary.
Next steps¶
- New here? Read Uses & for whom.
- Want to run it? See Install.
- Using it day to day? See the User documentation.
Frequently asked questions¶
What is CatalogueCanvas?
A domain-agnostic, self-hosted catalogue server. You ingest items as ZIP files, organise them into collections, and publish curated portfolios to a shareable link or to print. It runs from a single Docker container (bare metal install are possible, check install).
How is it licensed?
CatalogueCanvas is open source under the GNU AGPL v3.0. You can use it, modify it, distribute it, and build on it for any project, with no fees.
Can I contribute?
Yes. On the project's GitHub open an issue for problems or to suggest a new feature, or code them yourself and open a pull request. The link is at the top of the page.
Is CatalogueCanvas production ready?
CatalogueCanvas is a stable beta. It may change as development continues. I avoid breaking changes, but if one lands you will get a warning, and I will test transition scripts so existing installations keep working. I would not run it on a server exposed to the internet yet.
How do I report a problem or request a feature?
Open an issue on GitHub. For a bug, attach the system log from the settings script. For a feature, describe the use case.
-
LLM — a large language model. I have tested local models in LM Studio, Ollama, and llama-cpp. ↩