Documentation
Roadmap¶
Planned and proposed direction.
In progress / planned (not in order)¶
- Render/store p5.js processing outputs
- Watermark on portfolio images
- Security audit for exposed versions
- Code Test coverture
Shipped¶
- ZIP ingestion with content-hash deduplication
- WebP preview conversion; LZ4-compressed SVG storage
- Collections + built-in Favourites
- Markdown notes per item
- Bulk metadata download/upload edits.
- Bulk actions (tags, clear notes, favourite, download, collections)
- Keyboard navigation between items (← / →)
- Public portfolios at
/p/<slug>, withslide(printable deck) orscroll(continuous page) layout, four presentation themes, and a clickable Works index - Image lightbox with 1:1 zoom and pan
- Vision-LLM descriptions (OpenAI-compatible), with a configurable response timeout
- Collection quick-link ("View items" filter chip)
- Full-text search over all metadata, plus a JSON-LD metadata record per item (FAIR — Findable)
- Multi-library storage
- Database + full-asset backup/export
- Appearance settings (theme, accent, nav, density)
- Multi-user mode — Admin (full) and Reader (view-only, download-capable); unique password per user
- Auto-generated, persisted session signing key (no manual secret setup)
- Build provenance recorded (git SHA + build date)
- Redacted diagnostic report (Settings or CLI)
- On-brand 404 page
- Catalogue sorting & filtering controls
- Templates for portfolios — a rule-set skill for an LLM to generate one