Different formats. One library.
Feed adapters and explicit browser captures normalize into a common library. Articles, papers, podcasts, and videos retain their own metadata and reading cues.
Art-directed interface study · sample content
Calista brings reading, research, release feeds, and browser saves into a personal library. The content stays on your machine, and each source keeps the details that make it useful.
Useful material gets scattered across bookmarks, subscriptions, paper tabs, video queues, and release pages. Those formats need different treatment, but switching tools to organize them makes the library harder to use.
I built a connector-driven library with a shared item model. Source, date, format, reading state, and provenance travel with the content. A quiet feed, focused reader, boards, and local search support the daily work of deciding what to keep.
Feed adapters and explicit browser captures normalize into a common library. Articles, papers, podcasts, and videos retain their own metadata and reading cues.
The reader gives saved content a focused surface. Reading and saving are distinct states, so collecting something does not pretend you have finished it.
Sources expose their status and fetching history. Local search, boards, and portable SQLite storage make the library something you can organize and keep.
The framework-independent layout engine predicts card dimensions from item data. Placement is append-only: loading the next page preserves the positions already on screen.
Typed arrays store layout geometry, and per-column windowing limits mounted content. The browser does not need to measure every item to place it.
Fetching has size and rate bounds, DNS address pinning, and an SSRF gate. Source HTML passes through an allowlist sanitizer before the reading interface sees it.
Feeds, captures and sources
Identity, dates and geometry
SQLite library + FTS5
Read, save, search and organize
Interface walkthrough · sample data · no audio
A local library with a stable feed, focused reading, explicit captures, and searchable source context.
The current implementation focuses on local content management. Semantic search and broader cloud services are outside this version; the final license decision is pending.
Search & retrieval