Skip to main content
ContentsLab
All featuresSource library

Capture once, cite forever.

URLs, notes, quotes, files — all live in a shared Project library. Search by keyword or by meaning. Cite in one tap. Never lose a link again.

The problem

Why this matters

Research material lives everywhere — bookmarks, screenshots, Notion, paper notebooks, downloaded PDFs. By the time you remember you read something relevant, the link is gone. The Source Library fixes this by giving every Project a single, searchable home for everything you have ever read.

How it works

Four steps, no friction.

  1. 1

    Add a source

    Paste a URL, upload a PDF, or write a note. The Source Library pulls metadata — title, authors, year, abstract — automatically when it can.

  2. 2

    Extract the facts and quotes

    Mark up the parts that matter. Save quotes, facts, and figures as structured records attached to the source.

  3. 3

    Search by keyword or meaning

    Keyword search is instant. Semantic search uses embeddings to find sources that mean what you are asking, even if the words are different.

  4. 4

    Cite in one tap

    Drag a source — or just a quote from it — into your draft. The citation is generated in the style you choose.

Key features

What you get out of the box.

  • Auto-extracted metadata for URLs (Open Graph, oEmbed, JSON-LD)
  • PDF, EPUB, and plain-text upload with full-text indexing
  • Per-source facts and quotes as structured records
  • Keyword search plus semantic (vector) search
  • Per-source tags, verified flag, and DOI lookup
  • Cite in MLA, APA, Chicago, or custom format
contentslab.app/features/source-library
Capture once, cite forever.
Source library
1Add a source
2Extract the facts and quotes
3Search by keyword or meaning
4Cite in one tap

Screenshot of the Contents Lab Source Library showing a list of sources with their facts and quotes.

Deep dive 1

Semantic search across the Project.

The problem. Keyword search forgets what you meant when you did not use the exact word.

The Contents Lab way. Every source is embedded into a 1,536-dim vector at write time. Ask "studies on slow reading" and you get sources about deep reading even if they never use the word "slow".

  • Embeddings computed on the Lightbase edge — no per-request cost
  • Hybrid mode combines keyword and semantic results, ranked by relevance
  • Filters by tag, type, date, and verified status
Deep dive 2

Facts and quotes are first-class.

The problem. A "fact" buried inside a 4,000-word PDF is just as lost as the PDF itself.

The Contents Lab way. Lift facts and quotes out of a source into their own records. Search, tag, and cite them independently of the source they came from.

  • Each fact or quote links back to its parent source
  • Page numbers preserved for print sources
  • A fact can be cited directly in your draft without re-opening the source

Try Capture once, cite forever today.

Free plan, no card required. Cancel anytime.