# Use web links as sources for a nonfiction ebook

Add saved articles and useful URLs to a Scribora project without turning the finished book into a generic web summary.

Canonical: https://www.scribora.ai/docs/sources/web-links
Category: Sources
Last reviewed: 2026-09-29

## Links should fill a role in the book

A saved URL is most useful when you know why it belongs in the project: evidence for a claim, a current example, a counterargument, or background the reader needs. A hundred vaguely related bookmarks create more noise than a handful of purposeful sources.

Combine external links with your own notes or transcripts. That gives the project both factual context and an authorial spine.

## Review time-sensitive sources

Web pages change and facts become stale. Before export, revisit claims that depend on current prices, policies, product capabilities, dates, or market data. Source-first generation reduces invention; it does not make old evidence current.

## What the URL importer can fetch

The URL path accepts HTTP and HTTPS resources that resolve to public network addresses. Private, loopback, link-local, and reserved destinations are blocked before fetching.

For normal HTML pages, Scribora extracts the main readable text and falls back to a second HTML parser if needed. Direct PDF and PPTX URLs are also supported. The URL fetcher caps a response at 10 MB and follows only a bounded number of redirects.

> A page still needs to be publicly fetchable and contain extractable content; a URL by itself does not guarantee a usable source.