Download Latest Version v2.15.0 source code.zip (628.4 kB) Google Add to Preferred Sources
Home / v2.12.0
Name Modified Size InfoDownloads / Week
Parent folder
single-file-aarch64-apple-darwin 2026-09-13 69.0 MB
single-file-aarch64-linux 2026-09-13 106.7 MB
single-file-x86_64-apple-darwin 2026-09-13 82.0 MB
single-file-x86_64-linux 2026-09-13 106.2 MB
single-file.exe 2026-09-13 81.6 MB
README.md 2026-09-13 1.5 kB
v2.12.0 source code.tar.gz 2026-09-13 562.1 kB
v2.12.0 source code.zip 2026-09-13 618.8 kB
Totals: 8 Items   446.7 MB 0

SingleFile CLI 2.12.0

CLI fixes and improvements

  • --insert-canonical-link turns the canonical link off. The element is still inserted by default, so nothing changes unless you ask for it. The option existed in the core library but was forced on after the options were merged, so passing it did nothing at all
  • --read-maff-metadata saves the original URL and the archive time of a page that was extracted from a MAFF archive, read from the index.rdf file stored next to it. Without it the page is saved under the path of the extracted copy, dated today. It is off by default because it asks for an index.rdf beside every page saved, which is a request that fails on the open web

Fixes from single-file-core 1.5.130

  • Saved pages can no longer submit forms or set a base URI. Both were available to a saved page and neither has any use once the page is on disk
  • javascript: URIs are removed from every attribute, in every namespace. The previous pass read the resolved property, so it missed the href of an SVG link and never saw xlink:href
  • Reading the metadata of a page extracted from a MAFF archive no longer fails the capture when index.rdf declares an original URL with no RDF:resource attribute
  • The filename template data of a page extracted from a MAFF archive records the URL recovered from the archive, not the path of the extracted copy

Changes

  • zip.js is updated to 2.15.0

Co-authored by Claude (Claude Code)

Source: README.md, updated 2026-09-13