Effective: July 1, 2026

Harbor Web Clipper is a browser extension that saves web pages — as
simplified articles, full-page snapshots, bookmarks, or screenshots — into
your own Harbor account. This policy describes
everything the extension collects, stores, and transmits.

• The extension talks to exactly one service: your Harbor account. It
  bundles no third-party trackers or analytics SDKs, loads no third-party
  services, and shows no ads.
• Page content is read and transmitted only when you explicitly clip a
  page, and it is sent only to your Harbor account.
• Harbor does measure how the Web Clipper is used, on Harbor's side, to run
  and improve the product — see Usage & analytics below.
• The extension never sees or stores your password. You sign in on
  Harbor's own website; the extension holds only revocable OAuth tokens.

All local data lives in your browser's extension storage:

• OAuth tokens (access + refresh) for your Harbor account. Extension
  storage is not additionally encrypted — the threat model is the same as
  any OAuth-based app on your device. Tokens are revocable at any time
  (log out in the extension, or revoke the session from Harbor's settings).
• Your preferences (default notebook/format, shortcut bindings, image
  handling, post-save behavior) and small conveniences (last-used format and
  notebook).
• Cached lists of your notebook and tag names (refreshed every few
  minutes) so the pickers are instant.
• A diagnostics log of recent clip attempts (timestamps and human-readable
  status sentences only — never tokens, never page content). You can view,
  copy, and clear it in Settings → Logs.

Only when you take an action:

• Clipping sends the captured content (the article/page HTML, the
  bookmark link, or the screenshot image), your chosen title, notebook,
  tags, and comment to your Harbor account's API — nowhere else.
• Signing in opens Harbor's own login page. Your credentials go to
  Harbor directly; the extension receives only OAuth tokens.
• Opening pickers fetches your notebook/tag lists from your account.

The extension performs no background transmission of browsing activity.
It does not read pages you don't clip.


Like the Harbor apps, Harbor measures how the Web Clipper is used so we can
operate, secure, and improve it. This happens on Harbor's side, from the
requests the extension already makes to your account (signing in, saving a
clip, loading your notebooks) together with basic operational details such as
which browser and extension version a request came from. The extension itself
embeds no third-party analytics or advertising tools and does not
track your browsing. What Harbor collects and how it is used is governed by
Harbor's Privacy Policy.


Clipping requires reading the page you're looking at, and a clipper must
work on every site — that is why the extension requests access to all URLs.
The permission is used exclusively at the moment you clip (or preview a
clip) on the page you're clipping. No page data is read, stored, or
transmitted otherwise.


Everything the extension stores locally is removed when you log out (tokens)
or uninstall the extension (all of it). Clips you saved live in your Harbor
account under Harbor's own
terms and privacy policy, where you can delete them
at any time.


Changes to this policy are published in this repository
(docs/privacy.md) with the extension release notes. Questions:
support@cloudmanic.com or open an issue at
github.com/cloudmanic/harbor-webclipper.