Most Twitter/X → Nitter redirector extensions request <all_urls> (access to every website you visit) even though they only ever touch two domains. This one doesn’t. It requests host access only to the specific Nitter instances it uses, plus webNavigation/tabs/webRequest/storage — nothing else, and no permission for x.com/twitter.com at all (redirecting away from a page doesn’t require it).
It also validates the redirect instead of trusting it blindly: rate-limit detection, HTTP status checks, and empty-timeline detection (covers profile pages, search, hashtags), with automatic fallback to the next instance until one works. Most single-instance redirectors just break silently when their one hardcoded instance goes down.
- No telemetry, analytics, or backend of any kind
- Two files, no build step, no dependencies
- Remembers the last working instance so it doesn’t retry known-dead ones every time
Signed .xpi is on the releases page if you don’t want to build/sign it yourself. Feedback and instance-list corrections welcome; which public instances work changes often.