Overview & result
qs-lll shows twitter-jev-guard, built on Jev: Low-quality, spam and ad posts get a translucent watermark.
From the original post
twitter-jev-guard by @qslll: Low-quality, spam and ad posts get a translucent watermark.
How Jev fits in the loop
- The extension scans posts as you scroll the X timeline, with a cap on concurrent requests.
- Its background worker sends post text and public metadata straight to TypeSafe Jev.
- Posts above the low-quality probability threshold (75% by default) get a red STOP watermark with the probability.
- Posts that X marks as ads get an amber AD watermark; watermarks sit on the text area, not on images or video.
How to reproduce
- Review the source documentation and repository at https://github.com/qs-lll/twitter-jev-guard.
- Configure environment with TypeSafe Jev / System One API keys.
- Run test prompts against the decision endpoints.
Why this build matters
Shows Jev in practical use for content, marketing and SEO.
Limitations
- The author says it is a personal browsing aid, not a fact-checking tool or a moderation system.
- It analyses text and public metadata only; it does not understand images or video.
- Results vary with language, context and X's page structure, so the threshold needs adjusting to your taste.