AI Slop Alerter troch meow-d
Fast and lightweight AI slop detection that runs while you browse the internet.
Metadata útwreiding
Skermôfbyldingen
Oer dizze útwreiding
Browser extension that detects and flags AI slop, using a very bare-bones markdown conversion and a custom made fasttext model
how to use:
- on supported sites: it automatically runs "Analyze all text on page"
- or manually: right click > "Analyze all text on page"
- manually on specific sections: highlight sections you want > right click > "Analyze for AI slop"
supported sites:
- reddit
- twitter
- bsky
- linkedin (broken?)
- github (tons of false positives for now)
- generic article (fallback for everyting that isn't supported, if you decide to manually run Analyze all)
primarily motivated by non stop exposure to claude slop on ai reddit, normie twitter and 99% of linkedin. every single sentence just sounds like "it's not just a slop layer structure, it's a slop alerter — the core tension is that it's genuinely interesting: capture on-mobile signal -> brake + dragon layer" or something.
disclaimer: in my opinion, "AI content detection" itself is a theoretically impossible task. as such, the intent of the model is not accurate AI detection, but to imitate your human ai slop senses. so trust your own judgement over the model.
source code
how to use:
- on supported sites: it automatically runs "Analyze all text on page"
- or manually: right click > "Analyze all text on page"
- manually on specific sections: highlight sections you want > right click > "Analyze for AI slop"
supported sites:
- bsky
- linkedin (broken?)
- github (tons of false positives for now)
- generic article (fallback for everyting that isn't supported, if you decide to manually run Analyze all)
primarily motivated by non stop exposure to claude slop on ai reddit, normie twitter and 99% of linkedin. every single sentence just sounds like "it's not just a slop layer structure, it's a slop alerter — the core tension is that it's genuinely interesting: capture on-mobile signal -> brake + dragon layer" or something.
disclaimer: in my opinion, "AI content detection" itself is a theoretically impossible task. as such, the intent of the model is not accurate AI detection, but to imitate your human ai slop senses. so trust your own judgement over the model.
source code
Wurdearre: 0 troch 0 beoardielers
Tastimmingen en gegevens
Fereaske machtigingen:
- Jo gegevens foar websites yn it domein reddit.com benaderje
- Jo gegevens foar websites yn it domein linkedin.com benaderje
- Jo gegevens foar websites yn it domein youtube.com benaderje
- Jo gegevens foar websites yn it domein officeapps.live.com benaderje
- Jo gegevens foar twitter.com benaderje
- Jo gegevens foar x.com benaderje
- Jo gegevens foar bsky.app benaderje
- Jo gegevens foar witchsky.app benaderje
- Jo gegevens foar blacksky.community benaderje
- Jo gegevens foar github.com benaderje
- Jo gegevens foar webshell.suite.office.com benaderje
Opsjonele machtigingen:
- Jo gegevens foar websites yn it domein officeapps.live.com benaderje
- Jo gegevens foar webshell.suite.office.com benaderje
Gegevenssamling:
- De ûntwikkeler seit dat dizze útwreiding gjin gegevenssamling fereasket.
Mear ynformaasje
- Add-on-keppelingen
- Ferzje
- 0.1.2
- Grutte
- 32,85 MB
- Lêst bywurke
- 22 dagen lyn (12 jul. 2026)
- Lisinsje
- MIT-lisinsje
- Ferzjeskiednis
- Tafoegje oan kolleksje