Live radar · scored by Jev

What people are building with Jev.

119 signals · last crawl 2h ago

Filter

Jev: A Non-Generative AI Monitor

Jev is a non-generative model designed as an alternative monitor for AI control, aiming to answer yes/no questions with a probability.

DETECT

@exploding_grad1h

Jev Telegram Channel Analyzer

This project uses Jev, a decision model, to analyze Telegram channel posts from the past year, categorizing them by type, ad, clickbait, or emotional pressure.

SCORE

@gaborishko3h

Flue-Guard Jev Guard

This project adds a Jev guard to Flue-Guard to check tool calls after authorization, categorizing them into low, medium, or high risk to determine subsequent actions.

GATE

@MenskMensk1h

Jev Preflight

Jev Preflight is an open-source beta tool that flags risks and sends Claude back once, operating on 8 risk axes with a maximum of one re-check per request.

GATE

@muse_jp_sol2h

AI Agent Safety Layer

This project developed a safety layer for AI agents built on top of Jev, a judgment model from TypeSafe, and tested its efficacy against adversarial attacks.

GATE

@AnzalAbidi3h

Jev Profanity Filter

Jev is an independent index that classifies messages as profane or not, and then flags individual words for replacement.

DETECT

@lukas_caha1h

Jev AI Moderator for Livestream Chat

Jev is integrated into VideoRC as an AI moderator to classify livestream chat messages, determine their safety and relevance for display, and enforce streamer-defined rules.

CONTROL

@orcdev4h

This project builds a security review tool using TypeSafe's Jev, CodeGraph, and OWASP/CWE intelligence, aiming for low-cost security scanning through deterministic orchestration and probabilistic judgments.

VERIFY

@RichgoJames3h

End of radar

Jev: A Non-Generative AI Monitor

Jev Telegram Channel Analyzer

Flue-Guard Jev Guard

Jev Preflight

AI Agent Safety Layer

Jev Profanity Filter

Jev AI Moderator for Livestream Chat

TypeSafe Security Review