Guardrails & safety — page 4

ZCode PreToolUse guardrail built on the shared jev-evaluator core, combining deterministic rules and TypeSafe Jev for allow, ask, and deny decisions.

AI-assisted Excel operations engine: structured planning, JEV decision support, deterministic policy and execution, verification, and an audit trail.

Reads the open reports on a Mastodon server with a moderator's own read-only token and prints them as a queue sorted by severity, using TypeSafe's Jev model. It never takes a moderation action.

AI chat moderation for Minecraft (Paper/Folia) and Hytale servers, powered by TypeSafe's Jev model. Works with the TypeSafe API or Cloudflare AI Gateway.

Autonomous anti-spam and moderation engine for WhatsApp groups powered by TypeSafe Jev System-1 judgment primitives and zero-latency two-tier defense.

Decisões tipadas em milissegundos com o Jev (System One da TypeSafe) via OpenRouter — state + perguntas noul/choice/score entram, decisões calibradas saem. Validação pesada baseada nos conceitos do modelo, bandas auto/hitl/abstain, deteção de prompt injection e servidor MCP. Zero dependências (Node ≥20).

Jev (TypeSafe) as the decision layer for Claude Code: permission gate, skill routing, conditional instructions, effort-tiered subagents, context packs. Tested PoC with evals.

Policy + evaluation framework for Jev / Kev (TypeSafe System One): typed probabilities in, auditable actions out — calibration, budget-compiled thresholds, permutation & drift checks. Zero-dependency core, kev-native.

Pre-execution risk gate for AI coding agents (Claude Code, Codex, opencode, pi, Hermes). Blocks or asks on risky shell commands, guards against prompt injection in what agents read, escalates to System 2 and a human queue. Optional TypeSafe Jev.

Insurance claim triage console using TypeSafe AI's Jev model, with rule-based fallback, policy rules, tests and CI/CD

O Jev (TypeSafe AI) como guardrail em tempo real e como pré-avaliador de agentes conversacionais de negócio. Next.js + LangChain via Vercel AI Gateway, com estudo de calibração em português.

Ticket router built on TypeSafe's Jev, the companion code for a DataCamp tutorial. Six typed questions in one call, routing policy in plain Python.

Linux-only OpenCode shell permission guard using Jev; auto-approves simple reads and limited /tmp writes.

Ultra-fast, sub-100ms API & webhook guardrail and triage gateway powered by TypeSafe AI System One (Jev). Parallel 7-dimension speculative evaluation, deterministic policy router, zero-dependency SQLite audit trail, and automated outbound dispatch.

Describe a bagel and TypeSafe's Jev model rules on it under the Uptech Studio Bagel Policy. Made by an Uptech employee, not an official Uptech product.

Fail-closed PreToolUse safety hook for Antigravity shell commands using TypeSafe Jev.

Evidence-aware semiconductor alert triage research demo with TypeSafe Jev, policy guards, and reproducible evaluation

JEV evaluation (FM-JEV-01): replay-only, advisory-only evaluation of Foreman-style supervision with TypeSafe Jev. Measures how much safety comes from the model versus a deterministic evidence gate. No worker authority.

A Jev-powered policy gateway that checks an AI agent's proposed actions before executing them.

This repository contains a GitHub issue classifier built on Jev, TypeSafe AI's System One model. It labels every new issue with typed values and calibrated confidence in milliseconds, labelling what it is sure about and escalating what it is not. Three guardrail layers guard every write, and a frozen eval suite gates each deploy.

Jev (TypeSafe) vs every Claude model on prompt-injection validation and skill/agent routing: 1,000 labelled cases per task, measured latency and cost, films

Can a System One model (TypeSafe Jev, open-source Kev) replace an LLM-as-judge in agent guardrail callbacks?

Three.js Autonomous Driving Reflex & AI Safety Brake Simulator powered by TypeSafe Jev System 1/2 Dual-Brain Architecture

PoC: an AI Gateway guardrail policy for WSO2 API Platform, backed by TypeSafe AI's Jev model