Lanci del 26 settembre 2026

La CLI jg aiuta gli agenti a esplorare repository sconosciuti con file pertinenti, riferimenti di riga ed estratti del sorgente. Jev valuta cartelle, anteprime dei file e dichiarazioni mentre la ricerca segue i rami pertinenti.

Unofficial skill that teaches coding agents to build with TypeSafe AI's Jev: typed decisions, calibrated confidence, and prior art from 150+ community projects.

Classify text in PHP without an LLM bill: typed decisions in 100+ languages, self-hosted. Laravel-ready SDK for Laya, a Jev AI alternative.

Watch X for the news that matters to you, judged by Jev, and get alerted in Slack. Set up by an AI agent.

Unofficial Go and Rust clients for the TypeSafe AI System One API (Jev)

/fc for Claude Code: shrink old tool output in about a second — Jev keeps what's still needed, every cut saved to a file. Your /compact stays untouched.

An unofficial, provider-neutral Unix client for Jev from TypeSafe AI. Written in Go.

outcome-aware & adaptive model routing for coding agents with deterministic cache powered by Jev

A fuzzy linter for coding agents. It checks the code your agent writes against your team's best practices while the agent is still working, not at code review.

OpenCode plugin that adds an Auto (Jev) virtual model which routes each prompt to a configured real model using TypeSafe AI (Jev).

Local knowledge search with three modes: BM25 + Jev by default, agentic multi-query lexical search + Jev without vectors, and optional BM25 + embeddings/RRF + Jev, with grounded streaming answers.

A reproducible, self-hosted benchmark for comparing JEV and LLM decision-making on your own datasets, models, and evaluation policies.

zero-shot classifier models gateway and runner

Ctrl+F that finds by meaning - highlights the sentence you meant, with receipts. Extension + local helper + playground (early scaffold).

Trolley Promblem for the Agentic Era

MCP server and skill that let an agent drive an Android phone naturally over adb

Know when to hand a Claude Code session to a fresh one. A Stop hook that times the handoff with Jev or an open-weights model, sending only derived signals, never your chat.

VS Code Codex Plugin for Jev-gated tool output integration

Native DuckDB extension composing fast System One judgments with selective System Two reasoning, provenance, batching, and caching.

Built with Jev: how Fake / Real (fake-real.live) uses TypeSafe Jev to check claims against cited evidence.

Local playground and test bench for TypeSafe's Jev System One model: typed questions, calibrated-probability answers, and side-by-side comparison with an LLM baseline.

Jev 浏览器记忆助手：本地 BM25 + multilingual E5 召回，TypeSafe Jev 语义重排。

Fast, cheap guardrails for LLM apps, powered by TypeSafe's Jev model

JEV intent classification using Python

Jev System one AI model, very fast decision making engine. Jev AI will use input tokens but no output tokens

Tool search for LLM agents: BM25 vs embeddings vs rerankers vs Jev on 525 real MCP tools, plus an experimental Jev search engine

One-shot model and reasoning routing for Hermes Desktop. Manual control wins. Explicit upstream integration preview.

Zig SDK for TypeSafe AI's Jev decision API. Typed classification, scoring and yes/no probabilities with bounded buffers, retries and deadlines.

A hands-on educational lab for DeepEval + Jev: inspect real recorded judgments, explore release gates and drift playback, and run fresh evaluations on your own workloads.

Temporal evidence memory. Reconstruct historical context and trace decisions affected by corrections.

JEV and local LAYA for MCP. Typed decisions, bounded inference and durable decision records.

A real open AI agent + jev web interface that searches the web and creates research reports

Operating System powering the burgh's build chain. Made for AI Horizons 2026 Hackathon

Community TypeSafe Jev semantic decision integration for airbyte

Universal AI Agent Control Plane: LLM agents (Claude Code, Codex, Gemini, OpenCode) do the work, Jev judges typed decisions, executed checks provide the evidence.

Anchor: ground truth for Pi agents, backed by MongoDB Atlas verified state

Community TypeSafe Jev semantic decision integration for argo

An agent skill for structured decisions with Jev: choose options, score candidates, check claims, and rank files.

Use Jev instead of an LLM for RAG decisions: relevance grading, query routing, hallucination checks and LLM-as-a-judge evals. Faster, cheaper, measured first.

Decides route by route whether a full-page cache may serve one anonymous visitor's copy of a page to everyone, from response headers, WordPress and WooCommerce markers and TypeSafe's Jev model.

A jev Choice question that grows its own options: an LLM writes a new one when none fits.

Turns free-form git commits into a Keep a Changelog section: each commit's own first line, placed by its conventional-commit prefix or by TypeSafe's Jev model, with unsure commits listed for review.

Dagster asset checks with TypeSafe Jev semantic decisions

Deepsift. Not every bit deserves the trip to Earth.

Picks the reasoning effort for each turn of Claude Code with a System One classifier

Semantic emoji search - keyword + embedding recall, reranked by TypeSafe Jev. Built with TanStack Start, TanStack Query, and Emoji Mart.

Version-aware evidence decisions for RAG assistants using Jev, with inspectable routing, bounded recovery, and reproducible evaluation.

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

On-call runbook executor agent: triages alert emails, plans remediation with Jev risk scoring, sandbox-tests on TrueForge, auto-fixes safe steps and pauses for human approval on the rest.

Giving people the power to build community and bring the world closer together!