Jev发布帖
Diogo Almeida介绍了Jev,即TypeSafe的结构化决策模型。观看发布影片,并探索原始帖子了解该模型的方法、示例和作者报告的性能。
Diogo Almeida介绍了Jev,即TypeSafe的结构化决策模型。观看发布影片,并探索原始帖子了解该模型的方法、示例和作者报告的性能。
TypeSafe的正式发布公告介绍了实验室,并引导开发者使用Jev。公告包含Diogo Almeida的发布影片和原始介绍。
TypeSafe宣布Jev向公众开放,无需等待列表。官方控制台是将类型化决策应用于自己应用程序的起点。
TypeSafe强调Jev的类型化输出方法:应用程序提供状态和问题,然后根据概率、选择或评分进行操作。相关公告还介绍了Venice API集成。
Start with the Launch Post. Explore four official signals.
1,198 项资源
AI-assisted summaries and translations. Check original sources for context and performance claims.
Safety checks for Hermes and other AI agents before they act, ask for approval, or verify a change.
Source-linked curation
An agent project that enforces policies by judging Claude Code tool calls.
JEVLAB ARTSource-linked curation
A GitHub project where Jev plays Minesweeper, revealing its capability to handle real-time, turn-based game mechanics.
Source-linked curation
Typed semantic decisions for Unix pipelines and CI, powered by TypeSafe AI Jev.
Source-linked curation
Cut Claude Code's skill manifest by ~75% with TypeSafe Jev. Scores every installed skill for relevance and hides the rest via skillOverrides — 12,750 → 3,185 tokens on a 217-skill install, for $0.0009 a session.
Source-linked curation
X timeline labeler. A community post describing a Chrome extension that categorizes X posts into types like engagement bait or promo.
Source-linked curation
A developer project integrating Jev for judgment tasks within DeepSeek Harness.
Source-linked curation
Evidence-driven frontend QA built on Jev Ultrafast and Browser Harness, with a synthetic todo demo.
Source-linked curation
Semantic SQL for Postgres, powered by Jev.
Source-linked curation
An unofficial Laravel package offering typed responses and test fakes for Jev integration.
JEVLAB ARTSource-linked curation
An evaluation of typesafe AI chess. As it turns out, the AI isn't doing really well even though chess is not a particularly open-ended game. Still, it's only a prototype and this probably wasn't optimzied for games.
Source-linked curation
Jev assists a Mac app by selecting relevant support articles from built-in documentation.
Source-linked curation
Local Instagram, TikTok, and LinkedIn research where Jev chooses bounded socai browser operations and streams captured evidence into cited reports.
Source-linked curation
Jev is integrated with Cloudflare AI Gateway to provide structured answers to typed questions.
Source-linked curation
One AI trade decision every Monad block. Jev on Kuru MON-USDC.
Source-linked curation
A fire evacuation simulation where each person makes Jev-style decisions.
JEVLAB ARTSource-linked curation
An optimized inference engine to turn LLMs into Jev-like machines: optimized for quick, lightweight, and accurate decision-making, classification, and scoring.
Source-linked curation
TypeSafe Jev playground — custom Choice/Score/Noul builder with live distributions.
Source-linked curation
A developer project involving three Claude Code hooks that query Jev for specific information.
Source-linked curation
Guía del creador Fazt sobre el uso de Jev para decisiones tipadas, con una introducción y ejemplos de aplicaciones.
Source-linked curation
Six hook gates for Claude Code that catch broken rules, out-of-scope edits, unfinished asks, and false claims, each judged by Jev against a fixed checklist and one piece of evidence.
Source-linked curation
CLI that uses Jev to catch AI slop in UI, copy, and agent instructions before ship (probabilities on semantic taste checks).
Source-linked curation
A 26-sheet construction plan set was classified in 2.9 seconds using Jev with an LLM pipeline, costing $0.0052 and matching GPT-4.1 and GPT-6 Astra performance.
Source-linked curation
Pi extension: verbatim context compaction with TypeSafe Jev decisions.
Source-linked curation
Semantic MCP firewall powered by Jev — screens every tool call, tool result, and tool description with calibrated System One verification. 94% block recall, 0 false positives, ~$0.00002/check.
Source-linked curation
Papers, open reproductions and independent evaluations behind System One models and Jev.
Source-linked curation
OpenTelemetry Collector connector that uses Jev to assess metric metadata and apply retention policies before export.
Source-linked curation
One-pass typed decisions with calibrated probabilities (System One style model), fine-tuned from Qwen3.5-2B.
Source-linked curation
Jev classifies emails for fraud detection, with uncertain cases handled by Kimi K3 for high accuracy.
Source-linked curation
🚀 Blazing-fast native macOS menu bar companion for Notion. Pure Swift & SwiftUI (<3MB), 8 languages.
Source-linked curation
Screens 360 resumes in 24.2 seconds with low cost and automated scoring.
Source-linked curation
A local computer-use fast path for Codex and Waku, designed for developers.
Source-linked curation
Eight minimal working examples of TypeSafe's Jev (a System One model) applied to mechanical and electrical engineering: CAD/CAE/CAM routing, FEM result triage, DFM screening, BOM alignment, hallucination-proof extraction. Zero dependencies.
Source-linked curation
Hosted typesafe-ai/jev for AI SDK evaluate calls, no TypeSafe waitlist required.
Source-linked curation
A Chrome extension that organizes bookmarks using user-defined rules for categorization.
JEVLAB ARTSource-linked curation
Official TypeSafe reference: Advanced: structure.
Source-linked curation
Classifies tax documents using an LLM pipeline for efficient processing.
Source-linked curation
Validated enhancement pack for MiniMax Code CLI on arm64 DGX Spark + local GLM-5.3-EXL3: TUI scrollbar + context-meter patches, compaction repair for local vLLM, Jev Ultrafast + Playwright MCP browser integration, real usage ledger.
Source-linked curation
A community project embedding three Jev-powered bots into a first-person shooter game.
Source-linked curation
Fast structured Android control loops with TypeSafe Jev and Mobile MCP.
JEVLAB ARTSource-linked curation
A developer project that selects applicable rules for Claude, ensuring only relevant information is shown.
JEVLAB ARTSource-linked curation
A developer project using Jev to score sales calls, collected for the creative shelf.
Source-linked curation
A 30-minute internal session on Jev attended by over 50 engineers, generating more than 50 product ideas.
JEVLAB ARTSource-linked curation
Small dependency-free CLI for TypeSafe Jev.
Source-linked curation
A developer project providing a second MCP server for Jev from the PyModel initiative.
Source-linked curation
A small, extensible decision-to-action harness for TypeSafe Jev.
JEVLAB ARTSource-linked curation
Decision harness for TypeSafe Jev — confidence gates, shadow mode, recipes, and evals. Claude CLI 48.9s → Jev 1.3s on the same row-filter job.
Source-linked curation
Official TypeSafe reference: Confidence.
Source-linked curation
A community tool allowing Jev to make decisions through command-line interactions for agent systems.
JEVLAB ARTSource-linked curation
Official TypeSafe reference: Agent skill.
Source-linked curation
A community post demonstrating Jev's ability to score 700 leads and predict outreach message performance in 40 seconds.
Source-linked curation
A community post exploring Jev's role in improv judging, presented as a creative experiment.
Source-linked curation
A community post suggesting Jev can be hired as a QA tester through @typesafeai.
Source-linked curation
A community post describing control of a simulated fly.
Source-linked curation
I tortured Jev into being a RISC-V CPU.
Source-linked curation
A YouTube guide explaining Jev's functionality, including browser use, classification, and code review without an LLM.

Source-linked curation
A community post describing an SEO audit using Jev to analyze website internal links in 45 seconds.
Source-linked curation
Racing UI wired to Jev driving decisions.
Source-linked curation
A tool for tracking decisions made by Jev, using Arize's instrumentation for detailed analysis.
Source-linked curation
Jev tags X posts in real time with labels like 'breaking,' 'golden nugget,' or 'slop'.
Source-linked curation
.NET SDK for the TypeSafe AI platform.
Source-linked curation
Autonomous Jev pull-request review with typed decisions, calibrated approval gates, and trusted-owner escalation.
Source-linked curation
Browser Use Olympics by Almond: one prompt, five events, one clock. Plus fast loop, a ~200-line browser computer-use agent (Chrome DevTools + TypeSafe Jev).
Source-linked curation
An agent using headless Chromium to quickly navigate the web, demonstrated by browsing Wikipedia links efficiently.
Source-linked curation
Evaluating TypeSafe's Jev as a fast monitor and action gate for agent sabotage in SHADE-Arena, compared with Gemini 2.5 Flash/Pro.
Source-linked curation
Local browser automation UI that uses TypeSafe Jev to choose bounded page actions and a text model only for field values.
Source-linked curation
Benchmarks and a playground for TypeSafe's Jev (System One) model: chess, and who-is-the-player-talking-to for speech-to-text game NPCs.
Source-linked curation
Jev adapter to Mellea.
Source-linked curation
A collection of research papers and open reproductions supporting System One models.
JEVLAB ARTSource-linked curation
A reproduction of Jev that turns any Qwen model into a fast decision model, serving the same /v1/systemone schema (Choice, Score, Noul) with no training and no generated answer text.
Source-linked curation
A macOS app that automatically organizes downloaded files using rules.
Source-linked curation
A show-and-tell capability study for Jev, TypeSafe's System One decision model.
Source-linked curation