分类与标注

Android 助手通过无障碍服务与本地 OCR 读取可见对话，在悬浮窗显示分析及回复建议。 Jev 判断意图与回复选项，生成模型起草三条候选，再由 Jev 排序。

AnyJev 从开放模型读取概率，回答选择、是非和评分问题，无需生成文章。 这是独立的 Jev 风格方案；L2 预测头不会修改基础模型权重。 可用无需标签的 L0 读出，或借助 transformers、vLLM 为特定问题拟合 L2 预测头。

Shapeshift 将文字输入变成日程、清单、计时器、分账等面向具体任务的界面。 可选 Jev 判断意图和信号；日期、算术由确定性代码处理，默认使用离线关键词分类。 在基于 Bun 的应用中输入、选择并保存卡片；配置服务端 TypeSafe 密钥即可启用 Jev。

JevChat-Windows 通过本地 OCR 读取微信和 KakaoTalk 对话，在聊天窗口旁提供候选回复。 Jev 判断意图、紧张度并为三条候选排序；独立语言模型负责起草。 在 Windows 10 1903+ 或 Windows 11 上，分别配置判断服务与起草服务的密钥。

悬浮聊天面板显示意图、风险判断和候选回复，用户可复制或填入聊天输入框。 使用 macOS 13 或更新系统的 Apple Silicon Mac。截图读取需要屏幕录制权限，文本接口读取或填入需要辅助功能权限；在设置中选择判断与生成服务。

classifier.dev 通过 HTTP 和 CLI 提供文本分类，支持批处理及多个标签维度。 可在 URL 中传入标签和文本，安装 classify CLI，或将 items 和 dimensions POST 到 /v1/classify。

jev-voice-browser 将语音转录流发送给 Node 服务，通过 Playwright 控制可见 Chromium 窗口。 Jev 选择意图和目标；代码提取原文候选，并决定执行、等待还是询问。 需要 Node 20+、TypeSafe 密钥，以及 Chrome 或 Edge 麦克风控制页。

暂无项目说明。

200+ plug-and-play Jev recipes: small, calibrated AI decisions that route, grade, gate, compare, and label text for agents, RAG, support, code review, and music. Import from JavaScript or TypeScript, or call the CLI with JSON from any language.

Any chat model, any engine (SGLang, vLLM, transformers) as a Jev-compatible probability decision service: one prefill, one label token

⚡ High-speed, zero-hallucination System-1 intelligent tagging assistant for Obsidian powered by TypeSafe Jev

Home Assistant conversation integration powered by the Jev / TypeSafe AI API for fast, structured intent routing and device control

An intent router for Neovim that classifies AI commands with Jev (TypeSafe's System One model) via the OpenRouter Decisions API, then routes execution to the matching handler.

Minimal native macOS intent router for browser and profile selection

A nutrition label for privacy policies. Plain-language answers, confidence scores, and source excerpts powered by Jev.

JEV intent classification using Python

Async, typed intent & tool routing for Python agents: Jev decides, a confidence gate stops weak decisions before they reach a privileged tool.

Finds out whether Claude Code actually did what you asked, not just what it cost.

Label and prioritise Gmail with a language model. Reads mail through himalaya, classifies each thread with Jev (TypeSafe) over OpenRouter, and applies Gmail labels and colours. Dry run by default, idempotent, no state file.

Apply natural-language codebooks to tables: a CLI and Python API with resumable annotation, exports, and label evaluation.

Proof of concept: auto-tagging and triage of Readwise Reader articles using TypeSafe AI's Jev

Route an agent's intent to the one installed skill that fits, using TypeSafe's Jev

TypeSafe Jev-powered intent gate for OpenCode: confirm intent before the agent dives into underspecified requests.

AI terminal assistant for Git commit messages, code reviews, shell commands, and chat. Uses Jev (TypeSafe AI) for intent routing and execution planning, powered by Pi.