Submit a find
한국어
日本語
Español
Français
Deutsch
简体中文
繁體中文
Português
Bahasa Indonesia
ไทย
Italiano
Tiếng Việt
Türkçe
Polski
Русский
Back to the labJEVLAB / FIELD NOTES
Fresh from the lab
A dated reading list of newly added resources and JEVLAB NEWS. Come back for what changed, or follow the RSS feed in your own reader.
Remember this visit on this browser to highlight new additions next time. No sign-in or cross-device tracking.
Dates show when an item became available in JEVLAB, in UTC, not when its original source was published. Translation-only edits do not count as new resources.
36
jev-spam-eval
This project explores zero-shot spam filtering using TypeSafe Jev Noul questions. It provides a comparative analysis against traditional TF-IDF baselines to evaluate classification performance in text filtering tasks.
github.compi-jev-skill-picker
This tool utilizes Jev evaluation capabilities to rank Pi agent skills for specific tasks. It demonstrates a method for optimizing agent performance by matching specialized capabilities to user requirements through systematic assessment.
github.comgot-jev
This proof-of-concept project explores Jev TypeSafe AI integration by applying it to a Game of Thrones theme. It demonstrates how structured data models can be utilized within a narrative-driven application context.
github.comomp-jev-compaction
This tool demonstrates a method for verbatim Jev-scored context reduction within omp environments. It provides an interface for managing data flow across TypeSafe or OpenRouter platforms to optimize context usage for AI agents.
github.compi-typesafe
This project provides a coding agent extension built upon the TypeSafe AI System One API. It demonstrates how to integrate Jev-based infrastructure into development workflows to facilitate automated coding tasks and agentic interactions.
github.comtypesafe-ai
This Rust library provides typed clients for AI services, supporting both asynchronous and blocking backends. It demonstrates how to implement observable retry mechanisms to improve the reliability of AI-driven application interactions.
github.comjev_local
This project explores the feasibility of running Jev-based workflows using local large language models. It provides a framework for users to experiment with private, offline implementations of Jev architectures on their own hardware.
github.comjev-model-router
This plugin integrates with Claude Code and other tools to select AI models based on Jev decision logic and OpenRouter pricing. It aims to balance intelligence, speed, and cost for specific tasks.
github.comtypesafe-dotnet-sdk
This community-developed .NET SDK provides an interface for the TypeSafe AI System One API. It enables developers to implement typed noul, choice, and score questions while retrieving structured, confidence-scored responses within their .NET applications.
github.comtsai-sc
This project demonstrates an approach to controlling the original StarCraft shareware using TypeSafe Jev. It utilizes keyboard and mouse input emulation based on recorded action probabilities to interact with the game environment.
github.comJev: Typed Decisions
This guide explores the Jev framework, which prioritizes structured, typed decision-making over traditional text-based processing. It demonstrates how developers can implement type-safe logic to improve reliability in automated workflows.
flaviocopes.comAutomated Video Ad Scoring
This tool utilizes Maxfusion and Jev to evaluate individual shots within video advertisements. The creator claims the system processes hundreds of ads in minutes, demonstrating a method for rapid, automated visual content analysis and performance scoring.
x.comjev-search
This Python CLI tool demonstrates an experimental approach to semantic line searching using TypeSafe Jev and OpenRouter. It provides a lightweight interface for querying data without requiring complex runtime dependencies.
github.comStarchild Prompt Routing
Starchild utilizes Jev for real-time prompt classification. The creator claims this implementation reduces operational costs by 20x and increases processing speed by 6x, with an average latency of 140ms.
x.combrowser-ai
This project demonstrates a Jev browser agent built in Go that interacts with a live Chrome instance. It provides a framework for automating web-based tasks by bridging the Go runtime with browser automation capabilities.
github.comdecider
This project introduces a System One-style model fine-tuned from Qwen3.5-2B. It demonstrates a method for performing single-pass typed decision-making tasks while incorporating calibrated probability outputs for improved reliability in automated reasoning workflows.
github.comtoolgate
toolgate provides a security layer for agent tool and MCP calls. It enables developers to implement allow, deny, or human-in-the-loop approval workflows using TypeSafe Jev to manage external tool access.
github.comsystem-one-open
This project provides an open-source replica of the Jev architecture. It demonstrates how to achieve typed, calibrated decision-making processes within a single forward pass using Gemma 4 E2B or Gemma 3 270M models.
github.comDanish stock-market backtest
This project demonstrates a sentiment-based backtesting approach for the Danish stock market in 2025. It illustrates how Jev can aggregate diverse data sources to perform financial analysis at a low cost.
x.comopenjev
This research project explores local bilingual probability decision-making using context, questions, and candidate answers. It serves as an independent experiment inspired by TypeSafe Jev concepts to evaluate how models process language-specific inputs locally.
github.comJev Subway Surfers Gameplay
This demonstration showcases Jev managing multiple instances of Subway Surfers simultaneously. The creator claims the system achieves high-speed gameplay across 50 concurrent sessions for a low cost, though these performance metrics remain unverified by independent testing.
x.comjev-tetris
This project demonstrates a visual implementation of Tetris where Jev is utilized to determine piece placements. It serves as an experimental showcase for integrating TypeSafe logic into classic game mechanics.
github.comis-malicious
This tool utilizes Jev to scan codebases for deceptive or data-stealing patterns. It aims to identify suspicious files and specific line ranges, allowing developers to review potential security risks before executing the code.
github.comJev Recommendation Engine Integration
This research demonstrates how Jev integrates into recommendation systems. The creator claims the model generates over 70 personalized suggestions per user, citing specific operational costs per user for the implementation.
x.comSelf-consistency: choices
This TypeSafe AI reference guide explains the self-consistency method for improving model output reliability. It demonstrates how to evaluate multiple reasoning paths to select the most consistent response when addressing complex logical tasks.
docs.typesafe.aiThe Jev Build Report
This report provides a comprehensive overview of public Jev builds from their inaugural week. It aggregates repository metadata, development time, and cost metrics to serve as a research resource for analyzing early platform activity and build patterns.
madewithjev.comTypeSafe SDK Changelog
This official resource provides a chronological record of updates and modifications to the TypeSafe SDK. It serves as a primary reference for developers to track version history and identify specific changes implemented in the platform.
docs.typesafe.aiJido Agents Tic-Tac-Toe
This demonstration showcases Jido agents utilizing Jev to execute moves within a game of Tic-Tac-Toe. It serves as a practical example of how Jev can be applied to coordinate decision-making processes in multi-agent game environments.
x.comParallel adversarial browser testing
This project demonstrates a massively parallel browser-based testing suite designed for adversarial evaluation of software releases. The creator claims this approach enables cost-effective automated testing by running multiple browser instances simultaneously to identify potential vulnerabilities.
x.comTrustGate
TrustGate provides an independent media trust and safety gateway utilizing TypeSafe Jev. This tool demonstrates a framework for managing content moderation and safety protocols within decentralized media environments through structured, type-safe validation processes.
github.comOpenvons
Openvons is a decision-making layer designed to process text, images, and Japanese voice commands. It demonstrates a method for generating probabilistic responses based on a predefined set of limited choices.
github.comJev on Netlify AI Gateway
This integration enables Jev to function within the Netlify AI Gateway environment. It demonstrates a streamlined configuration process for deploying Jev-based workflows directly through Netlify infrastructure.
netlify.comJev Architecture Overview
These technical notes detail the architecture of Jev, specifically focusing on its implementation of parallel processing and its ability to output probabilities without relying on traditional text generation methods.
x.comtypesafe-sdk-go
This repository provides an unofficial Go software development kit for interacting with TypeSafe AI services. It demonstrates how developers can integrate TypeSafe AI functionality into Go-based applications through a structured client interface.
github.comopenjev-sglang
This project provides a Jev-compatible public API leveraging open models. It demonstrates an approach to achieving fast parallel processing for Jev-based workflows using the sglang framework.
x.comotto
otto is an open-source agent for macOS and Windows that enables native computer interaction. It utilizes TypeSafe Jev, local OCR, and selective planning to perform tasks directly on the user desktop environment.
github.com
