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
MOSS litter-picking simulation
This simulation demonstrates MOSS following instructions to collect litter. It serves as an experimental model for how automated decision-making might be applied to urban sanitation tasks, though it is not a verified real-world deployment.
x.comjevchat
jevchat is a research-oriented chatbot project designed to process and respond to user-typed inquiries. This repository serves as a functional demonstration of conversational interaction patterns within a broader collection of experimental language processing tools.
github.comadvocaat
advocaat is a compact, type-safe client designed for querying AI models using your own data. It demonstrates an implementation of TypeSafe Jev to facilitate structured interactions between local datasets and AI processing.
github.comContextual clipboard quick actions
This macOS utility demonstrates a workflow for triggering automated actions based on the specific content currently held in the user's clipboard. It aims to streamline productivity by providing context-aware shortcuts for common data processing tasks.
x.comBlitz Chess with Jev V13
This experiment showcases Jev V13 competing in blitz chess matches against various frontier language models. It serves as a demonstration of how different AI architectures handle strategic decision-making and tactical gameplay within a competitive, time-constrained environment.
x.comCartShield
CartShield utilizes TypeSafe Jev to assist small and medium-sized businesses in managing checkout fraud. This tool provides a structured approach for evaluating transaction risks and automating disposition processes within e-commerce environments.
github.compdoom-protocol
This project presents a cooperative platform shooter where a TypeSafe Jev agent acts as a teammate. It demonstrates an experimental integration of autonomous AI agents within a fast-paced gaming environment to facilitate collaborative player-AI gameplay dynamics.
github.comjevfire
jevfire implements JEV-inspired parallel decision-making for CUDA-based LLMs. The project provides a vLLM-compatible API and game-agent examples to demonstrate how multiple concurrent decisions can be processed from a single context.
github.comswitchloom
switchloom is a deterministic model routing system designed for coding agents. It demonstrates a specialized mechanism for managing model selection, specifically highlighting capabilities related to Codex integration for automated programming tasks.
github.comConsoleChaosRacing with Jev
This project demonstrates the integration of Jev to manage driving decisions within a racing game interface. It highlights an experimental approach to using autonomous logic for real-time control inputs in a simulated gaming environment.
x.comJev Ad Focus Group Experiment
This experiment demonstrates using Jev to simulate thirty distinct personas for evaluating advertising content. It explores how automated agents can make scroll or stop decisions on hundreds of ads to analyze engagement patterns.
x.comask-jev-skill
This project provides a skill for Hermes and similar AI agents to interface with Jev. It demonstrates how to integrate external agent platforms with Jev services to facilitate structured, typesafe queries within automated workflows.
github.comJev AI Micro-Decision Framework
This guide outlines the core architecture of Jev AI, specifically detailing three distinct categories of micro-decisions. It serves as an educational resource for understanding how the system processes granular operational choices to facilitate automated decision-making workflows.
youtube.comjev-acento
This research project investigates how the Jev architecture handles Spanish language processing. It serves as an experimental audit to evaluate linguistic performance and character handling within the Jev framework.
github.comjev-trader
This project demonstrates an automated trading implementation using Jev on the Kuru MON-USDC market. It executes a single trading decision for every Monad blockchain block, showcasing how Jev agents can interact with decentralized exchange liquidity.
github.comtypesafeai-go
This repository provides a Go SDK for interacting with the TypeSafe AI API. It enables developers to integrate TypeSafe AI functionality directly into Go applications by providing structured client methods for API communication.
github.cominvalidate
invalidate provides an invalidation layer for AI memory systems built on TypeSafe Jev. It implements a lease-based mechanism where stored facts are automatically retired when new, conflicting evidence is introduced.
github.comcapture
capture is a Mac voice diary application that utilizes Jev for automated content sorting. The tool integrates directly with Notion to organize recorded entries, providing a streamlined workflow for users to document their thoughts and daily activities.
github.comagentgateway Jev Guardrail Example
This example demonstrates integrating Jev as a prompt guardrail within the agentgateway proxy. It showcases how developers can implement automated request filtering alongside built-in tracing and cost tracking features for LLM interactions.
github.comjev-mcp-dispatcher
This project demonstrates a method for executing Model Context Protocol tool calls using Jev. It enables developers to trigger specific tool functions through plain-language inputs without requiring an underlying large language model.
github.comJev Email Processing Demonstration
This demonstration showcases Jev performing automated classification, priority assessment, spam detection, and reply prediction tasks. The creator reports testing these capabilities across a dataset of 1,000 emails to evaluate the system's functional performance in managing inbox workflows.
youtube.comjev-semgrep
This tool enables semantic code searching by scoring lines against specific meanings. It allows developers to perform cross-language queries using logical operators to filter code based on conceptual intent rather than just literal text matches.
github.comjev-align
This command-line tool facilitates active learning by integrating human feedback with GEPA. It demonstrates a workflow for refining Jev decision definitions through iterative labeling and alignment processes.
github.comjev-bun1
This repository serves as an introductory guide for developers looking to integrate Jev into their projects using the TypeScript SDK. It demonstrates the fundamental setup steps required to begin working with TypeSafe AI components in a Bun environment.
github.comUnderstanding State in TypeSafe AI
This official documentation defines the concept of state within the TypeSafe AI framework. It explains how state management functions as a core component for maintaining consistency and context across complex AI interactions and system workflows.
docs.typesafe.aiVoice-Controlled Beat Creation
This demonstration showcases a user generating musical beats through real-time voice commands directed at Jev. It illustrates an experimental interface for audio production that relies on natural language input to manipulate and arrange sound elements.
x.comJevTest
JevTest provides a framework for bounded exploratory browser testing using Jev. It enables developers to implement deterministic assertions and generate replayable evidence to support automated quality assurance workflows within web environments.
github.comJev Minecraft Agent
This work-in-progress demonstration showcases the Jev model navigating the Minecraft environment. The creator illustrates the agent performing basic survival tasks, such as actively avoiding hostile entities like zombies during nighttime gameplay.
reddit.comKiru Hai Coach
This tool functions as a mahjong assistant that utilizes Jev to provide tile discard suggestions. It is designed to help players improve their decision-making process during games by offering strategic guidance in Japanese.
github.comTetris with Jev Real-Time Logic
This project demonstrates Jev playing Tetris by executing real-time block placement decisions. It showcases the application of Jev-based logic in a classic arcade environment to manage game state and piece movement dynamically.
x.comVisual Reference Finder
This tool demonstrates a Jev-powered workflow that retrieves and generates a collection of 100 images from archives like NASA and The Met. It showcases how users can quickly aggregate visual references based on a single descriptive prompt.
x.comWorkflow evals
This resource outlines the methodology used to evaluate System One workflows. It provides comparative performance data across various models, offering a structured approach for assessing AI workflow reliability and output quality in practical applications.
evals.typesafe.aiSpeculative fan-out
This TypeSafe AI reference explains the speculative fan-out pattern. It demonstrates how to improve system responsiveness by initiating multiple concurrent requests and selecting the first successful result to reduce overall latency in AI-driven workflows.
docs.typesafe.aisimple-jev
This project provides a utility to transform open-source language models into functional classifier or Jev endpoints. It demonstrates a streamlined approach for developers to implement decision-making capabilities using existing model architectures without requiring complex infrastructure.
github.comjevtest
jevtest provides semantic test matchers for Vitest and Jest using the Jev model. It allows developers to write test expectations in plain English and receive calibrated probability outputs for their assertions.
github.comMakerMap
MakerMap is a discovery tool that utilizes Jev technology to connect users with local makers. The creator claims the platform provides high-accuracy matching for community-based artisan searches, though these performance claims remain independent of verified third-party benchmarks.
x.com
