Jev Agent Kit
CLI and MCP tools for agents using TypeSafe Jev to make fast typed decisions: route, triage, guard, grep, rank, compact, and judge. Includes runnable use-case demos and an offline evaluation script.
OPEN SOURCE · CURATED REPOSITORIES
Curated Jev SDKs, tools, demos, integrations, benchmarks, and research projects. Repository metrics come from GitHub; titles, summaries, and categories are editorial.
Ranked by GitHub Stars
Ranked by recent Star growth and update activity
CLI and MCP tools for agents using TypeSafe Jev to make fast typed decisions: route, triage, guard, grep, rank, compact, and judge. Includes runnable use-case demos and an offline evaluation script.
PiJev is a terminal coding agent built on Pi that uses TypeSafe Jev to rank files, pick skills, and triage tool failures while the coding model writes code. The repository includes runnable code, tests, and controlled experiments showing fewer tool calls and faster file discovery.
An autonomous Android sub-agent powered by TypeSafe Jev/System One for durable task control via USB ADB or MCP.
A reference implementation that lets TypeSafe's Jev (System One) model drive a browser via Claude Code and Codex skills. It observes page controls, asks Jev multiple-choice questions, validates and executes the picks, and includes a runnable demo and trace explainer.
A Windows port of TypeSafe Computer Use that uses Jev decisions for desktop automation, with local OCR/UIA and MCP host integration.
An agent skill for discovering TypeSafe Jev opportunities, designing typed questions, and running helper scripts for local scans and evaluations.
A typed, confidence-aware agent skill routing CLI and library that calls the TypeSafe Jev model. Achieved 94.4% routing accuracy (68/72) on synthetic tasks.
Evidence-first long-term experiment memory skill for AI coding agents, with optional Jev decision-model layers for context triage and semantic audit.
A single-file, zero-dependency port of browser-use/jev-ultrafast. It uses TypeSafe Jev to control the Ego Lite browser with one typed-choice request per step, and includes multi-model benchmarks and safety guardrails.
OpenCode skill providing LLM agents with fast typed probabilistic judgments from TypeSafe AI's Jev via System One API, including PII redaction.
A fast browser agent for ego lite powered by TypeSafe Jev. One TypeSafe request per step; an agent or Jev picks the move.
A workflow layer that uses TypeSafe AI's Jev model to route AI coding agents through quality gates and verify evidence.