Skip to content
Metatext Decide — typed decisions for your agents, with a probability on every answerGet early access
Metatext
Sign Up

jcode

RAM-efficient Rust coding-agent harness, listed as using Jev decisions for memory pruning, browser navigation and voice routing.

ServiceproductionMIT
View on GitHub ↗

Overview

jcode is an MIT-licensed coding-agent harness written in Rust that emphasises resource efficiency. Its README benchmarks RAM per session against other agents, reporting 27.8 MB with local embedding off versus 140-370 MB for tools such as Pi, Codex CLI, OpenCode and GitHub Copilot CLI. It ships installers for macOS, Linux and Windows, a TUI with in-place updates, an SDK and documentation at jcode.sh.

The yibie list describes native TypeSafe Jev typed-decision transport in jcode, used for memory pruning, browser navigation and voice interaction routing. The fetched README excerpt does not discuss Jev, so the details of those decisions are not confirmed here and should be checked in the docs. For readers evaluating it as a Jev integration, the practical route is to check jcode.sh's documentation and SDK pages for the decision transport the list describes.

What the decision model decides

TypeQuestionAnswers
Yes / noShould this memory or context item be kept?(inferred)yes · no
ChoiceWhere should this voice interaction be routed?(inferred)—

How it works

  1. Agent session runs in the jcode harness
  2. Jev answers typed questions for pruning, navigation or voice routing
  3. Harness applies the decision and continues

Strengths

  • Very low memory footprint per session
  • Typed decisions for housekeeping avoid spending LLM tokens

Caveats

  • Jev integration not described in the README excerpt

Getting started

`curl -fsSL https://jcode.sh/install | bash` on macOS and Linux.

At a glance

Type
Service
Status
production
Language
Rust
Runtime
CLI/TUI agent harness
Has code
yes
License
MIT

Use cases and integrations

How curated lists describe it

  • Agent runtimes: RAM-efficient autonomous agent harness implemented in Rust with native TypeSafe Jev typed decision transport for memory pruning, browser navigation, and voice interaction routing.

    yibie/awesome-jev

Frequently asked questions

What does Jev do in jcode?

Per the list, typed decisions for memory pruning, browser navigation and voice routing.

Why choose jcode?

Its README emphasises low RAM use and fast boot compared with other harnesses.

How do I install it?

Run the install script from jcode.sh, or use Homebrew or a source build.

Related projects

  • jev-voice-browser

    Node app that controls Chromium by voice, asking Jev a dozen typed questions per partial transcript to act before you finish.

    ProjectprototypeJavaScript
  • jev-windows-voice

    Electron app for controlling Windows 10/11 by Turkish or English speech, using OpenAI Realtime, Jev decisions, UI Automation and Playwright.

    ProjectprototypeTypeScript
  • keys-MiniMax-Code-CLI-Browser-Scroll-Context-Enhancement-Pack-with-Jev-Ultrafast-Integration

    Enhancement pack for MiniMax Code CLI on local DGX Spark hardware, including a Jev-style local decision browser agent and context patches.

    MCP serverprototype
  • jev-demo

    Zero-build local playground for Jev with browser-agent, intent routing and context-pruning presets, including a Playwright agent.

    AppbetaJavaScript
  • jev-kit

    Claude Code kit wiring Jev into a tool-call guard, sub-agent sizing, file search, a browser agent, review, belay and compaction.

    ProjectbetaPython
  • jev-stack

    One-command npx installer for a set of Jev-powered agent tools: context compaction, browser agent, prompt enhancer and skill router.

    ProjectbetaJavaScript