Yardsort
Desktop app running coding agents in parallel git worktrees, with optional Jev judgments of each changed file against the task.
Overview
Yardsort is an open-source desktop app for Linux, macOS and Windows that runs terminal coding agents in parallel. Each task gets its own branch, git worktree and terminal running the agent of your choice, such as Claude Code, Codex, Grok, OpenCode or Cursor, so agents cannot disturb each other or your checkout. A side panel shows the files each one touches with live diffs. It is modeled on tools like Conductor and Superset with first-class support for all three operating systems.
An optional Assist feature, enabled only if you switch it on and add your own key, asks Jev small typed questions about each changed file against what the workspace was asked to do. The app also tracks token usage, estimated costs and machine resources per agent. Per the list, it is GPL-3.0.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | Does this changed file fit what the workspace was asked to do?(inferred) | yes · no |
How it works
- Describe a task and pick an agent.
- Yardsort creates a branch, worktree and terminal.
- Watch changes live; optionally ask Jev about each changed file.
- Merge work back.
Strengths
- Isolated parallel agents.
- Cross-platform.
- Optional, opt-in Jev review of changes.
Caveats
- Jev Assist requires your own key.
- Large desktop app.
Getting started
Download from https://github.com/joaoh82/yardsort/releases/latestAt a glance
- Type
- App
- Status
- beta
- Runtime
- desktop app (Linux, macOS, Windows)
- Has code
- yes
- Has tests
- yes
- License
- GPL-3.0
Use cases and integrations
How curated lists describe it
Desktop app that runs terminal coding agents in parallel git worktrees, with Jev judging each changed file against what the workspace was asked to do.
hellogumbo/awesome-jevDesktop app for running coding agents in parallel, each in its own git worktree, with optional Jev judgements on each change. GPL-3.0.
MrJev/awesome-jev
Frequently asked questions
What does Jev judge?
Each changed file against what the workspace was asked to do, if Assist is on.
Which agents can it run?
Claude Code, Codex, Grok, OpenCode, Cursor and others.
Is Jev required?
No, Assist is optional and off by default.
Related projects
Local desktop pal that verifies what a coding agent really did, with optional Laya or Jev double-checks of unclear test results.
ProjectbetaTypeScriptClaude Code plugin with rules and checks that make agents prove their work; an optional Jev judge triages review findings.
PluginbetaMarkdown + shellLinear-backed work ledger for coding agents where a Jev-compatible model verifies done reports, blockers and plan changes against the task.
ProjectprototypeJavaScriptStop hook for Claude Code and Codex that refuses 'done' until a check passes after the last edit; Jev advises but never causes a block.
CLIbetaTypeScriptSoftware-factory supervisor where Jev independently judges coding agents' completion, test sufficiency, progress and need for a human.
ProjectbetaPythonClaude Code skills that wait for AI pull-request reviewers and judge whether their reports actually permit a merge, never inferring a pass.
Projectbeta