Parth1811

JevFlow — AI skill for Claude Code

AI community

A Claude Code plugin that keeps your AI agents honest.

How to install JevFlow

This entry records only its repository, not the path inside it, so there is no exact command to give. Open Parth1811/JevFlow and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What JevFlow does

A Claude Code plugin that keeps your AI agents honest. Claude plans tasks as phases with checks, and when it tries to stop, Jevflow runs them and asks Jev if the work is really done. Many agents, one plan, with a live viewer to watch it all.

Alternatives in AI

Real-World Pain Points — Project: https://github.com/Narwhal-Lab/MagicSkills For many teams building multi-agent systems, the first thi 277 ★
Laravel AI Translator — High quality automatic translate your language files into many languages using AI like Claude, GPT and etc 250 ★
Vibesubin — AI skills that keep your code honest, your secrets unleaked, and your repo from rotting 50 ★

README

AI coding agents are great at starting work and bad at finishing it:

They declare victory early. "All tests pass!" while a test is still red, or the README never got written.
They drift. A long session wanders off the goal, or loses it entirely after a context compaction.
They trip over each other. Two sessions or a few subagents in one repo, and nobody knows who is doing what.
They just stop. A crash, a rate limit or a hung process ends the run half done.

**Jevflow** is a Claude Code plugin that keeps Claude accountable. For any real task, Claude lays the work out as a small plan of phases, each with a definition of done and a check. Every time Claude tries to stop, Jevflow runs the checks and asks [Jev](https://typesafe.ai/blog/introducing-system-one-models-and-jev), a fast, calibrated judgment model, where the work really stands: is this phase finished, is Claude stuck, off the goal, or claiming done too early? A small, fully tested policy then decides: **keep going**, **move to the next phase**, or **you are really done**. Several agents can share one plan, every flow in a project is remembered, and a live viewer