---
title: "jev-guard"
description: "Prompt-injection and dangerous-action guard for coding agents, powered by Jev."
canonical: https://agentpluginsdirectory.com/plugins/jev-guard
last-updated: 2026-09-21
---

# jev-guard
Prompt-injection and dangerous-action guard for coding agents, powered by Jev.
- Slug: jev-guard
- Publisher: leepokai
- Repository: https://github.com/leepokai/jev-guard
- Manifest: plugin.json
- Version: 0.3.1
- License: MIT
- Category (editorial): other
- Skills: 0
- MCP servers: 0
- Stars: 18
- Repository created: 2026-09-17
- Repository last pushed: 2026-09-17
- Publisher type: User
- Listing: https://agentpluginsdirectory.com/plugins/jev-guard
- Schema: https://agent-plugins.org/schemas/1.0.0/plugin.schema.json

## What jev-guard does, in the publisher's words

jev-guard A security hook for coding agents, powered by Jev.

Claude Code's auto mode is described as: "A separate classifier model reviews actions before they run, blocking anything that escalates beyond your request, targets unrecognized infrastructure, or appears driven by hostile content Claude read." That is exactly the job jev-guard does, as three typed questions to Jev (risk, user_requested, from_untrusted) instead of a proprietary classifier, and it does it for Codex, Copilot, Gemini, Cursor, pi, OpenCode and ACP editors too, with the same policy and the same session memory everywhere. If you want auto mode outside Claude Code, or a second opinion inside it, this is the build.

Those two numbers are the whole reason this design works: cheap enough to run on every tool call and every tool result, fast enough that the agent doesn't notice, and typed so the policy lives in twenty lines of code you can read.

From the project README, punctuation lightly normalized. Full text: https://raw.githubusercontent.com/leepokai/jev-guard/HEAD/README.md