Jev Guardrails for LLMs and Agents

coldteadotai/abide

Hooks into Claude Code, Codex, and OpenCode, and asks Jev one question per rule whether each edit breaks your AGENTS.md or CLAUDE.md rules.

dealerdefi/Jevmind

Nine skills over one brain: a shell-command gate, a diff triage, a router and more, each decision appended to a hash-chained ledger that names any record edited afterwards. Runs offline on local reflexes or against Jev.

DevMortimer/pi-warden

Guardrails for the Pi coding agent. Jev judges every write and edit against the rules in pi-warden.md.

leepokai/jev-guard

Risk-scores every tool call against session context into deny, ask, or allow, and flags prompt injection in tool results.

qkal/Canny

Stop hook for Claude Code and Codex CLI that refuses a "done" while no check has passed since the last edit. Jev can only relax that refusal and never cause one, and it still blocks with no API key at all.

valentynkit/jev-belay

Claude Code Stop hook that checks the transcript for evidence before letting a "done" through, and spends one four-question Jev call only when files changed with no passing check since. Fails open on every error path.

luantak/is-malicious

Sends source, configuration, build, and CI files to Jev and points at the files and lines that look deceptive or data-stealing. Its README says a clean report is not proof a project is safe.

jev-sec/jev-ids

Intrusion detection that shows Jev one network flow and five labelled examples, benchmarked against LLM and classical baselines.

RiskAverseTech/toolgate

A calibrated tool-call firewall for agents, as a Claude Code mod or PreToolUse hook or as an MCP proxy in front of any MCP server: Jev checks each call for being off-task, breaking an explicit constraint or making a choice reserved for you.

devtooligan/jevscan-evm

Produces a heat map of likely bugs in EVM code. The author's own warning: a proof of concept whose code they did not read.

dark-hxx/jev-safety-gateway

Go reverse proxy that sits between nginx and a model backend, judges each request's user input, blocks the harmful ones and passes the rest through untouched. Chinese.

buberlo/dsh-jev

Decision layer for DeepSeek Harness whose failure policy rejects the value allow at configuration time, from plain JavaScript as well as TypeScript. Its verify script installs the built tarballs into a fresh consumer before testing them.

agent-chaperone/agent-chaperone

MCP proxy that screens a tool call before it runs and the tool result before the agent reads it, combining rules in code with typed judgments. Starts in a shadow mode that blocks nothing.

zszz3/Pi-Jev-Guide

Pi coding-agent plugin with rules configured by timing, plus risk checks, output redaction, and reminders on repeated failures. Chinese documentation.

everafterlabs/jes

Open-source guardrails for AI agents that check prompts, skills, retrieved content, tool calls and responses for prompt injection with a decision model such as Jev or Laya, with thresholds you set; LangChain integration and lessons. Apache-2.0.

kiwi0719/jev-edge

Admission control at the traffic edge: an OpenResty module, with Lua and JavaScript packages, that screens requests for prompt injection before they reach the app.

Kevthetech143/super-jev

Puts one judge between an agent and your data: it finds the file, checks the claim, and permits or refuses the action.

eugeniughelbur/jev-engineering

Two coding-agent tools on Jev with benchmarks you can rerun: jev-gate judges each shell command an agent is about to run, and review-router, a GitHub Action, decides from a diff whether an AI reviewer must read the whole pull request.

berget-ai/guardrails-md

Gate for opencode, pi and Claude Code that scores every bash command for destruction, secret leaks and breaches of your GUARDRAILS.md, and blocks it above a threshold with a reason the agent can act on.

atomicpages/sensored

Streaming PII redaction for TypeScript with regex and NER detectors, and an opt-in step where Jev confirms a detection before it is redacted.

noelzappy/tripwire

Runs seven checks on every LLM response in one Jev call, as AI SDK middleware or an OpenAI-compatible proxy.

rashedInt32/jev-gates

Seven calibrated gates for Claude Code (rules, scope, intent, done, claims, proof, and commit honesty) that escalate but never approve.