Will this override my explicit allow and deny rules?

No. The agent evaluates its own policy first and allow, deny, and configured ask outcomes are untouched. Jevvy only reviews other unresolved shell requests as complete commands and abstains unless every inquiry passes.

Does approving once permanently whitelist the command?

No. Approval covers only the current action, and any further prompt or check still runs. Anything uncertain falls through to the remaining permission flow unchanged.

Can a repo I clone loosen my approval policy?

It reads one global file that project repos cannot override, with JSONC comments and schema validation. The plugin uses only the selected provider and never falls back to another one.

Are the default thresholds just guesses?

Defaults were calibrated together against a bundled command corpus, and any must-ask auto-approval disqualifies a run. Replacing questions voids that evidence unless you recalibrate with the bundled skill.

What do I need to install Jevvy Permissions?

Node.js 22.19 or newer, then run npx @jevvy/permissions init. For OpenCode, use opencode plugin add @jevvy/permissions; for Claude Code, add the marketplace PanAchy/jevvy and install jevvy-permissions. Claude only sees readable ask settings, not managed, command-line, or session-only rules.

Where does its configuration live, and can a project change it?

One global file at ~/.config/jevvy/jevvy.jsonc, and project repositories cannot override it. JSONC comments and trailing commas are supported. The config picks a provider and, for providers other than Laya, can replace the default questions with your own risk-oriented nouls and thresholds.

Which providers can Jevvy point at?

OpenCode Zen, TypeSafe AI, OpenRouter, Vercel AI Gateway, local Laya, or a custom HTTP endpoint that accepts the System One request shape and returns typed answers. Each reads its key from the config or a matching environment variable, and the plugin uses only the selected provider, never falling back.

CommunityJev Agents & MCP

Jevvy

Permissions plugin for coding agents that reviews unresolved shell approval requests with Jev and auto-approves only harmless ones, leaving everything else in the normal permission flow.

See highlights

Open github.com

Stars: 20
Language: TypeScript
License: MIT
Published by: Community
Author: PanAchy
Last verified: Oct 2026

PanAchy/jevvy

User repository on GitHub

View on GitHub

Jev-powered plugins for coding agents

Forks: 1
Last pushed: Updated Oct 2026

Repo stats from the GitHub API, cached Oct 2026.

About this project

Fast probabilistic judgments inside coding-agent workflows. A plugin collection powered by TypeSafe's System One model. The permissions plugin auto-approves harmless shell permission requests. Anything uncertain continues through your agent's normal permission flow. In this isolated demo, a test reviewer requests the nonmatching OpenCode approvals. Claude checks readable ask settings; managed, command-line, and session-only rules are not visible to its hook. Node.js 22.19 or newer is required.

Quoted from the project's README. Read the full README.

Tagscommunity jevbest catalog-import

Works forJev for Classification Jev for Tool calling

Used inJev for Software & DevTools

Languages: TypeScript

Highlights

Agent policy runs first; Jevvy only judges leftover shell approvals and approves solely the current action.
Default four-Noul policy checks harmful, sensitive, untrusted, and obscured behavior against fixed thresholds.
One global config file outside repos, with six providers plus custom System One endpoints and no silent fallback.
OpenCode and Claude Code setup with demos, custom-question support, and a calibration skill producing JSONL evidence.

Watch out

Existing allow and deny decisions stay final and visible configured ask rules always remain human prompts. Custom endpoints use default policy unless replaced, and custom-question evidence only transfers when serving the calibrated model. Copy reviewed against the repository README on 2026-10-04, not from running it.

Questions about this entry

Short answers drawn from the linked project or page.

More Jev resources like this

duet-agent

Agent harness for work that outlives one chat, with resumable turns, routed state machines, and observational memory across sessions and processes.

Jev Agents & MCP#community#jevbest#catalog-import

Community

dzhng/duet-agentTypeScript

skilled-agent-harness_spec-driven-loops

Assistant framework that stores specs, memory, and specialist agents as plain markdown files in your repo for Claude Code, Codex, Opencode, and similar runtimes.

Jev Agents & MCP#community#jevbest#catalog-import

Community

MichelKerkmeester/skilled-agent-harness_spec-driven-loopsTypeScript

pi-jev-skill-picker

Pi extension that strips the full Agent Skills catalog from every request and loads only the skills a Jev ranking call judges relevant.

Jev Agents & MCP#community#jevbest#catalog-import

Community

safzanpirani/pi-jev-skill-pickerTypeScript

jevgrep

102

Behavioral code search that asks Jev to score repository fragments for a plain-language question and returns exact excerpts with paths and lines.

Jev Agents & MCP#community#jevbest#catalog-import

Community

nassim-arifette/jevgrepTypeScript

Multi-Agent

Local engine with TypeScript and Python SDKs that runs Claude Code and Codex sessions as a warm team with a durable mailbox.

Jev Agents & MCP#community#jevbest#catalog-import

Community

masonlee39/Multi-AgentTypeScript

oh-my-harness

Single-command harness that turns a plain-language stack description into synced docs plus blocking hooks for Claude Code, Codex, and Pi, with deterministic presets or one Jev chooser call.

Jev Agents & MCP#community#jevbest#catalog-import

Community

kyu1204/oh-my-harnessTypeScript

Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai