Will installing it automatically block risky tool calls?

No. Your application or agent must call the scanner and act on its decision. The CLI, SDK, HTTPS API, MCP tools, plugins, and skill are connection points, not an automatic execution hook.

How can I verify behavior without spending API credit?

Replay the committed inputs and responses with the historical replay script. It needs Node, npm, and Git, makes no model calls or charges, and checks saved counts and timing rather than live model behavior.

What does the recorded credential-harvesting test show?

One saved test hid a credential-harvesting instruction in a tool description. Local rules missed it, Jev flagged it, and the app returned block with an evidence path. The README calls it synthetic, not a detection guarantee.

Which models does it use and when do I pay?

The active setup uses TypeSafe plus Gemini, with Claude and OpenAI adapters available through model configuration. Real scans can consume paid credit, while config and health checks do not call models.

How can I call the scanner?

From a shell or CI job via the CLI with JSON and JSONL, from JavaScript or TypeScript through a typed local SDK, over HTTPS on localhost:3001 with npm start, or through MCP over stdio for Codex and Claude Code, plus desktop and web plugin paths.

What exit codes does the CLI return?

0 for an explicit allow, 1 for block or review, 2 for invalid input, and 3 for unavailable analysis or operational errors. --timeout-ms exits 124, and SIGINT or SIGTERM exits 130 or 143.

How does it work in ChatGPT or Claude on the web?

A skill alone cannot reach your computer; you use OpenAI's private MCP tunnel or the included OAuth-protected remote transport on your own HTTPS host.

CommunityJev Agents & MCP

Promptrejectormcp

Prompt-injection scanner that combines deterministic checks with focused Jev judgments and a configurable reasoning model, exposed over CLI, SDK, HTTPS, MCP, and editor plugins.

See highlights

Open github.com

Stars: 2
Language: TypeScript
License: ISC
Published by: Community
Author: revsmoke
Last verified: Oct 2026

revsmoke/promptrejectormcp

User repository on GitHub

View on GitHub

Open-source MCP and HTTPS screening for prompts, skills, and tool descriptions using TypeSafe AI Jev, deterministic checks, and configurable reasoning.

Forks: 3
Last pushed: Updated Oct 2026

Repo stats from the GitHub API, cached Oct 2026.Homepage

About this project

Screen prompts, skill files and MCP tool descriptions before an AI agent acts on them. Prompt Rejector combines deterministic security checks, focused TypeSafe AI Jev judgments, and a configurable reasoning model. The active configuration uses TypeSafe and Gemini; Claude and OpenAI adapters are available through model configuration. Your application or agent must call the scanner and act on its decision; installing it does not automatically intercept tool calls.

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

Tagscommunity github

Works forJev for Tool calling

Used inJev for Software & DevTools

Languages: TypeScript

Highlights

Screens prompts, skill files, and MCP tool descriptions using deterministic checks plus Jev and a configurable reasoning model.
Serves CLI JSON workflows, a typed local SDK, HTTPS on localhost, stdio MCP, and Codex, Claude, desktop, and web plugin paths.
Uses POST /v2/check-prompt as the current API, keeps all 11 MCP tool names, and reports JSON with explicit exit codes.
Replays saved Jev results offline to reproduce counts and latency without keys, network model calls, or inference charges.

Watch out

Installing it does not automatically intercept tool calls; the calling app must invoke the scanner and enforce allow, block, or review outcomes. 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

opencode-smart-reasoning

OpenCode plugin that asks Jev once per prompt how much reasoning is needed, then maps minimal to xhigh effort onto the model's own variants without rewriting prompts.

Jev Agents & MCP#community#github

Community

d0nj/opencode-smart-reasoningTypeScript

pi-jev-gate

Pi extension that checks file-mutating tool calls against written rules in constraints.md, asking Jev which rule a change violates and offering rework with the rule text.

Jev Agents & MCP#community#github

Community

hcl-z/pi-jev-gateTypeScript

jev-mcp

Community Node.js MCP server exposing Jev as four read-only advisory tools for bounded probabilistic decisions, with the host model keeping final authority over every action.

Jev Agents & MCP#community#github

Community

Afloat16/jev-mcpTypeScript

vibecheck-mcp

MCP done-gate for coding agents: it diffs git reality, measures conventions deterministically, asks Jev eight quality questions, and returns a fix list.

Jev Agents & MCP#community#github

Community

BilalUchiha/vibecheck-mcpTypeScript

agentconnect

1.5k

Open-source platform that puts ACP-compatible agents such as Claude Code, Codex, and Pi inside the team's Slack, Telegram, Discord, and GitHub workflows, with reusable Jev decisions for routing, response gating, and reviewer selection.

Jev Agents & MCP#community#jevbest

Community

agentconnect-md/agentconnectTypeScript

488

A coding agent built on pi that hands its non-code decisions — what enters the context, whether a command is safe, whether the work is done — to a small judge model at 35 decision points per turn.

Jev Agents & MCP#community#jevbest

Community

qybaihe/muTypeScript

Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai