Vibe Check MCP Server

Monitor agent activity with research-backed Chain-Pattern Interrupt oversight.

Data last scanned yesterday · Reviewed 3 weeks ago

48★ 503TypeScriptMIT

Overview

Vibe Check MCP Server provides a plug-and-play mentor layer designed to keep AI agents aligned and focused on minimal viable paths. By implementing research-backed Chain-Pattern Interrupt (CPI) methodologies, the server aims to reduce over-engineering and harmful actions during complex task execution. Currently in maintenance mode, the project supports STDIO and streamable HTTP transports. It remains fully functional for developers seeking an oversight tool to evaluate agent reliability and reflection, with feature development limited to security and bug fix patches.

Our verdict

Vibe Check MCP Server carries a high-risk score of 48/100, primarily because its authentication methods and read-only status have not yet been editorially reviewed. As a community-maintained tool, it lacks official vendor backing. While it offers research-backed benefits for agent oversight, users should implement it with caution until security audits for its communication and authentication protocols are formalized.

Reduces harmful AI actions based on documented CPI research.
Maintains a small footprint with only 11 direct dependencies.
Active community with 10 contributors and 503 GitHub stars.
Version 2.9.0 remains fully functional for production oversight.

Project is currently in maintenance mode with no new features.
Authentication methods remain unreviewed for potential security vulnerabilities.
Read-only capabilities have not undergone editorial safety assessment.

Setup

Claude Desktop

{
  "mcpServers": {
    "vibe-check-mcp-server": {
      "args": [
        "-y",
        "@pv-bhat/vibe-check-mcp"
      ],
      "command": "npx"
    }
  }
}

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Some servers require API keys or tokens as environment variables — check the project's README and add an "env" object if needed.

No read-only mode — running this grants full read/write access.
No built-in authentication — anyone with access to this config can invoke every tool it exposes.

Compatibility

Client	Local	Docker	Remote	Read-only
ChatGPT
Claude Desktop
Cursor
VS Code
Windsurf

Frequently asked questions

›Is this project still being actively developed?

The project is currently in maintenance mode, meaning active feature development has ended. Only maintenance patches, such as security and bug fixes, are being released.

›What are the system requirements for running the server?

You must have Node.js version 20 or higher installed to run the server.

›Does the server support different communication methods?

Yes, the server supports both STDIO transport for MCP-aware clients and streamable HTTP transport for manual inspection.

›How can I verify that the HTTP server is running?

You can confirm the service is live by sending a request to the health check endpoint at http://127.0.0.1:2091/healthz.