🧠 Vibe Check MCP

Name: 🧠 Vibe Check MCP
Author: PV-Bhat

@PV-Bhat

🧠 Vibe Check MCP について

Vibe Check is a tool that provides mentor-like feedback to AI Agents, preventing tunnel-vision, over-engineering and reasoning lock-in for complex and long-horizon agent workflows. KISS your over-eager AI Agents goodbye! Effective for: Coding, Ambiguous Tasks, High-Risk tasks

#automation #mcp #error-handling #agent-based #workflow-automation #ai-agents #metacognition #agentic-workflow #agentic-ai #model-context-protocol #mcp-server #vibe-coding #vibecoding #model-context-protocol-servers

基本情報

カテゴリ

その他

ライセンス

MIT

ランタイム

node

トランスポート

stdio

公開者

PV-Bhat

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "vibe-check": {
      "command": "node",
      "args": [
        "/path/to/vibe-check-mcp/build/index.js"
      ],
      "env": {
        "GEMINI_API_KEY": "YOUR_GEMINI_API_KEY"
      }
    }
  }
}

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is Vibe Check MCP?

Vibe Check MCP is a lightweight, plug-and-play oversight server implementing Anthropic’s Model Context Protocol. It acts as an AI meta-mentor for LLM agents, using Chain-Pattern Interrupts (CPI) to prevent reasoning lock-in and keep agents on the minimal viable path. Research shows agents using Vibe Check improved success by 27% and halved harmful actions by 41%.

How to use Vibe Check MCP?

Run directly via npx with npx -y @pv-bhat/vibe-check-mcp start --stdio for MCP clients or --http --port 2091 for manual HTTP inspection. Requires Node >=20. Add the appropriate command block to your MCP client config. Optionally set API keys in a .env file (Gemini, OpenAI, Anthropic, or OpenRouter) and provide a sessionId for history continuity.

Key features of Vibe Check MCP

CPI adaptive interrupts that challenge assumptions and keep agents aligned.
Multi-provider LLM support: Gemini, OpenAI, Anthropic, and OpenRouter.
History continuity: summarizes prior advice when a sessionId is supplied.
Optional vibe_learn logs mistakes and fixes for future reflection.
Session constitution: enforce per-session rules via update_constitution, reset_constitution, and check_constitution tools.
Research-backed with published results; listed in Anthropic’s official MCP repo and MCP Registry.

Use cases of Vibe Check MCP

Preventing agents from over-engineering by providing quick sanity checks.
Keeping LLMs aligned and reflective before irreversible actions.
Reducing harmful agent behaviors in production or experimental workflows.
Maintaining consistency across sessions with per-session rule sets.
Integrating meta-cognitive feedback into agent system prompts for improved common-sense reasoning.

FAQ from Vibe Check MCP

What runtime does Vibe Check MCP require?

Node >=20 is required. It runs on any platform that supports Node.js.

What transport options are available?

STDIO transport for MCP-aware clients (Claude Desktop, Cursor, Windsurf, etc.) and HTTP transport for manual inspection or custom integrations. Default port for HTTP is 2091.

Is Vibe Check MCP still actively developed?

The project is in maintenance mode as of the latest release (v2.8.1). Active feature development has ended; only security and bug fix patches are published. Community forks are welcome under the MIT license.

Which LLM providers are supported?

Gemini (default), OpenAI, Anthropic, and OpenRouter. Provider is configured via environment variables and can be overridden with DEFAULT_LLM_PROVIDER.

How do I enforce rules per session?

Use the update_constitution tool to set rules for a given sessionId (e.g., “no external network calls” or “prefer unit tests before refactors”). Rules are honored by CPI interrupts. Clear them with reset_constitution or inspect them with check_constitution.