vibe-check-mcp-server MCP — 사용 사례, 설치 & 라이브 데모

Name: vibe-check-mcp-server MCP Server
Author: PV-Bhat

왜 쓰나요

핵심 기능

vibe_check tool interrupts with pushback before irreversible actions
vibe_learn captures per-session lessons for in-context improvement
Constitution tools define per-session rules the agent must follow
Pure prompt-wrapping — no data leaves your chosen LLM provider

라이브 데모

실제 사용 모습

vibe-check.replay ▶ 준비됨

0/0

설치

클라이언트 선택

~/Library/Application Support/Claude/claude_desktop_config.json · Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "vibe-check": {
      "command": "npx",
      "args": [
        "-y",
        "vibe-check-mcp-server"
      ],
      "_inferred": true
    }
  }
}

Claude Desktop → Settings → Developer → Edit Config 열기. 저장 후 앱 재시작.

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "vibe-check": {
      "command": "npx",
      "args": [
        "-y",
        "vibe-check-mcp-server"
      ],
      "_inferred": true
    }
  }
}

Cursor는 Claude Desktop과 동일한 mcpServers 스키마 사용. 프로젝트 설정이 전역보다 우선.

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "vibe-check": {
      "command": "npx",
      "args": [
        "-y",
        "vibe-check-mcp-server"
      ],
      "_inferred": true
    }
  }
}

Cline 사이드바의 MCP Servers 아이콘 클릭 후 "Edit Configuration" 선택.

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "vibe-check": {
      "command": "npx",
      "args": [
        "-y",
        "vibe-check-mcp-server"
      ],
      "_inferred": true
    }
  }
}

Claude Desktop과 같은 형식. Windsurf 재시작 후 적용.

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "vibe-check",
      "command": "npx",
      "args": [
        "-y",
        "vibe-check-mcp-server"
      ]
    }
  ]
}

Continue는 맵이 아닌 서버 오브젝트 배열 사용.

~/.config/zed/settings.json

{
  "context_servers": {
    "vibe-check": {
      "command": {
        "path": "npx",
        "args": [
          "-y",
          "vibe-check-mcp-server"
        ]
      }
    }
  }
}

context_servers에 추가. 저장 시 Zed가 핫 리로드.

claude mcp add vibe-check -- npx -y vibe-check-mcp-server

한 줄 명령. claude mcp list로 확인, claude mcp remove로 제거.

사용 사례

실전 활용법: vibe-check-mcp-server

Sanity-check a plan before a destructive action with vibe-check

👤 Anyone running agents with write access ⏱ ~10 min beginner

언제 쓸까: The agent is about to delete files, run a migration, or deploy — you want a second opinion.

흐름

Invoke vibe_check on the plan

Before running the migration, call vibe_check on this plan: [paste]. What could go wrong?✓ 복사됨

→ Pushback list: unchecked assumptions, missing rollback, scope creep
Resolve or ignore each concern explicitly

Address each concern — either mitigate or explicitly dismiss as out of scope.✓ 복사됨

→ Annotated plan ready to run

결과: Fewer 'oh no' moments after destructive actions.

함정

Agent dismisses all concerns rubber-stamp style — Require the agent to restate each concern and the specific mitigation

함께 쓰기: sequentialthinking-tools

Enforce session rules with vibe-check's constitution

👤 Teams sharing agent workflows ⏱ ~10 min intermediate

언제 쓸까: You want certain rules (never push to main, always write tests first) enforced across a session.

흐름

Set the constitution

Use update_constitution to set these rules for this session: [list]. Enforce via CPI on every tool call that could violate them.✓ 복사됨

→ Rules acknowledged; subsequent violations get interrupted
Inspect when in doubt

Call check_constitution to see what's currently in effect.✓ 복사됨

→ Current rules list

결과: Consistent agent behavior within a session without repeating rules every message.

Capture lessons from agent mistakes with vibe_learn

👤 Agent developers iterating on prompts ⏱ ~5 min beginner

언제 쓸까: The agent made a classic mistake; you want it not to repeat.

흐름

Log the lesson

Call vibe_learn with: mistake='deleted prod table without snapshot', lesson='always take a snapshot before DDL on prod'.✓ 복사됨

→ Lesson stored for the session
Rely on it later

(later in same session) Attempt a DDL — vibe_check surfaces the stored lesson✓ 복사됨

→ Interrupt with the relevant past lesson

결과: In-session memory of 'don't do that again' without prompt engineering.

조합

다른 MCP와 조합해 10배 효율

vibe-check + sequentialthinking-tools

Plan + pushback loop

Plan with sequentialthinking_tools, then call vibe_check on the plan before executing step 1.✓ 복사됨

vibe-check + overture

Visual plan + vibe-check layer

After Overture produces the plan graph, run vibe_check on it and surface concerns as annotations.✓ 복사됨

도구

이 MCP가 노출하는 것

도구	입력	언제 호출	비용
vibe_check	plan: str, context?: str	Before any non-trivial or irreversible action	1 LLM call via your configured provider
vibe_learn	mistake: str, lesson: str	Immediately after a mistake, while fresh	0
update_constitution	rules: str[]	Session start or policy change	0
check_constitution		Debug unexpected interrupts	0
reset_constitution		Start fresh	0

비용 및 제한

운영 비용

API 쿼터: Uses your chosen LLM's quota
호출당 토큰: 500-3000 for vibe_check
금액: LLM token cost only
팁: Only vibe_check on substantive plans — skipping it on trivial edits saves tokens

보안

권한, 시크릿, 파급범위

자격 증명 저장: LLM API key in env (OPENAI_API_KEY, ANTHROPIC_API_KEY, etc.)

데이터 외부 송신: Only to your configured LLM provider

문제 해결

자주 발생하는 오류와 해결

vibe_check returns empty concerns

Plan input was too vague — include concrete steps and side effects

Agent ignores vibe_check output

Make the system prompt require addressing each concern before proceeding

npx cannot find @pv-bhat/vibe-check-mcp

The published scope is @pv-bhat; pin latest explicitly: npx -y @pv-bhat/vibe-check-mcp@latest

대안

vibe-check-mcp-server 다른 것과 비교

대안	언제 쓰나	단점/장점
sequentialthinking-tools	You want planning, not pushback	Different shape — no adversarial second model
overture	You want visual gating instead of text pushback	Heavier setup

vibe-check-mcp-server