vibe-check
Stop AI coding disasters before they cost you weeks. Real-time anti-pattern detection for vibe coders who love AI tools but need a safety net to avoid expensive overengineering traps.
claude mcp add --transport stdio kesslerio-vibe-check-mcp python -m vibe_check_mcp \ --env VIBE_CHECK_LOG_LEVEL="info" \ --env VIBE_CHECK_CONFIG_PATH="path/to/config.yaml (optional)"
How to use
Vibe Check MCP provides an AI-assisted coding safety net that monitors and interrupts potentially risky coding decisions in real time. It offers three modes of analysis: a Senior Engineer Mentor for collaborative reasoning, Fast Analysis for quick pattern detection and lightweight reviews, and Deep Analysis for comprehensive, Claude-powered reasoning with actionable remediation steps. The system is designed to catch anti-patterns such as building infrastructure before validating against standard APIs, treating symptoms rather than root causes, and unnecessary complexity escalation, all while providing multi-perspective feedback from engineers and product perspectives. To use it, run the MCP server using the provided command, then interact with the mentor via the MCP protocol to receive static, dynamic, or hybrid responses depending on query confidence. The built-in sampling and interrupt mode help prevent costly architectural mistakes during development.
How to install
Prerequisites:
- Python 3.8+ (the README indicates Python 3.8+ compatibility)
- pip (Python package installer)
- Basic knowledge of running Python modules
-
Set up a Python virtual environment (recommended): python3 -m venv venv source venv/bin/activate
-
Install the Vibe Check MCP package (example package name based on project): pip install vibe-check-mcp
-
Run the MCP server using the module entry point (as configured in mcp_config): python -m vibe_check_mcp
-
Verify the server starts successfully and listening on the expected interface/port as per your environment. If you need to customize configuration, set VIBE_CHECK_CONFIG_PATH or pass a config file path via environment variables as described in the docs.
Optional: If you distribute via pipx or another method, adapt the installation step accordingly (pipx run vibe-check-mcp).
Additional notes
Tips and notes:
- Ensure Python 3.8+ is installed to match compatibility expectations.
- If you encounter network or dependency issues, consider creating an isolated virtual environment and upgrading pip.
- Use VIBE_CHECK_LOG_LEVEL to adjust verbosity during debugging (debug, info, warning, error).
- The MCP environment supports an interrupt mode to halt risky suggestions in real time—familiarize yourself with how to trigger and interpret interrupts in your workflow.
- If there are configuration options, keep sensitive data out of logs; use secret redaction features if available.
- When upgrading, review changes in newer MCP versions (e.g., v0.5.1) for any breaking changes in command-line interface or required environment variables.
Related MCP Servers
claude-talk-to-figma
A Model Context Protocol (MCP) that allows Claude Desktop and other AI tools (Claude Code, Cursor, Antigravity, etc.) to read, analyze, and modify Figma designs
claude-code-open
Open source AI coding platform with Web IDE, multi-agent system, 37+ tools, MCP protocol. MIT licensed.
Pare
Dev tools, optimized for agents. Structured, token-efficient MCP servers for git, test runners, npm, Docker, and more.
metabase
Metabase MCP server provides integration with the Metabase API, enabling LLM with MCP capabilites to directly interact with your analytics data, this server acts as a bridge between your analytics platform and conversational AI.
codingbuddy
Codingbuddy orchestrates 29 specialized AI agents to deliver code quality comparable to a team of human experts through a PLAN → ACT → EVAL workflow.
claude-vigil
🏺 An MCP server for checkpointing and file recovery in Claude Code