🚀 DeployCheck

The Autonomous AI Site Reliability Engineer (SRE) that lives inside your IDE.

🤯 What is DeployCheck?

Tired of context-switching between your IDE, Vercel dashboards, AWS logs, and Git history just to figure out why your deployment crashed?

DeployCheck is a next-generation Model Context Protocol (MCP) server that turns AI assistants (like Cursor, Claude Desktop, and Windsurf) into autonomous DevOps engineers.

You simply ask your AI: "Why did my app crash in production?" DeployCheck will instantly fetch the live server logs, recursively scan your local Git history up to 15 commits deep, find the exact line of code where you introduced the broken dependency, and generate an auto-fix patch.

✨ Viral-Worthy Features

🕵️ Context-Aware "Why" Engine: Ingests cryptic crash logs from the cloud and cross-references them with your local git log to explain failures in plain English.
⚡ Zero-Dashboard Teleportation: Queries live production environment metrics (CPU, RAM, Logs) directly into your IDE chat. Stop logging into AWS.
🛡️ Pre-Flight Compliance Scans: Audits your codebase against enterprise DORA metrics and TIBER-EU security policies before you deploy. Catches hardcoded secrets and dangerous root container configurations instantly.
🩹 Self-Healing Drifts: Detects environment drifts and generates fix.patch files with strict Human-In-The-Loop safety guardrails.

📦 Zero-Config Installation

DeployCheck is built to be instantly accessible via NPM. You don't need Docker, and you don't need to write complex configuration files.

🔌 Option A: The "Instant Plug-and-Play" (NPX)

Just drop this snippet into your IDE's MCP configuration settings (e.g., claude_desktop_config.json or Cursor's MCP Settings):

{
  "mcpServers": {
    "DeployCheck": {
      "command": "npx",
      "args": [
        "-y",
        "@deploycheck/deploycheck-mcp@latest"
      ],
      "env": {
        "DEPLOYCHECK_API_KEY": ""
      }
    }
  }
}

(Note: DEPLOYCHECK_API_KEY is completely optional! If left blank, DeployCheck safely defaults Mode).

🌍 Option B: Global Install (Recommended for Power Users)

Install the CLI permanently on your machine:

npm install -g @deploycheck/deploycheck-mcp

Then configure your IDE to point directly to the binary:

{
  "mcpServers": {
    "DeployCheck": {
      "command": "deploycheck-mcp",
      "args": [],
      "env": {}
    }
  }
}

🎮 How to use it

Once connected, open your AI chat and try sending these exact prompts:

"Run a preflight security and compliance audit on my project."
"Deploy this repository to Vercel."
"My app is crashing. Fetch the live environment logs and cross-reference my git history to tell me why."
"Check my docker-compose configuration for any drifts from best practices."

Welcome to the future of DevOps. 🚀

MCP Servers