cloudsealed/cloudsealed-mcp

Overview

MCP server that exposes CloudSealed's billing-waste anomaly detector and architecture risk scorer as tools for AI agents. It integrates with clients like Claude Code and Cursor via the Model Context Protocol, letting agents invoke these analyses directly. Written in Python.

Best for

Best for
Developers using AI agents who want CloudSealed's cost and risk insights in their workflow

Use cases

• Detect cloud billing waste from an AI agent
• Score architecture risk during development
• Expose CloudSealed analysis to Claude Code or Cursor

How to use

Install

uvx cloudsealed-mcp

Tools exposed

• cloudsealed_analyze_billing_waste
• cloudsealed_score_architecture_risk
• cloudsealed_correlate_cost_and_risk

Tested with

Claude Desktop, Claude Code, Cursor, Windsurf

Example client config

{\n  "mcpServers": {\n    "cloudsealed": {\n      "command": "uvx",\n      "args": ["cloudsealed-mcp"]\n    }\n  }\n}

Notes

MCP server that exposes CloudSealed’s billing-waste anomaly detector and architecture risk scorer as tools for AI agents. It integrates with clients like Claude Code and Cursor via the Model Context Protocol, letting agents invoke these analyses directly. Written in Python.

0 stars on GitHub. Last updated 2026-08-09. Licensed MIT.

Use cases

• Detect cloud billing waste from an AI agent
• Score architecture risk during development
• Expose CloudSealed analysis to Claude Code or Cursor

Pros

• Standard MCP protocol works across multiple agent clients
• Combines billing-waste detection and architecture risk scoring in one server
• Python-based, easy to run for Python developers

Cons

• No community traction indicated by zero stars
• Scope limited to billing waste and architecture risk
• Requires setup of an MCP client to use

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.