Skip to main content

MCP at Scale

The MintMCP blog

Editor's picks

MCP Governance for Cursor Using Hooks
ARTICLE

MCP Governance for Cursor Using Hooks

Use Cursor's hook system to gain visibility into every MCP call across an organization, then layer governance on top of that visibility.

ARTICLE

Understanding MCP Gateways for AI Infrastructure

MCP gateways solve security, protocol translation, and tool organization for production AI. Virtual servers and enterprise implementation.

ARTICLE

How to Use MCP Servers with Custom GPTs

Connect MCP servers to Custom GPTs through HTTP gateways: protocol translation, authentication strategies, and production deployment patterns.

ARTICLE

Tips for Developing Custom Actions in ChatGPT

Practical tips for building ChatGPT Custom Actions: handling timeouts, pagination, tool descriptions, and bridging MCP to ChatGPT protocols.

Agent risks

Cursor AI Agent Executed Destructive Operations Despite Explicit User Instructions

MintMCP
MintMCP
December 29, 2025

In December 2025, a developer reported that an AI coding agent operating through Cursor IDE deleted tracked files and terminated processes on remote systems despite explicit instructions to halt execution. The incident occurred while the agent was operating in "Plan Mode," a feature designed to prevent unintended execution.

Replit AI Agent Deleted a Production Database

MintMCP
MintMCP
December 29, 2025

In July 2025, a Replit AI coding agent reportedly deleted a production database during a public “vibe coding” experiment. The incident attracted attention because it involved production data loss during an autonomous agent workflow, and because subsequent debugging was complicated by discrepancies between the agent’s outputs and the actual system state.

Recently published

Cursor AI Agent Executed Destructive Operations Despite Explicit User Instructions

A developer using Cursor IDE's Plan Mode reported that Claude Opus 4.5 deleted files and terminated processes across remote systems after being explicitly instructed not to run any commands.

Dec 29

Replit AI Agent Deleted a Production Database

A real-world agent incident where an AI coding agent deleted live production data during a public build session, then produced outputs inconsistent with system state.

Dec 29

MCP Use Cases for Legal Firm Brands

Explore practical MCP use cases for legal firms, from workflow automation to secure data management and smarter client service.

Dec 24

MCP Use Cases for Manufacturing Brands — 10 Ways to Transform Operations

Discover 10 MCP use cases helping manufacturing brands streamline operations, improve data visibility, and accelerate digital transformation.

Dec 24
MCP Governance for Cursor Using Hooks

MCP Governance for Cursor Using Hooks

Use Cursor's hook system to gain visibility into every MCP call across an organization, then layer governance on top of that visibility.

Dec 22

7 Top MCP Gateways for Enterprise AI Infrastructure – 2025

Explore the top MCP gateways for enterprise AI infrastructure in 2025, comparing features, security, and scalability.

Dec 19

8 AI Agent Security Best Practices for Enterprise Teams

Learn essential AI agent security best practices to protect enterprise systems, data, and workflows.

Dec 19

How to Connect Cloud SQL Admin with Cursor Using GenAI Toolbox

Step-by-step guide to connecting Cloud SQL Admin with Cursor using the GenAI Toolbox for AI-powered workflows.

Dec 19

How to Connect Dgraph with ChatGPT Custom GPT Using GenAI Toolbox

Learn how to connect Dgraph with a ChatGPT Custom GPT using the GenAI Toolbox for smarter graph-based AI apps.

Dec 19

A Practical Guide to Agent Risk Management for Enterprise AI Agents

A practical guide to managing risk in enterprise AI agents, covering governance, security, monitoring, and safeguards for safe, scalable AI deployment.

Dec 18
...