Quick Start
Getting Started with MCPHero
Introduction
MCPHero is a managed platform for building MCP servers using AI (aka vibecode) and deploying them in one click. It’s inspired by platforms like v0 and Lovable, but focused specifically on MCP servers.
How MCPHero Works
MCPHero is built around a structured AI Chat UI. Instead of a free-form chat, the AI follows a strict, step-by-step flow:
Conversation Step
A natural conversation where the AI guides the user and understands what kind of MCP server is needed.
Tool Selection Step
The AI suggests a set of tools based on the conversation. The user then selects which tools should be included in the MCP server.
Auth Configuration Step
Configure how you authenticate with MCP server - header, OAuth, or no authentication at all.
Values Step
If the MCP server requires credentials, connection strings, or other runtime values, the AI will ask for them explicitly in this step.
Deploy Step
The final step defines where and how the MCP server is deployed. Available deployment options depend on the user’s plan.
This rigid structure significantly reduces AI hallucinations and limits room for error, ensuring generated MCP server can be successfully run.
Servers
MCPHero supports multiple deployment options depending on your plan.
Shared Runtime : A managed, shared environment designed for experimentation and the free tier.
- Shared compute and limited resources
- No ephemeral credentials or runtime-injected values
- Limited, whitelisted Python libraries
- Free tier only
Dedicated VPS : A fully isolated environment with maximum flexibility and control.
- Dedicated compute resources
- Ephemeral credentials and arbitrary runtime values supported
- Full freedom over code and dependencies
- Paid plans only