MCP Config Generator for Claude Code
Build a Claude Code MCP configuration without hand-writing JSON. Choose the transport, enter the server details, then copy the project-scoped CLI command or .mcp.json entry.
Remote HTTP is the preferred remote transport in current Claude Code documentation. SSE remains supported but is described as deprecated.
# Claude Code CLI
claude mcp add --scope project --transport http my-server https://mcp.example.com/mcp
# .mcp.json
{
"mcpServers": {
"my-server": {
"type": "http",
"url": "https://mcp.example.com/mcp"
}
}
}
How to use this tool
Choose HTTP for a remote server, stdio for a local process, or SSE only when required by a legacy server.
Enter the server name plus URL or local command.
Copy the CLI command or .mcp.json configuration and review it before trusting the server.
HTTP vs stdio vs SSE
Current Claude Code documentation recommends HTTP for remote MCP servers. stdio remains the natural option for local processes. SSE is still supported but documented as deprecated in favor of HTTP where available.
Project-scoped MCP configuration
Using --scope project stores shared MCP server definitions in .mcp.json so a team can version the configuration with the repository. Claude Code still requires workspace trust and approval for project servers.
Review trust boundaries
An MCP server can expose tools, APIs and data to Claude. Verify the server, understand its permissions and avoid committing secrets directly into shared configuration.
Related Claude resources
Frequently asked questions
What is .mcp.json?
It is the project-scoped MCP configuration file Claude Code uses for shared server definitions.
Which remote transport should I choose?
Use HTTP when the server supports it. Claude Code documentation describes HTTP as the recommended remote transport and SSE as deprecated.
Does Claude Code support local MCP servers?
Yes. Local MCP servers commonly use stdio and run as local processes.