Free tool
MCP Config Generator
Generate the mcpServers config to connect any Model Context
Protocol server to Claude Desktop, Cursor, VS Code, or Claude Code. Everything runs in your
browser — nothing is sent to any server.
Server
A short key for this server in your config.
Added as an Authorization: Bearer header.
Where does this go?
Save this into your MCP client's config file.
Restart your client after saving so it picks up the new server.
How MCP servers connect
The Model Context Protocol (MCP) lets AI agents call external tools and data. You register a server in your client's config; the client connects and exposes its tools to the model.
Remote (HTTP/SSE)
Point the client at a hosted server URL, with an optional Authorization header. Best for team/SaaS servers like SkillInvoke.
Local (stdio)
The client launches a local process (via command + args) and talks to it over stdio. Best for local filesystem or dev tools.
Tools, on demand
Once connected, the agent can list and call the server's tools mid-conversation — no code changes to your app.
Where each client stores MCP config
| Client | Config location |
|---|---|
| Claude Desktop | ~/Library/Application Support/Claude/claude_desktop_config.json |
| Cursor | ~/.cursor/mcp.json (or .cursor/mcp.json per project) |
| VS Code | .vscode/mcp.json |
| Claude Code | claude mcp add … (CLI) |
Give your agent your team's skills over MCP
SkillInvoke is a hosted MCP server for your team's approved AI skills — connect Claude, Cursor, or VS Code and invoke governed skills on demand.