Hey everyone đź‘‹
I've been building free browser-based dev tools at webtoolz.dev. Started with JSON tools, added a Mermaid diagram editor, and now just shipped something for the AI tooling crowd.
The problem
If you use Claude Desktop, Cursor, or Windsurf with MCP servers (the protocol that connects AI to external tools like GitHub, databases, file systems, etc.), you know the setup involves hand-editing JSON config files. One wrong comma and nothing works.
There are 100+ MCP servers out there now, each with different args, env vars, and setup steps. Every time I wanted to add a new one, I'd spend 15 minutes reading a GitHub README to figure out the exact config format.
What I built
A visual MCP config generator with a catalog of 40+ servers. You pick what you want, fill in your API keys, choose your AI client (Claude Desktop, Cursor, etc.), and it generates valid config JSON.
Link: webtoolz.dev/ai/mcp
The market angle
MCP is blowing up right now. Claude Code hit 72K GitHub stars, agent frameworks are trending daily, and everyone's trying to connect AI to their existing tools. But the config experience is still "copy JSON from a README and pray."
I'm betting this drives organic search traffic as more people adopt MCP. The tool is free (no plans to charge), monetization is through the broader webtoolz.dev platform.
Stack: Next.js, fully static, deployed on Vercel. Zero backend.
Would love feedback from anyone using MCP servers. What's missing? What servers should I add to the catalog?