The most useful ones are memory and sequential thinking. Imo

How do you add these to chatgpt?

Chatgpt asks for a host for the mcp server.

All the MCPS I find give a config like

```{ "mcpServers": { "sequential-thinking": { "command": "npx", "args": [ "-y", "@modelcontextprotocol/server-sequential-thinking" ] } } }```

I still don’t fully understand the sequential thinking MCP. I have to assume those who like it did some kind of bake-off where they decided that the LlM has better results with it than without but I am skeptical.

It feels like wizardry a little to me.