Its more targeted? Sure I can allow an LLM to access CLI, but then it will try to pipe to other command, write scripts etc. MCP is simply request/response. Its not really that technically different, but the way you and LLMs interact with it allows for easier control. IMHO.
I think this misses one of the main selling points of PHP, the shared nothing, single thread, fire and forget methodology that I can throw behind a load balancer and spin up 100 servers.
And the waiting is somewhat frustrating, what am I supposed to do while I wait? I could just sit and watch, or context switch to another task then forget the details on what I was originally doing.
It would be nice if I could tell it what page to look at (maybe you can, I am not sure). Often if I am getting an LLM to write some code that I can see is obviously wrong, I would love to say here is the docs ... use that to formulate your response.