CLI is the same API in more concise format. At minimum, the same amount of context overhead exist for MCP, but most of the time more because the boxes have size.
CLI can be secure, AWS CLI is doing just fine. You can also play simple tricks to hide secret in a daemon or run them remotely, and all of them are still smaller than a MCP.
I've always wondered: Doesn't the fact that the MCP input/output is more structured lead to higher reliability? With MCP you declare the types for input (string, int, list, etc) and output.
As part of our product, we have an MCP server. Since many of our MCP tools are expensive, for our tests we simply give the LLM all the tool descriptions (but in text form, not structured) and ask it which tool it would call for a given query and assert on the response.
The tests are flaky. In practice, I've always seen the LLM make the right tool call with the proper formatting of args, etc. In the tests (same LLM model), it occasionally makes mistakes on the argument types and it has to try again before it gets it right.
My assumption was that the structure MCP provides was the reason there was a discrepancy.
Tool discovery is only one very small part of the “context problem”. The bigger problem is the outputs. You can’t compost them. Ever watch Claude code run some crazy shit through python and jq to take some input, transform it in some crazy way and output exactly what it needs back into its context? You simply can’t do that with mcp. It’s basically forced to accept the exact shape of the mcp output into its context and then take that intermediate output and dump it right back into another tool. That is incredibly wasteful!
If your lucky the mcp might expose a way to ship its output into a text file so at least the agent can have a go at it with CLI tools.
CLI is the same API in more concise format. At minimum, the same amount of context overhead exist for MCP, but most of the time more because the boxes have size.
CLI can be secure, AWS CLI is doing just fine. You can also play simple tricks to hide secret in a daemon or run them remotely, and all of them are still smaller than a MCP.