The difference is that the server gives a description of the api it understands in enough detail that the llm can make use of it.
MCP is still going to be handy enough for iot type devices, where an llm can discover what's actually supported by that device without needing to query about the specific version.
Swagger / OpenAPI just aren't detailed enough to use without other documentation.
Skills & instructions will always have the limit that they run locally, so if they don't match the server there is a problem.
Literally no tolerance? I always figured that like everything else, there’s a cost/benefit analysis to be done, and you try and tweak enforcement levels to the point where it gives decent results without costing the earth. No law is enforced with literally no tolerance in practice.
You would not believe how bad publisher data is. I run a book website, and Dune is often tagged nonfiction in the data we get from publishers. I don't think they know how to use the BISAC system the industry uses (https://bisg.org/page/BISACEdition). With Dune, they were marking it "AI," which is a nonfiction-only category.
MCP is still going to be handy enough for iot type devices, where an llm can discover what's actually supported by that device without needing to query about the specific version.
Swagger / OpenAPI just aren't detailed enough to use without other documentation.
Skills & instructions will always have the limit that they run locally, so if they don't match the server there is a problem.
reply