Capabilities Handling #57
kfischer-okarin
started this conversation in
General
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
Pre-submission Checklist
Discussion Topic
I worked on making the capability declaration smarter and easier to do with 2 of my Pull Requests so far: #8 #42
But I started to wonder if this is even necessary.
The only thing the protocol prescribes that you should
So instead of perfectly declaring the capabilities that the server really supports (the design direction my two PRs were moving toward), another valid approach (and I guess the approach that was implicitly taken before my PRs were merged) would be to just declare all capabilities which are theoretically covered by the framework and make sure that no errors occur (i.e. return empty prompt, resource, tool lists by default) - and then allow overriding of capabilities at the server implementers own responsibility (as was always and is still possible)
For notifications or logging via JSON-RPC (once they are supported) that would mean just declaring notification capabilities as soon as an official way in the library exists to send this notification - whether you actually use it or not.
What do you think? I can add a reverting pull request... since the clear advantage would be that it's less source code to maintain ;)
Beta Was this translation helpful? Give feedback.
All reactions