The TS SDK has a similar test, "Input Required Flow | should handle elicitation during tool execution" that verifies that the elicitation request can be received via a call to tasks/result and that it will call the normal elicitation handler similar to what's described in https://modelcontextprotocol.io/specification/2025-11-25/basic/utilities/tasks#task-augmented-tool-call-with-elicitation
https://github.com/modelcontextprotocol/typescript-sdk/blob/00249ce86dac558fb1089aea46d4d6d14e9a56c6/test/integration/test/taskLifecycle.test.ts#L579-L588
It doesn't look like we have similar logic in the server to include the elicitation/sampling request in the tasks/result response and in the client to parse the elicitation out of the tasks/result response and call the appropriate handler. Do we? I expect we need something similar to TS's TaskMessageQueue. If we don't have support for this, would it make sense to do this as a follow up? Or is it not worth doing?
Originally posted by @halter73 in #1170 (comment)
The TS SDK has a similar test, "Input Required Flow | should handle elicitation during tool execution" that verifies that the elicitation request can be received via a call to
tasks/resultand that it will call the normal elicitation handler similar to what's described in https://modelcontextprotocol.io/specification/2025-11-25/basic/utilities/tasks#task-augmented-tool-call-with-elicitationhttps://github.com/modelcontextprotocol/typescript-sdk/blob/00249ce86dac558fb1089aea46d4d6d14e9a56c6/test/integration/test/taskLifecycle.test.ts#L579-L588
It doesn't look like we have similar logic in the server to include the elicitation/sampling request in the
tasks/resultresponse and in the client to parse the elicitation out of thetasks/resultresponse and call the appropriate handler. Do we? I expect we need something similar to TS'sTaskMessageQueue. If we don't have support for this, would it make sense to do this as a follow up? Or is it not worth doing?Originally posted by @halter73 in #1170 (comment)