Relevance expansion is not straightforward as ID is missing

[Josipmrden]

There is no strategy what to do if a pivot node is found. ID is not passed (not saying that it should be passed), but there needs to be a strategy for which we can go from one pivot query to relevance expansion.

[antejavor]

So this is the key point, the agent has that capability inside the vector, text search (pivot) and then custom applying the relevance expansion for getting the node.

Does this makes sense in MCP/Agent experience to you? @Josipmrden