Enhance User Input Buffering Logic for Delegated Agents

- Updated the BufferUserInput method to ensure only the root stack buffers user input, preventing duplication in delegated agents.
- Added comments in the Stream and processNextResponse methods to clarify that user input is already buffered by the root agent, allowing delegated agents to skip this step.
- Improved code clarity and maintainability by documenting delegation behavior in the context of user input handling.

agent/assistant/agent.go

@@ -176,6 +176,7 @@ func (ast *Assistant) Stream(ctx *context.Context, inputMessages []context.Messa
 			ctx.Logger.Debug("Create hook delegating to agent: %s", createResponse.Delegate.AgentID)
 
 			// Delegate to target agent (reuse existing delegation logic from next.go)
+			// Note: User input is already buffered by root agent, delegated agent will skip buffering
 			delegateResponse, err := ast.handleDelegation(ctx, createResponse.Delegate, streamHandler)
 			if err != nil {
 				finalStatus = context.ResumeStatusFailed

agent/assistant/chat.go

@@ -237,6 +237,12 @@ func (ast *Assistant) BufferUserInput(ctx *agentcontext.Context, inputMessages [
 		return
 	}
 
+	// Only root stack should buffer user input
+	// Delegated agents share the same buffer but should not duplicate user input
+	if ctx.Stack != nil && !ctx.Stack.IsRoot() {
+		return
+	}
+
 	// Convert input messages to buffer format
 	for _, msg := range inputMessages {
 		// Extract content from message

agent/assistant/next.go

@@ -15,6 +15,7 @@ func (ast *Assistant) processNextResponse(npc *NextProcessContext) (*agentContex
 	}
 
 	// Handle Delegate: call another agent
+	// Note: User input is already buffered by root agent, delegated agent will skip buffering
 	if npc.NextResponse.Delegate != nil {
 		return ast.handleDelegation(npc.Context, npc.NextResponse.Delegate, npc.StreamHandler)
 	}