feat(ai-core): trigger memory extraction after each conversation round

The WS hub and platform bridge now call Pipeline.FinishRound when the
response stream completes, recording the assistant reply and invoking
the LLM memory extractor asynchronously.

backend/cmd/cyrene/main.go

@@ -388,10 +388,12 @@ func loadAdaptersFromConfig(router *bridge.Router, cfg *bridge.BridgeConfig, orc
 	}
 
 	router.SetMessageHandler(func(msg *bridge.UnifiedMessage) {
-		tokenCh, _, err := orch.ProcessInput(context.Background(), "bridge-"+msg.Source, msg.Content)
+		sessionID := "bridge-" + msg.Source
+		tokenCh, _, err := orch.ProcessInput(context.Background(), sessionID, msg.Content)
 		if err != nil { return }
 		var fc string
 		for t := range tokenCh { fc += t.Content }
+		orch.FinishRound(sessionID, msg.SenderID, fc)
 		if fc != "" {
 			router.SendTo(msg.Platform, msg.Source, fc, bridge.SendOptions{MaxLength: 500, MinGap: 3 * time.Second})
 		}

backend/services/gateway/ws/hub.go

@@ -6,6 +6,7 @@ import (
 	"encoding/json"
 	"git.yeij.top/AskaEth/Cyrene_v2/internal/logger"
 	"net/http"
+	"strings"
 	"sync"
 
 	"github.com/gorilla/websocket"
@@ -22,6 +23,9 @@ var upgrader = websocket.Upgrader{
 // Orchestrator is the interface for the AI dialogue pipeline.
 type Orchestrator interface {
 	ProcessInput(ctx context.Context, sessionID, userMessage string) (<-chan model.StreamToken, *model.Intent, error)
+	// FinishRound records the assistant reply and triggers post-round work
+	// (e.g. memory extraction) after the response stream completes.
+	FinishRound(sessionID, userID, assistantReply string)
 }
 
 // Hub manages all active WebSocket connections.
@@ -168,13 +172,18 @@ func (c *Client) handleChatMessage(msg WSMessage) {
 	}
 
 	// Stream tokens
+	var fullContent strings.Builder
 	for token := range tokenCh {
+		fullContent.WriteString(token.Content)
 		if token.Done {
 			c.sendJSON(WSMessage{Type: "stream_end", SessionID: msg.SessionID})
 			break
 		}
 		c.sendJSON(WSMessage{Type: "stream_chunk", Content: token.Content, SessionID: msg.SessionID})
 	}
+
+	// Record the reply and trigger post-round work (memory extraction etc.)
+	c.hub.orchestrator.FinishRound(msg.SessionID, c.UserID, fullContent.String())
 }
 
 func (c *Client) writePump() {