fix(agent): guard async maintenance goroutine and add timeout to diagnosis ctx

- Add stopCh check in runPostResponseMaintenanceAsync to respect agent
  shutdown, preventing goroutine leak on Agent.Stop()
- Replace bare context.Background() in handleTraderDiagnosisSkill with
  a 30s timeout context for proper deadline propagation

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

agent/planner_runtime.go

@@ -2628,6 +2628,12 @@ func (a *Agent) runPostResponseMaintenanceAsync(userID int64) {
 		}()
 		ctx, cancel := context.WithTimeout(context.Background(), 20*time.Second)
 		defer cancel()
+		// Respect agent shutdown: abort early if stopCh is closed.
+		select {
+		case <-a.stopCh:
+			return
+		default:
+		}
 		a.maybeUpdateTaskStateIncrementally(ctx, userID)
 		a.maybeCompressHistory(ctx, userID)
 	}()

agent/skill_execution_handlers.go

@@ -7,6 +7,7 @@ import (
 	"regexp"
 	"strconv"
 	"strings"
+	"time"
 
 	"nofx/mcp"
 	"nofx/store"
@@ -2757,7 +2758,9 @@ func (a *Agent) handleTraderDiagnosisSkill(storeUserID, lang, text string) strin
 	}
 
 	evidence := a.collectTraderDiagnosisEvidence(storeUserID, target.ID, target.Name)
-	if answer, ok := a.generateTraderDiagnosisAnswerWithLLM(context.Background(), lang, text, evidence); ok {
+	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
+	defer cancel()
+	if answer, ok := a.generateTraderDiagnosisAnswerWithLLM(ctx, lang, text, evidence); ok {
 		return answer
 	}
 	return formatTraderDiagnosisEvidence(lang, evidence)