fix(ci): resolve lint errors and test failures

- Fix errcheck: handle watcher.Close() and Set() return values
- Fix unparam: remove always-nil error from NewActivityWatcher
- Fix unparam: remove unused sinceTime param, delete dead code
- Fix version mismatch: update MCP __init__.py to 0.48.0
- Fix routing tests: change CWD so routing can find town root

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

cmd/bd/activity.go

@@ -185,13 +185,8 @@ func runActivityFollow(sinceTime time.Time) {
 	// Create filesystem watcher for near-instant wake-up
 	// Falls back to polling internally if fsnotify fails
 	beadsDir := filepath.Dir(dbPath)
-	watcher, err := NewActivityWatcher(beadsDir, activityInterval)
-	if err != nil {
-		// Watcher creation failed entirely - fall back to legacy polling
-		runActivityFollowPolling(sinceTime, lastPoll)
-		return
-	}
-	defer watcher.Close()
+	watcher := NewActivityWatcher(beadsDir, activityInterval)
+	defer func() { _ = watcher.Close() }()
 
 	// Start watching
 	watcher.Start(rootCtx)
@@ -261,69 +256,6 @@ func runActivityFollow(sinceTime time.Time) {
 	}
 }
 
-// runActivityFollowPolling is the legacy polling-based follow mode.
-// Used as fallback when ActivityWatcher cannot be created.
-func runActivityFollowPolling(sinceTime time.Time, lastPoll time.Time) {
-	ticker := time.NewTicker(activityInterval)
-	defer ticker.Stop()
-
-	consecutiveFailures := 0
-	const failureWarningThreshold = 5
-	lastWarningTime := time.Time{}
-
-	for {
-		select {
-		case <-rootCtx.Done():
-			return
-		case <-ticker.C:
-			newEvents, err := fetchMutations(lastPoll)
-			if err != nil {
-				consecutiveFailures++
-				if consecutiveFailures >= failureWarningThreshold {
-					if time.Since(lastWarningTime) >= 30*time.Second {
-						if jsonOutput {
-							errorEvent := map[string]interface{}{
-								"type":      "error",
-								"message":   fmt.Sprintf("daemon unreachable (%d failures)", consecutiveFailures),
-								"timestamp": time.Now().Format(time.RFC3339),
-							}
-							data, _ := json.Marshal(errorEvent)
-							fmt.Fprintln(os.Stderr, string(data))
-						} else {
-							timestamp := time.Now().Format("15:04:05")
-							fmt.Fprintf(os.Stderr, "[%s] %s daemon unreachable (%d consecutive failures)\n",
-								timestamp, ui.RenderWarn("!"), consecutiveFailures)
-						}
-						lastWarningTime = time.Now()
-					}
-				}
-				continue
-			}
-
-			if consecutiveFailures > 0 {
-				if consecutiveFailures >= failureWarningThreshold && !jsonOutput {
-					timestamp := time.Now().Format("15:04:05")
-					fmt.Fprintf(os.Stderr, "[%s] %s daemon reconnected\n", timestamp, ui.RenderPass("✓"))
-				}
-				consecutiveFailures = 0
-			}
-
-			newEvents = filterEvents(newEvents)
-			for _, e := range newEvents {
-				if jsonOutput {
-					data, _ := json.Marshal(formatEvent(e))
-					fmt.Println(string(data))
-				} else {
-					printEvent(e)
-				}
-				if e.Timestamp.After(lastPoll) {
-					lastPoll = e.Timestamp
-				}
-			}
-		}
-	}
-}
-
 // fetchMutations retrieves mutations from the daemon
 func fetchMutations(since time.Time) ([]rpc.MutationEvent, error) {
 	var sinceMillis int64