feat(metrics): start module run durations buckets at 0.25s instead of 1s

This should better capture scripts that execute extremely quickly. From
testing and sampling, small scripts are often running on the order of
nanoseconds for noop type scripts and microseconds for very simple
things, for reference.

internal/manager/metrics.go

@@ -29,7 +29,7 @@ var (
 		prometheus.HistogramOpts{
 			Name:    "mango_manager_module_run_duration_seconds",
 			Help:    "Histogram of durations of how long it took a given module to run, in seconds",
-			Buckets: prometheus.ExponentialBuckets(1, 2, 8),
+			Buckets: prometheus.ExponentialBuckets(0.25, 2, 10),
 		},
 		[]string{"module", "script"},
 	)