Optimize usage dashboard: partial indexes, request cache, parallel histograms

- Add migration 14: partial covering indexes WHERE job_state='running'
  for user/project/subcluster groupings (tiny B-tree vs full table)
- Inline literal state value in BuildWhereClause so SQLite matches
  partial indexes instead of parameterized placeholders
- Add per-request statsGroupCache (sync.Once per filter+groupBy key)
  so identical grouped stats queries execute only once per GQL operation
- Parallelize 4 histogram queries in AddHistograms using errgroup
- Consolidate frontend from 6 GQL aliases to 2, sort+slice top-10
  client-side via $derived

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Entire-Checkpoint: 5b26a6e5ff10

internal/repository/migrations/sqlite3/14_running-partial-indexes.down.sql

@@ -0,0 +1,5 @@
+-- Reverse migration 14: Drop partial indexes for running jobs
+
+DROP INDEX IF EXISTS jobs_running_user_stats;
+DROP INDEX IF EXISTS jobs_running_project_stats;
+DROP INDEX IF EXISTS jobs_running_subcluster_stats;

internal/repository/migrations/sqlite3/14_running-partial-indexes.up.sql

@@ -0,0 +1,18 @@
+-- Migration 14: Partial covering indexes for running jobs
+-- Only running jobs are in the B-tree, so these indexes are tiny compared to
+-- the full-table indexes from migration 13. SQLite uses them when the query
+-- contains the literal `job_state = 'running'` (not a parameter placeholder).
+
+CREATE INDEX IF NOT EXISTS jobs_running_user_stats
+  ON job (cluster, hpc_user, num_nodes, num_hwthreads, num_acc, duration, start_time)
+  WHERE job_state = 'running';
+
+CREATE INDEX IF NOT EXISTS jobs_running_project_stats
+  ON job (cluster, project, num_nodes, num_hwthreads, num_acc, duration, start_time)
+  WHERE job_state = 'running';
+
+CREATE INDEX IF NOT EXISTS jobs_running_subcluster_stats
+  ON job (cluster, subcluster, num_nodes, num_hwthreads, num_acc, duration, start_time)
+  WHERE job_state = 'running';
+
+PRAGMA optimize;