feat: add benchmark filtering (--max-size, --category, --skip-longctx)

Both run-baseline.sh and run-suite.sh now support:
- --max-size GB: skip models larger than N GB (prevents OOM)
- --category LIST: filter by catalog category (smoke,dense,moe)
- --skip-longctx: skip 32K context tests (saves time + memory)
- --reps N: configure repetition count
- --help: shows usage with examples

Safe pre-optimization run: benchmark baseline --max-size 20 --skip-longctx
Full post-optimization: benchmark baseline (no filters, all models + longctx)

Also: 4 new BATS tests for flag parsing (98 total, all passing)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

bin/benchmark

@@ -11,10 +11,22 @@ case "${1:-help}" in
     compare)  exec bash "$SCRIPT_DIR/scripts/benchmark/compare.sh" "${@:2}" ;;
     *)
         echo "Usage: benchmark <command> [options]"
+        echo ""
+        echo "Commands:"
         echo "  setup     Ensure toolboxes and test models are ready"
         echo "  baseline  Capture pre-optimization baseline"
-        echo "  run       Run full benchmark suite (--tag NAME, --backends LIST)"
+        echo "  run       Run full benchmark suite"
         echo "  compare   Compare two runs (DIR1 DIR2)"
+        echo ""
+        echo "Filtering options (baseline and run):"
+        echo "  --max-size GB       Only models up to this file size"
+        echo "  --category LIST     Comma-separated: smoke,dense,moe"
+        echo "  --skip-longctx      Skip long-context (32K) tests"
+        echo "  --reps N            Standard test repetitions (default: 5)"
+        echo ""
+        echo "Examples:"
+        echo "  benchmark baseline --max-size 20 --skip-longctx"
+        echo "  benchmark run --tag post-opt --category moe"
         exit 1
         ;;
 esac