feat: SuperCache Lab interactive explorer

[Code0987]

Summary

• SuperCache Lab (examples/lab): React + Vite explorer for every keyspace mode, with a 3-node canvas and attach-to-remote cache gRPC.
• Read-only Engine.LocalView / BloomDump plus bloom.Indexes so the UI can show owner/replica/tombstone and the Bloom bit grid.
• Default go run ./examples/lab is UI-only (no mesh); -cluster or the in-page Connect / Local 3-node control starts or attaches a backend.

Test plan

• go test ./pkg/engine ./examples/lab
• go test ./... (local, before UI polish)
• CI test + bench; read <!-- supercache-bench-comment --> (Get-hit allocs flat, shared cells ±10%)

[github-actions]

SuperCache bench vs main

Both sides ran once on this same GitHub runner (main, then this PR). Not a merge gate.

Baseline: 02d0b31 (2026-09-07T09:47:39Z) → this run: 02d0b31 (2026-09-07T09:46:55Z)

scbench smoke

cell	ops/s (main → PR)	Δ ops/s	p50	p99
get/hit n=1 c=1 uniform	11850 → 11887	+0.3%	84µs → 84µs	145µs → 145µs
get/hit n=1 c=10 uniform	44943 → 44967	+0.1%	204µs → 204µs	595µs → 600µs
get/hit n=3 c=10 uniform	51578 → 51888	+0.6%	171µs → 170µs	661µs → 650µs
miss/miss-cacheonly n=1 c=1 uniform	11593 → 11566	-0.2%	85µs → 85µs	162µs → 161µs
miss/miss-cacheonly n=1 c=10 uniform	46660 → 45392	-2.7%	196µs → 199µs	612µs → 641µs
set/put n=1 c=1 uniform	11262 → 11438	+1.6%	87µs → 86µs	156µs → 150µs
set/put n=1 c=10 uniform	42973 → 43820	+2.0%	213µs → 209µs	637µs → 616µs

Δ ops/s: green-ish is higher throughput. Latency: lower is better.

go test -bench (ns/op)

benchmark	ns/op (main → PR)	Δ	allocs/op
BenchmarkEngineDelete	604 → 585	-3.2%	6 → 6
BenchmarkEngineGetHit	914 → 914	+0.0%	15 → 15
BenchmarkEngineGetHitParallel	963 → 961	-0.2%	15 → 15
BenchmarkEngineGetMissCacheOnly	661 → 655	-0.9%	14 → 14
BenchmarkEngineGetMissCacheOnlyParallel	532 → 518	-2.7%	14 → 14
BenchmarkEngineGetMissLoadThrough	1272 → 1258	-1.1%	24 → 24
BenchmarkEngineGetMissLoadThroughParallel	1411 → 1407	-0.3%	23 → 23
BenchmarkEnginePut	1200 → 1185	-1.2%	18 → 18
BenchmarkEnginePutParallel	1070 → 1115	+4.2%	18 → 18
BenchmarkStoreBitSet	89 → 87	-1.7%	1 → 1
BenchmarkStoreCGetHit	56 → 56	+0.4%	0 → 0
BenchmarkStoreCIncr	79 → 79	+0.6%	1 → 1
BenchmarkStoreCMSIncr	94 → 94	+0.5%	0 → 0
BenchmarkStoreCMSIncrN	93 → 93	+0.1%	0 → 0
BenchmarkStoreGeoAdd	7450 → 7401	-0.7%	2 → 2
BenchmarkStoreGeoPosHit	68 → 69	+0.6%	0 → 0
BenchmarkStoreGeoPosMiss	67 → 67	-0.3%	0 → 0
BenchmarkStoreGetHit	226 → 228	+1.0%	2 → 2
BenchmarkStoreGetHitParallel	271 → 276	+1.8%	2 → 2
BenchmarkStoreHGetHit	71 → 71	+0.3%	1 → 1
BenchmarkStoreHGetMiss	66 → 66	+0.1%	0 → 0
BenchmarkStoreHLLAdd	77 → 79	+2.3%	0 → 0
BenchmarkStoreHSet	8251 → 8491	+2.9%	3 → 3
BenchmarkStoreJsonSet	20240 → 21173	+4.6%	207 → 207
BenchmarkStorePut	235 → 242	+2.9%	2 → 2
BenchmarkStorePutParallel	272 → 274	+0.7%	2 → 2
BenchmarkStoreSetAdd	7176 → 7299	+1.7%	2 → 2
BenchmarkStoreSetContainsHit	67 → 68	+1.6%	0 → 0
BenchmarkStoreSetContainsMiss	65 → 66	+0.7%	0 → 0
BenchmarkStoreTopKAdd	772 → 778	+0.8%	20 → 20
BenchmarkStoreVAdd	83669 → 85250	+1.9%	514 → 513
BenchmarkStoreVSim	73780 → 71723	-2.8%	530 → 530
BenchmarkStoreZAdd	7419 → 7592	+2.3%	2 → 2
BenchmarkStoreZScoreHit	67 → 67	+1.1%	0 → 0
BenchmarkStoreZScoreMiss	67 → 67	+0.9%	0 → 0

_{Posted by CI bench job. Not a merge gate.}