bench(fullhistory): interactive HTML results explorer

cmd/stellar-rpc/scripts/bench-fullhistory/README.md

@@ -215,4 +215,6 @@ bench-fullhistory build-txhash-index --in-dir=/path/to/out/cold/txhash
 - `corpus.go`, `cache*.go`, `tx_hash_helpers.go`, `metrics_helpers.go` — shared helpers.
 - `run-all-benches.sh` — suite driver (builds once, runs every read + ingest
   bench over both decode paths; `INGEST_FIRST=1` self-bootstraps the stores).
-- `results/` — cross-machine benchmark reports (see `results/*.md`).
+- `results/` — cross-machine benchmark reports (`results/*.md`), an interactive
+  self-contained HTML explorer (`results/*-explorer.html`), and the generator
+  `results/make_explorer.py` (turns a dir of per-machine run dirs into one).

cmd/stellar-rpc/scripts/bench-fullhistory/results/2026-06-03-cross-machine.md

@@ -5,7 +5,11 @@ Cross-machine summary of `cmd/stellar-rpc/scripts/bench-fullhistory` runs from
 per-iter and per-sweep CSVs live at
 `gs://rpc-full-history/benchmarks/2026-06-03/<machine-dir>/`; every number here
 is recomputed from those CSVs. Raw per-iter data is on GCS, so this report keeps
-only the aggregated tables.
+only the aggregated tables. For an **interactive view** (toggle tier, decode
+path, percentiles, machines, concurrency; overlaid line chart for cross-machine
+comparison + bar chart + sortable table) open
+[`2026-06-03-explorer.html`](./2026-06-03-explorer.html) — a self-contained,
+offline file with all data embedded (regenerate via `make_explorer.py`).
 
 All four machines were re-run on the fixed harness, which addresses the PR #750
 review: