feat(scaling): bench run and load test detail pages

[0x00101010]

Summary

Why

A base-bench run is split across two old pages (/benchmark/run/<id> and /benchmark/run-comparison/<id>, a wall of ungrouped charts), and a load test leads with a workload-specific swaps/s number. These pages put each run on one page in the scaling style, with Mgas/s as the unit.

What

• /scaling/runs/bench/<id>: one row per configuration of the benchmark run (an id can cover several), thresholds as a table (metric, value, warning, error, status), per-block charts grouped Build / State root / Persist / Sparse trie / Validate with only Build open and sequencer and validator as two series, raw config, JSON/CSV downloads. Every metric in metricDefinitions.ts is assigned to one group (tested).
• /scaling/runs/load/<network>/<timestamp>: observed Mgas/s headline with ≈ N tx/s at this mix (gas/s ÷ gas per confirmed tx), target / duration / senders / in-flight inline with the full config behind an info icon, throughput chart, block and Flashblocks latency percentile bars, gas stats, failure reasons, previous/next run.
• A small SVG line chart in app/scaling/runs/components/ (no new dependency); nothing imported from app/benchmark/components or views. downloadUtils.ts moves to app/scaling/runs/lib/ unchanged.
• Thresholds keyed with a role prefix (sequencer/latency/get_payload, how report-api writes them) now count toward run status; before, they were skipped, so a breached threshold still read success.
• Both routes join the CI check that internal-only routes 404 on the public build.

Notes

• Stack: feat(scaling): snapshot schema v2 with ladder, ceilings and guardrails #194 → feat(scaling): first screen with ladder, ceilings, guardrails and progress #196 → feat(scaling): detail pages and unified run index #199 → this PR.
• The translation reads tx/s, not swaps/s: the load tester's mix can include transfers, and the mix is printed beside it.
• Internal-only through the existing scaling surface; the public build returns 404 for both routes.

Screenshots

Internal dev build with report-api responses replaced by the committed synthetic fixture (invented numbers).

/scaling/runs/bench/test-synthetic-1 at 1280 px and 360 px:

/scaling/runs/load/sepolia/<timestamp> at 1280 px and 360 px, and a run that stopped with a fatal error:

Test plan

• npm test, npm run typecheck, npm run lint, npm run build, npm run build:internal, npm run llms -- --check, npm run agents -- --check
• BenchRun.test.tsx: configurations, thresholds (role-prefixed one marks its config error), stage order with only Build open, alias-resolved metric, metrics fetch error, every metric in one stage.
• LoadRun.test.tsx: headline, translation, knobs, chart, latency bars, previous/next, fatal run.
• Rendered the newest live bench run and Sepolia load test plus the fixtures in dev:internal at 360 px and 1280 px; public build returns 404 for both routes.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
omni-ui	Ready	Preview	Oct 16, 2026 1:53pm UTC

[cb-heimdall]

✅ Heimdall Review Status

Requirement	Status	More Info
Reviews	✅ 1/1	Denominator calculation Show calculation 1 if user is bot 0 1 if user is external 0 2 if repo is sensitive 0 From .codeflow.yml 1 Additional review requirements Show calculation Max 0 0 From CODEOWNERS 0 Global minimum 0 Max 1 1 1 if commit is unverified 0 Sum 1