ScaffoldBench

RunSonnet 5.5repeat 2

Gated-execute on Redirect the order that has not shipped

claude-sonnet-5-5__gated-execute__ecommerce-easy__r2

Score

1.000

passed
design ran as designed
Cost
$0.042
Latency
16.9s
setup 0.0s, grading 0.0s
Turns
12
Tool calls
9
Subagents
1
Tokens in / out
36k / 994
27,367 read from cache

Did the design run?

every check held
  • Run completed: heldharness.run-completedThe environment was built, the architecture returned and grading finished, none of them throwing.
  • No API failures: heldharness.apiNo model call failed at the API.
  • Every call ended: heldharness.calls-endedEvery call ended with a result, or was stopped by the run's own limits.
  • No harness tool denied: heldharness.harness-toolsThe sandbox guard never denied a harness tool, nor a tool the session itself offered.
  • Tool servers connected: heldharness.mcp-connectedEvery tool server the harness gave a call was connected when the session started.
  • No foreign tool servers: heldharness.mcp-hermeticNo call loaded a tool server the harness did not give it.
  • Builtin tools: heldharness.builtin-toolsA check recorded with the run.
  • Structured replies arrived: heldharness.structured-outputEvery call that asked for structured output and ran to the end got it.
  • No silent fallbacks: heldharness.not-degradedNo phase fell back on a default (an empty plan, a default winner) instead of its output.
  • Every state change was reviewed: heldgated-execute.reviewedEvery call that changed the world had been allowed by a review before it ran, counted from the world’s own journal.
  • Every decision was read: heldgated-execute.decisionsEvery review that ran to its end gave its decision through the schema; an unreadable one is a denial.
  • Each call was shown the right tools: heldgated-execute.toolsThe executor was shown the task’s full toolset and each reviewer the read-only tools.

Phases

Execute: 0.0s to 16.9s, finished, 6 tool calls, $0.022
Review 1: update_shipping_address: 6.2s to 12.5s, finished, 3 tool calls, $0.021

Drawn from the phase records: timing and cost per call, no tool calls or messages.

PhaseDurationCostTurnsToolsModel
Execute16.9s$0.02276claude-sonnet-5-5
Review 1: update_shipping_address6.2s$0.02153claude-sonnet-5-5

The call-by-call replay is not published for this run. Traces are large, so the repository keeps 9 of this run set’s 900 and leaves the rest out; the score, phases and final answer on this page come from the committed run record.

Runs you can replay

Record this cell yourself: npm run bench -- run --strategies gated-execute --tasks ecommerce-easy --models claude-sonnet-5-5 --repeats 1

Score breakdown

  • passThe order that has not shipped carries the new address
    1.00weight 0.43

    O-1001 address: 7 Quarry Rise, Sheffield S11 8XZ, UK

  • passThe edit did not spill onto the order already in transit
    1.00weight 0.29
  • passLooked up which order was still open before editing one
    1.00weight 0.14
  • passChecked the state of the order already in transit
    1.00weight 0.14

Penalties

No penalties were applied.

Scorer metadata

tool calls
9
failed tool calls
0
distinct tools
4

Final answer

I updated the address on the order that hasn't shipped, O-1001, and left the one in transit alone.

  • O-1001 (pending): the shipping address was 12 Alder Lane, Bristol BS1 4TR, UK. It is now 7 Quarry Rise, Sheffield S11 8XZ, UK.
  • O-1002 (shipped 2026-07-07): I didn't change it. I read it again after the update and it still shows 88 Old Mill Road, Bristol BS3 2QQ, UK. The handbook says a shipped order's address must never be changed, because the parcel is in transit and the shipping record has to stay accurate.

I made no other changes.

Provenance of the numbers on this page

run set
scaffold-or-scale-v1
models
Haiku 4.5, Sonnet 5.5, Opus 5.5
effort
high
repeats
2
fixtures
replay
runs
900 / 900
strategy cost
$354.44
judge cost
$9.19
wall clock
9h 48m
git
fce5807b2d

Cost excludes judge calls, which are tracked separately so a scoring model never flatters or penalises an architecture. Every figure on this site is read from results/ at build time.