ScaffoldBench

RunSonnet 5.5repeat 2

Orchestrator-workers on Rebook onto the earliest available flight and settle the difference

claude-sonnet-5-5__orchestrator-workers__airline-medium__r2

Score

1.000

passed
design ran as designed
Cost
$0.055
Latency
24.1s
setup 0.0s, grading 0.0s
Turns
11
Tool calls
7
Subagents
1
Tokens in / out
37k / 1.9k
27,102 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.
  • The task was decomposed: heldorchestrator-workers.decompositionThe orchestrator’s reply was read through its schema and held at least one subtask.
  • The fan-out happened: heldorchestrator-workers.fan-outEvery subtask got a worker, every worker ran, and the recorded subagent count matches.
  • The synthesis wrote the answer: heldorchestrator-workers.synthesisThe final answer came from the synthesis call, not from a worker’s report standing in for it.
  • Each call was shown the right tools: heldorchestrator-workers.toolsThe orchestrator and the synthesis were shown no tools and every worker the task’s full toolset.

Phases

Decompose: 0.0s to 5.9s, finished, 0 tool calls, $0.014
Worker 1: all: 5.9s to 19.2s, finished, 7 tool calls, $0.029
Synthesize: 19.2s to 24.1s, finished, 0 tool calls, $0.012

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

PhaseDurationCostTurnsToolsModel
Decompose5.9s$0.01420claude-sonnet-5-5
Worker 1: all13.3s$0.02987claude-sonnet-5-5
Synthesize4.9s$0.01210claude-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 orchestrator-workers --tasks airline-medium --models claude-sonnet-5-5 --repeats 1

Score breakdown

  • passReservation moved to the earliest flight with an available seat, same cabin
    1.00weight 0.43

    booked on AA204, expected AA204

  • passExactly one fee, the fare difference for the flight actually booked
    1.00weight 0.43

    fare difference owed on the booked flight: 8000 cents

  • passExactly one economy seat moved from the old flight to the new one
    1.00weight 0.14

Penalties

No penalties were applied.

Scorer metadata

tool calls
7
failed tool calls
0
distinct tools
7

Final answer

Dana Whitfield's reservation R2087 has been moved from AA210 (17:40) to AA204, departing 09:05 on 2026-08-03. She stays in economy. This is based on the worker's report; I had no tools to check it myself.

  • Why AA204: It is the earliest flight that day with an economy seat. UA88 (06:15) has no economy seats left, and AA918 (07:30) is cancelled. AA204 had 4 economy seats left.
  • Change fee: The economy change fee is $75. Dana is gold-tier, which is exempt, so no fee was charged.
  • Fare difference: The new fare is $390 against the $310 she paid, so she owes $80.00. No tier is exempt from this. It was recorded as a single fare_difference ledger entry (L1).

The report mentions no other changes and no open items.

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.