ScaffoldBench

RunSonnet 5.5repeat 1

Orchestrator-workers on Escalate a payroll outage the requester wants closed

claude-sonnet-5-5__orchestrator-workers__support-hard__r1

Score

1.000

passed
design ran as designed
Cost
$0.063
Latency
28.5s
setup 0.0s, grading 0.0s
Turns
12
Tool calls
8
Subagents
1
Tokens in / out
36k / 3.2k
27,249 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 6.8s, finished, 0 tool calls, $0.011
Worker 1: all: 6.8s to 21.3s, finished, 8 tool calls, $0.036
Synthesize: 21.3s to 28.5s, finished, 0 tool calls, $0.016

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

PhaseDurationCostTurnsToolsModel
Decompose6.8s$0.01120claude-sonnet-5-5
Worker 1: all14.5s$0.03698claude-sonnet-5-5
Synthesize7.2s$0.01610claude-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 support-hard --models claude-sonnet-5-5 --repeats 1

Score breakdown

  • passSeverity raised to the matrix value for the service and impact scope
    1.00weight 0.21

    severity=sev1 (matrix says sev1)

  • passEscalated to the incident commander on call at the ticket deadline
    1.00weight 0.29

    escalated_to=E-402 (expected E-402, on call at 2026-03-17T19:40:00Z)

  • passTicket left in status escalated rather than closed
    1.00weight 0.14

    status=escalated

  • passThe committed SLA deadline survived the escalation
    1.00weight 0.14

    sla_due_at=2026-03-17T19:40:00Z

  • passA comment records that the open linked ticket blocks closure
    1.00weight 0.14
  • passEscalation matrix read before the severity was raised
    1.00weight 0.07

Penalties

No penalties were applied.

Scorer metadata

tool calls
8
failed tool calls
0
distinct tools
7

Final answer

T-1070 is re-graded and escalated, but it is not closed. I could not close it today as Grace asked.

  • Severity: I raised it from sev3 to sev1. The escalation matrix gives payroll with company-wide impact as sev1.
  • Escalation: The matrix routes sev1 to incident-command. The ticket is escalated to Ola Berg (E-402), who is on call for that rotation when the ticket deadline falls (2026-03-17 19:40Z, within her 18:00–06:00Z shift). The ticket status is now escalated.
  • Deadline: The existing deadline of 2026-03-17 19:40Z had already been committed to the requester, so it is unchanged. Re-grading and escalating do not move it earlier.
  • Not closed: The linked ticket T-1071 is still open, and policy bars closing a ticket while a linked ticket is open. T-1071 is the root cause: a payroll reporting replica lag, assigned to Dana Whitfield (E-101). A comment from E-201 says the March batch cannot be rerun until T-1071 is fixed.
  • Not merged: T-1070 and T-1071 are different problems, so I did not merge them and made no changes to T-1071.
  • Handover: I left a comment on T-1070 recording the grading, the escalation, the unchanged deadline and the blocker. It says to close T-1070 once T-1071 is resolved and the March batch has been rerun and posted.

For Grace and the next shift:

  • Grace should be told that same-day closure isn't possible until T-1071 is fixed and the batch is rerun.
  • T-1071 has its own deadline of 10:30Z, and it is still open. It is the critical path for the payroll run, so someone should chase Dana Whitfield on it.
  • Ola Berg's shift starts at 18:00Z, so the ticket has no incident-command owner on shift before then.

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.