ScaffoldBench

RunSonnet 5.5repeat 2

Gated-execute on Escalate a payroll outage the requester wants closed

claude-sonnet-5-5__gated-execute__support-hard__r2

Score

1.000

passed
design ran as designed
Cost
$0.091
Latency
38.5s
setup 0.0s, grading 0.0s
Turns
28
Tool calls
21
Subagents
3
Tokens in / out
74k / 3.3k
57,971 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 38.5s, finished, 8 tool calls, $0.034
Review 1: set_severity: 6.2s to 12.5s, finished, 2 tool calls, $0.013
Review 2: escalate_ticket: 14.2s to 22.6s, finished, 5 tool calls, $0.019
Review 3: post_comment: 24.9s to 34.5s, finished, 6 tool calls, $0.025

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

PhaseDurationCostTurnsToolsModel
Execute38.5s$0.03498claude-sonnet-5-5
Review 1: set_severity6.3s$0.01342claude-sonnet-5-5
Review 2: escalate_ticket8.4s$0.01975claude-sonnet-5-5
Review 3: post_comment9.6s$0.02586claude-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 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
21
failed tool calls
0
distinct tools
8

Final answer

I did not close T-1070. T-1071 (the replica lag that caused the failure) is still open, and the rules don't allow closing a ticket while a linked ticket is open. The payroll batch also can't be rerun until T-1071 is fixed.

  • Severity: The matrix gives sev1 for payroll with company-wide impact, so I raised T-1070 from sev3 to sev1.
  • Escalation: I escalated it to incident-command, the rotation the matrix names for sev1. I picked Ola Berg (E-402) because she is the engineer on call at 19:40Z, when the ticket's deadline falls. The ticket is now in status escalated.
  • Deadline: It stays at 2026-03-17T19:40Z. A committed deadline can't be moved earlier, and it already had one.
  • Handover: I left a comment on T-1070 naming T-1071 as the blocker. It tells the next shift to fix T-1071, rerun the batch, confirm payments post before Friday's payday, and then close T-1070.
  • Scope: I didn't touch T-1071 or merge the two tickets, since one is the cause and the other is the effect, not the same report.

Grace should be told the ticket stays open until payroll actually posts.

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.