ScaffoldBench

RunOpus 5.5repeat 2

Single on Consolidate duplicate CRM reports

claude-opus-5-5__single__support-medium__r2

Score

1.000

passed
design ran as designed
Cost
$0.103
Latency
25.1s
setup 0.0s, grading 0.0s
Turns
16
Tool calls
15
Subagents
0
Tokens in / out
50k / 2.4k
42,886 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.
  • Each call was shown the right tools: heldsingle.toolsThe agent was shown exactly the task’s tools, none missing and none extra.

Phases

Main: 0.0s to 25.1s, finished, 15 tool calls, $0.103

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

PhaseDurationCostTurnsToolsModel
Main25.1s$0.1031615claude-opus-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 single --tasks support-medium --models claude-opus-5-5 --repeats 1

Score breakdown

  • passSurvivor carries the earliest SLA deadline of the merged set
    1.00weight 0.23

    sla_due_at=2026-03-17T12:40:00Z (expected 2026-03-17T12:40:00Z)

  • passBoth duplicate reports were merged into the earliest-created ticket
    1.00weight 0.23

    T-1053: status=merged, merged_into=T-1052; T-1054: status=merged, merged_into=T-1052

  • passThe engineering tracking ticket now points at the survivor
    1.00weight 0.15

    T-1061 links=[T-1052]

  • passSurvivor re-graded from the matrix for its widened impact scope
    1.00weight 0.15

    severity=sev2 (matrix says sev2)

  • passSurvivor assigned to the engineer on call for the owning rotation at its deadline
    1.00weight 0.15

    assignee=E-201 (expected E-201)

  • passEvery ticket folded in was read before the merge
    1.00weight 0.08

Penalties

No penalties were applied.

Scorer metadata

tool calls
15
failed tool calls
0
distinct tools
10

Final answer

The three reports of the CRM account search outage are now one ticket, T-1052. It's graded sev2 and assigned to the right engineer.

Merged: T-1053 and T-1054 into T-1052. All three describe the same problem: every account search returns an HTTP 500 error, starting around 07:00, all reported by Revenue Ops.

  • Surviving ticket: T-1052, because it was reported first (07:05).
  • Deadline: 12:40 today (the service desk clock reads 2026-03-17). That was already promised to the requester on T-1053 and was the earliest of the three, so T-1052's own 15:05 deadline was brought forward to match.
  • Severity: sev2. T-1053 reported that the whole department is blocked, so the merged ticket took that wider impact. The escalation matrix rates CRM with department-wide impact as sev2.
  • Assignee: E-201 Ines Kovac. Business-apps is the team that owns CRM, and she is their on-call engineer at the 12:40 deadline (her shift runs 08:00–20:00). T-1053 had been assigned to E-202 Tomas Lindqvist, but he isn't the one on call at the deadline, so he isn't on the merged ticket.
  • I left a comment on T-1052 explaining all of this.

Left alone:

  • T-1055 (CSV export gives an empty file) is a different problem. The requester says search works fine for them.
  • T-1061 (search index rebuild) is engineering's internal tracking ticket, not a customer report, so it wasn't merged. It's still linked to T-1052; the link moved over from T-1054 in the merge.

I didn't escalate the ticket. The task asked for it to be graded and assigned, and the rules say the team that owns CRM works it. Under the matrix, a sev2 escalation would go to the platform-primary team. Someone should make that call if the 12:40 deadline is at risk.

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.