ScaffoldBench

RunSonnet 5.5repeat 2

Plan-execute-verify on Triage a new service desk ticket

claude-sonnet-5-5__plan-execute-verify__support-easy__r2

Score

1.000

passed
design ran as designed
Cost
$0.062
Latency
23.0s
setup 0.0s, grading 0.0s
Turns
17
Tool calls
12
Subagents
0
Tokens in / out
48k / 1.9k
36,002 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 plan was a plan: heldplan-execute-verify.planThe planner’s reply was read through its schema and held at least one step.
  • Every verdict was read: heldplan-execute-verify.verdictsEvery review round ended with a verdict read from the schema reply, not a default.
  • The loop did what its verdicts said: heldplan-execute-verify.loopEvery verdict that called for a revision got one, and nothing was revised without such a verdict.
  • Each call was shown the right tools: heldplan-execute-verify.toolsThe planner was shown no tools, the verifier the read-only ones, and the executor and reviser the full toolset.

Phases

Plan: 0.0s to 6.5s, finished, 0 tool calls, $0.0086
Execute: 6.5s to 17.1s, finished, 8 tool calls, $0.029
Verify: 17.1s to 23.0s, finished, 4 tool calls, $0.025

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

PhaseDurationCostTurnsToolsModel
Plan6.5s$0.008620claude-sonnet-5-5
Execute10.5s$0.02998claude-sonnet-5-5
Verify6.0s$0.02564claude-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 plan-execute-verify --tasks support-easy --models claude-sonnet-5-5 --repeats 1

Score breakdown

  • passSeverity matches the escalation matrix for the service and impact scope
    1.00weight 0.27

    severity=sev3 (matrix says sev3)

  • passSLA deadline is the triage time plus the response target for that severity
    1.00weight 0.27

    sla_due_at=2026-03-17T17:20:00Z (expected 2026-03-17T17:20:00Z)

  • passAssigned to the engineer on call for the owning rotation at the deadline
    1.00weight 0.27

    assignee=E-301 (expected E-301)

  • passTicket left in status open
    1.00weight 0.09
  • passEscalation matrix read before the severity was set
    1.00weight 0.09

Penalties

No penalties were applied.

Scorer metadata

tool calls
12
failed tool calls
0
distinct tools
8

Final answer

I triaged T-1041. I didn't re-read the ticket afterwards, but each update was confirmed in its tool response. I changed no other ticket.

  • Severity: sev3. The ticket is VPN with team-level impact (the Berlin sales team, about nine people), and the matrix gives sev3 for that.
  • Deadline: 2026-03-17T17:20:00Z. The ticket had no deadline. Triage time was 09:20Z and the sev3 response target is 480 minutes.
  • Assignee: E-301 Yuki Sato. VPN is owned by the network-primary rotation, and she is the engineer on call when the deadline falls (shift 08:00–20:00Z). She is a member of that rotation. The ticket is now open.

I didn't escalate it. The matrix gives sev3 no escalation rotation, and the task only asked for triage.

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.