RunSonnet 5.5repeat 2
Orchestrator-workers on Run the major-incident runbook for a CI outage
claude-sonnet-5-5__orchestrator-workers__support-major-incident__r2
Score
0.820
- Cost
- $0.141
- Latency
- 52.4s
- setup 0.0s, grading 0.0s
- Turns
- 42
- Tool calls
- 38
- Subagents
- 1
- Tokens in / out
- 115k / 6.7k
- 98,596 read from cache
Did the design run?
- 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
Drawn from the phase records: timing and cost per call, no tool calls or messages.
| Phase | Duration | Cost | Turns | Tools | Model |
|---|---|---|---|---|---|
| Decompose | 8.4s | $0.012 | 2 | 0 | claude-sonnet-5-5 |
| Worker 1: all | 32.2s | $0.108 | 39 | 38 | claude-sonnet-5-5 |
| Synthesize | 11.8s | $0.022 | 1 | 0 | claude-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
- A gate that blocked the right moveGated-execute · Haiku 4.5
- Three proposals beat one agentPropose-select · Haiku 4.5
- A verifier that earned its keepPlan-execute-verify · Sonnet 5.5
- A revision that did not helpPlan-execute-verify · Haiku 4.5
- The control, missing a stepSingle · Haiku 4.5
Record this cell yourself: npm run bench -- run --strategies orchestrator-workers --tasks support-major-incident --models claude-sonnet-5-5 --repeats 1
Score breakdown
- passAn incident is open for the build pipeline1.00weight 0.04
- passThe incident-command engineer on call now commands the incident1.00weight 0.08
commander=E-401 (expected E-401)
- passThe commander was paged1.00weight 0.04
- passT-3101 linked to the incident1.00weight 0.04
- passT-3101 kept open, raised to sev1 with its deadline intact1.00weight 0.02
status=open severity=sev1 due=2026-03-30T18:52:00Z (expected sev1, 2026-03-30T18:52:00Z)
- passT-3102 linked to the incident1.00weight 0.04
- passT-3102 kept open, raised to sev1 with its deadline intact1.00weight 0.02
status=open severity=sev1 due=2026-03-30T18:58:00Z (expected sev1, 2026-03-30T18:58:00Z)
- passT-3103 linked to the incident1.00weight 0.04
- passT-3103 kept open, raised to sev1 with its deadline intact1.00weight 0.02
status=open severity=sev1 due=2026-03-30T19:01:00Z (expected sev1, 2026-03-30T19:01:00Z)
- passT-3104 linked to the incident1.00weight 0.04
- passT-3104 kept open, raised to sev1 with its deadline intact1.00weight 0.02
status=open severity=sev1 due=2026-03-31T15:05:00Z (expected sev1, 2026-03-31T15:05:00Z)
- passT-3105 linked to the incident1.00weight 0.04
- passT-3105 kept open, raised to sev1 with its deadline intact1.00weight 0.02
status=open severity=sev1 due=2026-04-01T15:07:00Z (expected sev1, 2026-04-01T15:07:00Z)
- passT-3106 linked to the incident1.00weight 0.04
- passT-3106 kept open, raised to sev1 with its deadline intact1.00weight 0.02
status=new severity=sev1 due=2026-03-30T16:20:00Z (expected sev1, 2026-03-30T16:20:00Z)
- passT-3109 linked to the incident1.00weight 0.04
- passT-3109 kept open, raised to sev1 with its deadline intact1.00weight 0.02
status=new severity=sev1 due=2026-03-30T16:20:00Z (expected sev1, 2026-03-30T16:20:00Z)
- failT-3112 linked to the incident0.00weight 0.04
- failT-3112 kept open, raised to sev1 with its deadline intact0.00weight 0.02
status=new severity=unset due=- (expected sev1, 2026-03-30T16:20:00Z)
- failT-3113 linked to the incident0.00weight 0.04
- failT-3113 kept open, raised to sev1 with its deadline intact0.00weight 0.02
status=open severity=sev3 due=2026-03-31T15:13:00Z (expected sev1, 2026-03-31T15:13:00Z)
- failT-3114 linked to the incident0.00weight 0.04
- failT-3114 kept open, raised to sev1 with its deadline intact0.00weight 0.02
status=open severity=sev3 due=2026-03-31T15:16:00Z (expected sev1, 2026-03-31T15:16:00Z)
- passThe engineer covering comms right now is the communications lead1.00weight 0.08
comms=E-803 (expected E-803)
- passA status update in the runbook form, with the next update due within 30 minutes1.00weight 0.08
- passA postmortem ticket for the service, linked to the incident, due at close of business five business days on1.00weight 0.08
T-3115 due=2026-04-08T17:00:00Z
Penalties
No penalties were applied.
Scorer metadata
- tool calls
- 38
- failed tool calls
- 0
- distinct tools
- 18
Final answer
The major incident is open and the runbook steps the worker reported are done, but the incident still needs follow-up. This summary rests on a single worker report that I could not independently verify.
Done, per the worker's report
- Incident: INC-1 was opened on build-pipeline at 15:20 UTC. The worker followed KB-40 and ignored KB-39 because it is superseded.
- Commander: Priya Raman (E-401), on call for incident-command at that moment, was paged once (PG-1) and made commander. No one else was paged.
- Linked reports: T-3101 to T-3106 and T-3109 were linked. All describe runners not picking up jobs.
- Left out: T-3107 (flaky test) and T-3108 (local Xcode link failure) were not touched.
- Severity: All seven linked reports were set to sev1. The report doesn't say the escalation matrix or any knowledge-base amendment was checked for this, so that is unconfirmed.
- Deadlines: Existing committed deadlines were left unchanged. T-3106 and T-3109 had none, so they got 16:20 UTC (triage time plus the 60-minute sev1 target).
- Comms lead: Mia Laurent (E-801) is on sick leave, so the comms-secondary backup rotation covered. Elif Demir (E-803) is comms lead.
- Status update: One update was published at 15:20 UTC, giving the start as 14:52 UTC (the earliest report's creation time) and the next update at 15:45 UTC.
- Postmortem: T-3115 was created and linked to INC-1, due 2026-04-08 17:00 UTC. The worker says this is the 5th business day after opening, with Good Friday (Apr 3) and Easter Monday (Apr 6) excluded.
Open or unclear
- Status updates: The next update is due at 15:45 UTC, with hourly updates after that while the incident continues. No one has been assigned to post them.
- Assignments: T-3101 is still assigned to E-101 and the other reports are unassigned. The worker left them for the commander to direct, so they have not been reassigned or escalated.
- T-3115 severity: The runbook doesn't call for one, so it is unset.
- Dates: The report never states the incident's calendar date. The April postmortem date also doesn't match today's date in my environment (2026-10-13). I can't tell whether that is a real problem or just a difference in the test setup.
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.