Claude Haiku 5.5 (effort default (no --effort flag) · thinking on) on Find support tickets that breached SLA, measured in business hours across time zones, holidays and customer-wait pauses

AClaude Haiku 5.5default · thinking on · Claude Code 2.1.299 (headless -p) · cloud · 2026-10-21 · by @bobsiopener

Trust 80%Fully autonomoussuccess · first tryCompare

Estimated cost: $0.0092
Wall time: 32s
Human step-ins: 0
Outcome: Pass

Compare with other models on this task →

Exact prompt

# Which support tickets breached their SLA last month? Inputs: `tickets.csv` (ticket, customer, priority), `customers.csv` (plan), `events.csv` (every status change of every ticket, timestamps with their own UTC offsets) and `sla.json` (targets, support hours, holidays). For each ticket, measure time to resolution in business hours and compare it with its target. - The clock runs only during support hours on working days, in the support team's time zone given in `sla.json`. Holidays are not working days. - The clock is paused while a ticket is waiting on the customer. - The target depends on priority and on the customer's plan as `sla.json` sets out. - A ticket breaches if its business time is strictly greater than its target. Write `out/sla.json`: `{"breached": [<ticket ids, sorted>], "breach_count": <n>, "median_hours_by_priority": {"P1": <h>, "P2": <h>, "P3": <h>}, "worst_overrun": "<ticket id>"}`. Hours are business hours rounded to 2 decimals, half-up. `worst_overrun` is the ticket with the largest business time minus target. ## How this run works - Work on your own: nobody will answer questions in this task, and writing to `ask/` counts against the run. - Write only under `out/`. Create `out/` if it is missing. Inputs are synthetic; don't fetch anything from the network. - Scoring is partial credit: each listed check is pass/fail, and the score is the share passed. Numbers are compared to within 0.005 unless the task says otherwise.

Measure 36-48 tickets against priority- and plan-based targets in business hours. Status events arrive shuffled with mixed UTC offsets, the clock pauses while waiting on the customer, two holidays fall in the month, and premium customers get half the target. Wall-clock, holiday-blind and pause-blind readings all give a different breach list. Original benchmark task written for usecaserank.com (not based on a specific public post). Headless, offline run on a reproducible fixture. Each run's notes say how its cost was measured (estimates are marked). Folder-restore note: this run's output files under work/out/ were lost in a 1:22 AM ET restore on Oct 21, 2026. The trace, step-in exchanges and output-hash list were checked against the recorded SHA-256 values and match; the hashes of the individual output files inside that list can no longer be recomputed. Measured data unchanged.

Success criteria: breached_set breach_count medians worst_overrun

Setup notes: Batch 19 case u02-sla-breach-business-hours (offline, synthetic inputs, seed 1357615286). Autonomous case: no stand-in user and no step-ins expected; any file the solver wrote to ask/ is counted as a measured intervention (0) and blocks a pass. Prompt = TASK.md verbatim. Hidden key kept outside the solver's work dir. Scored with the case's scorer (python3 score.py WORKDIR KEYDIR, absolute paths); pass = every check true and no ask/ files. Tools: Read/Write/Edit/Glob/Grep/Bash, web disabled. Single run, no retries. 1-min load average at scoring 0.37 (8 cores). Time-zone results depend on the system time-zone database: tzdata 2026c-0+deb13u1 (Python zoneinfo). Claude Code's own reported cost for claude-haiku-5.5: $0.009225. Score: 4/4. Cost is an ESTIMATE: measured tokens x pricing.json list price (Claude Pro subscription, nothing billed); input_tokens include cache reads and writes. Single run.

Estimated cost List-price estimate; not actually billed.

Metrics

Estimated cost: $0.0092
Wall time: 32s
Human step-ins: 0
Turns: 5

Model usage
Model	Input tok	Output tok	Cost
AClaude Haiku 5.5anthropic/claude-haiku-5.5effort default (no --effort flag) · thinking on	135,474	6,399	$0.0092

Reproduce this run

Use the use case prompt on the use case page.
Same setup: Claude Haiku 5.5 (anthropic/claude-haiku-5.5, effort default (no --effort flag) · thinking on) via Claude Code 2.1.299 (headless -p), cloud.
It counts as a success only if it meets every success criterion: breached_set breach_count medians worst_overrun
Log model, tokens, wall time, human step-ins and the evidence log, then submit it as a reproduction. A different model? Submit a head-to-head rerun.

How runs are scored: Methodology.

Was this useful?

⇄ Rerun on another model

Automated checks

Trust 80% comes from these checks (how Trust % works).

Run date is not in the future : passed
Model is in the dated pricing table : passed
Cost recomputed from tokens matches : passed
Cached tokens are consistent : passed
Output throughput is plausible : passed
Autonomy matches step-ins : passed
Step-ins do not exceed turns : passed
Retries agree with first-try result : passed
Evidence log SHA-256 matches : passed
Evidence log is substantial : passed

Evidence log

[
 {
  "duration_api_ms": 31217,
  "stop_reason": "end_turn",
  "session_id": "[REDACTED]",
  "total_cost_usd": 0.00922458,
  "usage": {
   "input_tokens": 10,
   "cache_creation_input_tokens": 24576,
   "cache_read_input_tokens": 110888,
   "output_tokens": 6399,
   "output_tokens_details": {
    "thinking_tokens": 3432
   },
   "server_tool_use": {
    "web_search_requests": 0,
    "web_fetch_requests": 0
   },
   "service_tier": "standard",
   "cache_creation": {
    "ephemeral_1h_input_tokens": 24576,
    "ephemeral_5m_input_tokens": 0
   },
   "inference_geo": "global",
   "iterations": [
    {
     "input_tokens": 2,
     "output_tokens": 1751,
     "cache_read_input_tokens": 32225,
     "cache_creation_input_tokens": 4890,
     "cache_creation": {
      "ephemeral_5m_input_tokens": 0,
      "ephemeral_1h_input_tokens": 4890
     },
     "type": "message"
    }
   ],
   "speed": "standard",
   "fallback_credit": null
  },
  "modelUsage": {
   "claude-haiku-5-5": {
    "inputTokens": 10,
    "outputTokens": 6399,
    "cacheReadInputTokens": 110888,
    "cacheCreationInputTokens": 24576,
    "webSearchRequests": 0,
    "costUSD": 0.00922458,
    "contextWindow": 1000000,
    "maxOutputTokens": 128000,
    "thinkingTokens": 3432,
    "canonicalModel": "claude-haiku-5-5",
    "provider": "firstParty",
    "costBasis": "list"
   }
  },
  "permission_denials": [],
  "terminal_reason": "completed",
  "fast_mode_state": "off",
  "fast_mode_disabled_reason": "sdk_opt_in_required",
  "subagent_stats": {
   "spawned": 0,
   "requested": {
    "background": 0,
    "foreground": 0,
    "unset": 0
   },
   "started_in_background": 0,
   "max_depth": 0,
   "spawned_by_subagents": 0,
   "completed": 0,
   "failed": 0,
   "killed": {
    "parent": 0,
    "user": 0,
    "system": 0
   },
   "refused": {
    "depth_limit": 0,
    "concurrency_limit": 0,
    "budget": 0
   },
   "by_type": {}
  },
  "safety_stops": 0,
  "is_error": false,
  "num_turns": 5,
  "subtype": "success",
  "api_error_status": null,
  "result": "I wrote `out/sla.json`:\n\n```json\n{\"breached\": [\"SUP-4113\", \"SUP-4117\", \"SUP-4119\", \"SUP-4122\", \"SUP-4150\", \"SUP-4196\", \"SUP-4204\", \"SUP-4205\", \"SUP-4226\", \"SUP-4234\"], \"breach_count\": 10, \"median_hours_by_priority\": {\"P1\": 0.92, \"P2\": 2.95, \"P3\": 8.0}, \"worst_overrun\": \"SUP-4204\"}\n```\n\nI checked the event data before computing: every ticket has exactly one `new` and one `resolved` event, and all the ticket activity falls in September 2026, so no ticket needed filtering.\n\nInterpretations I made:\n- **Clock:** starts at the `new` event and stops at the `resolved` event, both converted to UTC. Business time counts 09:00\u201317:00 at UTC+01:00 on weekdays, excluding the two listed holidays.\n- **Pauses:** each `waiting_on_customer` period runs until the next status event, and its business time is subtracted.\n- **Targets:** premium customers get half the listed target (P1 2h, P2 4h, P3 12h). Standard customers get the listed target.\n- **Medians:** taken over each priority's actual business hours, not the plan-adjusted targets, rounded half-up to 2 decimals.\n- **Boundary case:** SUP-4151 used exactly 8.00 business hours against an 8h target, so it is not a breach because the comparison is strict.\n\nThe script ran inline and nothing was written outside `out/`.",
  "ttft_ms": 1548,
  "type": "result",
  "duration_ms": 31644,
  "uuid": "[REDACTED]",
  "ttft_stream_ms": 1357,
  "time_to_request_ms": 244,
  "first_content_frame_ms": 1358,
  "queued_turn_count": 0,
  "result_index": 0
 }
]