{
  "schema_version": "preflight/v2",
  "base_url": "http://127.0.0.1:30002/v1",
  "model": "gpt-oss-puzzle-88b",
  "thinking": {
    "mode": "default",
    "chat_template_kwargs": null,
    "reasoning_effort": "low"
  },
  "budget": {
    "visible_answer_tokens": 512,
    "reasoning_headroom_tokens": 4096,
    "max_completion_tokens": 4608
  },
  "checks": [
    "smoke",
    "json",
    "needle",
    "tools"
  ],
  "results": [
    {
      "name": "smoke (short coding)",
      "passed": true,
      "detail": "2.5s got='```python\\nsum(xs)\\n```' finish='stop' visible=21 reasoning_chars=172 reasoning_tokens=None"
    },
    {
      "name": "structured JSON",
      "passed": true,
      "detail": "parsed keys=['language', 'ok'] finish='stop' visible=31 reasoning_chars=20 reasoning_tokens=None"
    },
    {
      "name": "needle @ ~120000 ctx",
      "passed": true,
      "detail": "18.0s ctx~120000 got='ZEBRA-42917-QUARTZ' finish='stop' visible=18 reasoning_chars=213 reasoning_tokens=None"
    },
    {
      "name": "shared-prefix tool batch x20",
      "passed": true,
      "detail": "20/20 clean (sample: 1.1s valid tool_call get_weather(city='Oakland') finish='tool_calls' visible=0 reasoning_chars=48 reasoning_tokens=None)"
    }
  ],
  "observations": [
    {
      "content": "```python\nsum(xs)\n```",
      "finish_reason": "stop",
      "content_chars": 21,
      "content_excerpt": "```python\nsum(xs)\n```",
      "reasoning_field": "reasoning",
      "reasoning_chars": 172,
      "reasoning_excerpt": "We need to output a Python one-liner that returns the sum of a list xs. Probably something like: `sum(xs)`. But as a one-liner function: `lambda xs: sum(xs)`. Provide code.",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 83,
        "total_tokens": 146,
        "completion_tokens": 63,
        "prompt_tokens_details": null
      },
      "test": "smoke",
      "seconds": 2.502,
      "passed": true,
      "validation_detail": "contains sum("
    },
    {
      "content": "{\"language\":\"python\",\"ok\":true}",
      "finish_reason": "stop",
      "content_chars": 31,
      "content_excerpt": "{\"language\":\"python\",\"ok\":true}",
      "reasoning_field": "reasoning",
      "reasoning_chars": 20,
      "reasoning_excerpt": "We just output JSON.",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 85,
        "total_tokens": 109,
        "completion_tokens": 24,
        "prompt_tokens_details": null
      },
      "test": "json",
      "seconds": 0.199,
      "passed": true,
      "validation_detail": "parsed keys=['language', 'ok']"
    },
    {
      "content": "ZEBRA-42917-QUARTZ",
      "finish_reason": "stop",
      "content_chars": 18,
      "content_excerpt": "ZEBRA-42917-QUARTZ",
      "reasoning_field": "reasoning",
      "reasoning_chars": 213,
      "reasoning_excerpt": "The user asks: \"What is the launch code? Reply with ONLY the code.\" The launch code is given in the prompt: \"IMPORTANT: The launch code is ZEBRA-42917-QUARTZ.\" So answer should be exactly that code, n",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 99100,
        "total_tokens": 99176,
        "completion_tokens": 76,
        "prompt_tokens_details": null
      },
      "test": "needle",
      "seconds": 18.019,
      "passed": true,
      "validation_detail": "needle present"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.146,
      "request_index": 1,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.219,
      "request_index": 5,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.222,
      "request_index": 0,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.219,
      "request_index": 7,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.219,
      "request_index": 6,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.221,
      "request_index": 4,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.221,
      "request_index": 2,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.221,
      "request_index": 3,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.554,
      "request_index": 8,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.615,
      "request_index": 12,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.614,
      "request_index": 13,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.615,
      "request_index": 11,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.613,
      "request_index": 16,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.616,
      "request_index": 10,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.615,
      "request_index": 14,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.616,
      "request_index": 9,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.934,
      "request_index": 15,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.941,
      "request_index": 17,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.941,
      "request_index": 18,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    },
    {
      "content": "",
      "finish_reason": "tool_calls",
      "content_chars": 0,
      "content_excerpt": "",
      "reasoning_field": "reasoning",
      "reasoning_chars": 48,
      "reasoning_excerpt": "We need to call get_weather with city \"Oakland\".",
      "reasoning_tokens": null,
      "usage": {
        "prompt_tokens": 4933,
        "total_tokens": 4974,
        "completion_tokens": 41,
        "prompt_tokens_details": null
      },
      "test": "tools",
      "seconds": 1.942,
      "request_index": 19,
      "passed": true,
      "validation_detail": "valid tool_call get_weather(city='Oakland')"
    }
  ],
  "evidence_policy": {
    "reasoning": "required",
    "allowed_finish_reasons": [
      "stop",
      "tool_calls"
    ],
    "errors": []
  },
  "passed": true
}