{
  "schema": "anvil-serving.benchmark-campaign-state/v1",
  "campaign_id": "2026-09-08-glm53-linux-wsl-comparison",
  "updated_at": "2026-09-08T07:13:50.406765+00:00",
  "status": "completed-with-failures",
  "current_stage": "completed",
  "launcher": {
    "version": "1.0.0",
    "revision": "8abcc5dc75a1b2389210b553120abb76e5864d3a",
    "module": "anvil_serving/cli.py",
    "branch": "codex/glm53-linux-benchmark-20260908",
    "initial_tracked_dirty": false,
    "verified": true,
    "invocation": "python3 -m anvil_serving.cli",
    "module_source": "anvil_serving/cli.py"
  },
  "stages": [
    {
      "id": "research",
      "status": "completed",
      "evidence": [],
      "failures": []
    },
    {
      "id": "feasibility",
      "status": "completed",
      "evidence": [],
      "failures": []
    },
    {
      "id": "scout",
      "status": "completed-with-failures",
      "evidence": [
        "controlled-scout-4k-r3.json",
        "linux-unique-natural-4k-r10.json"
      ],
      "failures": []
    },
    {
      "id": "finalist",
      "status": "not-run",
      "evidence": [],
      "failures": [
        "Strict scout and unique canary gates failed"
      ]
    },
    {
      "id": "quality",
      "status": "completed-with-failures",
      "evidence": [
        "context-summary.json",
        "summary.json"
      ],
      "failures": []
    },
    {
      "id": "restoration",
      "status": "completed",
      "evidence": [
        "restoration.json"
      ],
      "failures": []
    },
    {
      "id": "publication",
      "status": "completed",
      "evidence": [
        "publication-summary.md"
      ],
      "failures": []
    }
  ],
  "active_assignments": [],
  "completed_cells": [
    "direct-smoke",
    "direct-preflight-disabled (image phrase failure)",
    "direct-preflight-thinking-enabled",
    "historical capacity four contexts x3",
    "strict controlled-output scout failed 3/3",
    "unique natural 4K: 1/10 canary compliant; dependent cells stopped",
    "coding quality 15/15",
    "image corpus 12/12",
    "endurance60 passed",
    "routed disabled 9/10: nominal380k needle HTTP413",
    "routed260k needle passed",
    "routed enabled answers pass; required reasoning evidence fails",
    "agentic deep 30/30",
    "SWE official smoke resolved 1/1",
    "context quality128/150",
    "post-workload routed smoke/json pass; unchanged serving state"
  ],
  "next_actions": [],
  "promotion_authorized": false
}