{
  "format": "proofrun-publication",
  "schemaVersion": 5,
  "contentHash": "sha256:73d9a4812694fff69905a145f865dd3470d50074494d3292772d1ad7f6c49371",
  "preparedAt": "2026-08-28T10:06:17.518Z",
  "editorial": {
    "format": "editorial-feature",
    "slug": "three-attempt-records-one-model-outcome-qwen3-8-flash-scores-76-on-triage-desk",
    "headline": "Three Attempt Records, One Model Outcome: Qwen3.8 Flash Scores 76 on Triage Desk",
    "dek": "After a Proofrun persistence defect and a provider 429 produced no submissions, the only valid artifact passed all ten frozen requirements and earned 76/100—showing why attempt provenance belongs beside model judgment.",
    "status": "draft",
    "publishedAt": null
  },
  "experiment": {
    "protocol": "proofrun-v0.8",
    "protocolId": "proofrun-review",
    "protocolRevision": 3,
    "protocolHash": "sha256:6bda63bb3ec8881dc9d84ffdc1bb1e4fbfd3faee9d850680bfaf862f20a5e1d6",
    "mode": "solo",
    "runOutcome": "solo_artifact",
    "runId": "PR-MTBS8SW7-BF",
    "startedAt": "2026-08-27T17:14:26.228Z",
    "identityState": "revealed",
    "evaluationState": "locked",
    "legacyProtocolException": false,
    "blindOrder": {
      "scheme": "sha256-128bit-assignment-commit-reveal-v1",
      "commitment": "sha256:66050c3df8368975aa7e8b516bac8e294b795b2b686630ae10506b6d20a0450e",
      "secret": "AEC3D4464D5B704092C2F934FD94F564",
      "verified": true
    }
  },
  "test": {
    "id": "triage-desk",
    "title": "Triage Desk",
    "category": "Product workflow",
    "summary": "A compact support-operations workspace that tests information architecture, validation, and synchronized multi-step state.",
    "prompt": "Build a responsive one-page support operations workspace called TRIAGE DESK for a small software team. Seed exactly four incidents: INC-204, Checkout timeout, Northstar Bikes, Critical, Web, 6 min, Unassigned; INC-203, Duplicate invoice, Kanso Studio, High, Email, 18 min, Noor Patel; INC-202, Export stuck at 99%, Fieldnote Labs, Medium, Chat, 42 min, Imani Brooks; and INC-201, Cannot rename workspace, Alder & Co, Low, Email, 1 hr, Leo Martins. All four begin Open. Make urgency and current queue state easy to scan. Include text search, status filters for All, Open, and Resolved, and a severity filter; combinations must work together and show a clear empty-result state. Selecting a queue item must open a useful incident detail with the incident facts, a short customer report, and an activity timeline. In the detail, let the operator change severity, choose among Unassigned, Mira Chen, Noor Patel, Imani Brooks, and Leo Martins, write an internal note, and resolve the incident. Resolution must be blocked when the note is blank, with visible guidance. A successful resolution must preserve the note in the timeline, show a clear Resolved state, and update the queue, filters, and counts without losing the selected incident. Keep the workflow useful at narrow mobile widths and support keyboard operation with visible focus and properly associated labels. For standardized evidence, use these exact hooks: id=\"ticket-search\" on the search input, id=\"severity-select\" on the detail severity select with lowercase option values, id=\"owner-select\" on the owner select with value \"mira\" for Mira Chen, id=\"internal-note\" on the note field, and id=\"incident-status\" on the selected incident's visible status. Give the selected incident control an accessible name containing its ID and title, and give the final action the accessible name \"Resolve incident\". Use one self-contained offline HTML file with vanilla CSS and JavaScript, with no external assets, fonts, libraries, APIs, or network requests.",
    "origin": "curated",
    "version": 1,
    "hash": "curated:triage-desk:v1",
    "scope": "standard",
    "runtime": "10-20 min",
    "executionMode": "web-single-file-v2",
    "signals": [
      "Workflow judgment",
      "Information architecture",
      "State consistency",
      "Accessibility",
      "Responsive product craft"
    ],
    "integrity": "versioned"
  },
  "settings": {
    "reasoningEffort": "high",
    "maxTokens": 40000,
    "completionBudget": {
      "version": "artifact-first-v2",
      "totalCompletionTokens": 40000,
      "visibleArtifactBudgetTokens": 24000,
      "reasoning": {
        "mode": "explicit-max",
        "effort": "high",
        "maxTokens": 16000,
        "providerEffort": null
      },
      "artifactBudgetEnforcement": "guaranteed"
    },
    "temperature": 0.7,
    "htmlStartDeadlineMinutes": 20,
    "writingGraceMinutes": 10,
    "maxCostMicrousdPerAttempt": 30000
  },
  "verdict": {
    "winner": null,
    "winnerModel": null,
    "confidence": "high",
    "rationale": "Comparable overall polish to Terra, but the hierarchy and queue/detail layout feel less resolved and require slightly more effort to scan.",
    "scores": {
      "A": 76,
      "B": null
    },
    "evaluationVersion": 3,
    "scoreConstruct": "craft-score",
    "compliance": {
      "A": 100,
      "B": null
    },
    "identityGuess": null
  },
  "requirements": [
    {
      "id": "requirement-1",
      "label": "Exactly the four frozen incidents and their supplied facts are present, with every incident initially Open.",
      "A": "pass",
      "B": "unscored"
    },
    {
      "id": "requirement-2",
      "label": "The queue makes urgency, assignment, status, and useful aggregate counts easy to scan.",
      "A": "pass",
      "B": "unscored"
    },
    {
      "id": "requirement-3",
      "label": "Text search, status filters, and severity filtering compose correctly and include a clear empty-result state.",
      "A": "pass",
      "B": "unscored"
    },
    {
      "id": "requirement-4",
      "label": "Selecting an incident keeps the queue context visible and reveals its facts, customer report, and activity timeline.",
      "A": "pass",
      "B": "unscored"
    },
    {
      "id": "requirement-5",
      "label": "Severity and owner changes remain synchronized between the selected detail, queue, filters, and counts.",
      "A": "pass",
      "B": "unscored"
    },
    {
      "id": "requirement-6",
      "label": "Attempting to resolve with a blank internal note is blocked and produces visible, actionable guidance.",
      "A": "pass",
      "B": "unscored"
    },
    {
      "id": "requirement-7",
      "label": "A successful resolution preserves the note in the timeline and clearly changes the incident to Resolved.",
      "A": "pass",
      "B": "unscored"
    },
    {
      "id": "requirement-8",
      "label": "After resolution, the selected detail, queue row, filters, and aggregate counts remain mutually consistent.",
      "A": "pass",
      "B": "unscored"
    },
    {
      "id": "requirement-9",
      "label": "The workflow remains usable at narrow widths and provides semantic labels, keyboard operation, and visible focus.",
      "A": "pass",
      "B": "unscored"
    },
    {
      "id": "requirement-10",
      "label": "The artifact is one responsive, self-contained HTML document that works offline without external dependencies.",
      "A": "pass",
      "B": "unscored"
    }
  ],
  "criteria": [
    {
      "id": "workflow-reasoning",
      "label": "Workflow reasoning",
      "description": "Whether selection, assignment, notes, validation, resolution, and dependent state changes form one coherent operational flow.",
      "weight": 30,
      "anchors": {
        "low": "Disconnected controls or broken state transitions",
        "mid": "The main path works with omissions or weak feedback",
        "high": "Every transition is coherent, guarded, and clearly communicated"
      },
      "A": 7,
      "B": 0
    },
    {
      "id": "information-design",
      "label": "Information design",
      "description": "Scanability, hierarchy, density, prioritization, and the relationship between the queue and incident detail.",
      "weight": 25,
      "anchors": {
        "low": "Urgency and context are difficult to find",
        "mid": "Readable structure with some clutter or ambiguity",
        "high": "Dense operational information feels immediate and effortless"
      },
      "A": 6,
      "B": 0
    },
    {
      "id": "interaction-accessibility",
      "label": "Interaction & accessibility",
      "description": "Control clarity, keyboard use, focus visibility, responsive adaptation, validation, and useful feedback.",
      "weight": 25,
      "anchors": {
        "low": "Important actions are hidden, inaccessible, or confusing",
        "mid": "Usable across the core path with rough edges",
        "high": "Clear, accessible, responsive, and confidence-inspiring throughout"
      },
      "A": 9,
      "B": 0
    },
    {
      "id": "state-robustness",
      "label": "State robustness",
      "description": "Consistency of incident data, filters, counts, selected detail, and edge cases as the workflow changes.",
      "weight": 20,
      "anchors": {
        "low": "Repeatable state drift, data loss, or runtime failure",
        "mid": "Core state remains sound with minor inconsistencies",
        "high": "All dependent views remain synchronized under scrutiny"
      },
      "A": 10,
      "B": 0
    }
  ],
  "contestants": [
    {
      "side": "A",
      "model": "qwen/qwen3.8-flash",
      "resolvedModel": "qwen/qwen3.8-flash",
      "provider": "Alibaba",
      "status": "complete",
      "artifactOutcome": "completed_submission",
      "selectedAttemptOrdinal": 3,
      "routing": {
        "mode": "pinned",
        "endpointSlug": "alibaba",
        "allowFallbacks": false,
        "resolvedProvider": "Alibaba",
        "resolvedEndpoint": "alibaba"
      },
      "screenshot": {
        "path": "screenshots/contestant-a-c-users-burnt-downloads-pr-mtbs8sw7-bf-c-99314880.png",
        "capturedAt": "2026-08-28T07:35:11.123Z",
        "width": 1574,
        "height": 1404,
        "byteLength": 371644,
        "method": "reviewer-upload-v1",
        "rendererVersion": null,
        "cameraProvenance": null,
        "camera": null,
        "recipeReceipt": null
      },
      "scene": null,
      "artifact": {
        "format": "proofrun-web-artifact",
        "schemaVersion": 1,
        "side": "A",
        "path": "artifacts/contestant-a.html",
        "contentHash": "sha256:5551e3d7dfc955ee53e78077f9f3f8d46c9ca595742d4236f841562c31e73145",
        "byteLength": 48973,
        "mediaType": "text/html;charset=utf-8",
        "executionMode": "web-single-file-v2",
        "sandboxProfile": "proofrun-public-web-v1"
      },
      "evidenceImages": [
        {
          "id": "EV-7fbe472f-968c-477c-b676-bdb27a8874ec",
          "path": "screenshots/contestant-a-resolved-critical-incident-7a8874ec.png",
          "label": "Resolved critical incident",
          "caption": "The frozen recipe searched for INC-204, opened it, assigned Mira Chen, recorded the standardized internal note, resolved the incident, and captured the synchronized queue and detail state.",
          "provenance": "harness-recipe",
          "state": "demonstration",
          "primary": false,
          "capturedAt": "2026-08-28T07:07:05.902Z",
          "width": 1280,
          "height": 720,
          "byteLength": 179903,
          "method": "browser-paint-evidence-recipe-v1",
          "rendererVersion": null,
          "cameraProvenance": null,
          "camera": null,
          "recipeReceipt": {
            "schemaVersion": 1,
            "id": "ER-mtclzlua",
            "recipeId": "triage-desk-resolved-incident-demonstration",
            "recipeVersion": 1,
            "recipeHash": "sha256:d2d100f5e92c755ee8a84a6f31ed8a89324405c4898bcea302ece859c3077261",
            "testId": "triage-desk",
            "side": "A",
            "status": "succeeded",
            "startedAt": "2026-08-28T07:07:03.591Z",
            "completedAt": "2026-08-28T07:07:05.910Z",
            "viewport": {
              "width": 1280,
              "height": 720,
              "deviceScale": 1
            },
            "resetBeforeStart": true,
            "actionTranscript": [
              {
                "id": "reset-before-start",
                "label": "Reset before start",
                "type": "reset",
                "status": "passed",
                "detail": "Activated ×."
              },
              {
                "id": "search-checkout",
                "label": "Find the checkout incident",
                "type": "type",
                "status": "passed",
                "detail": "Entered the frozen recipe text into #ticket-search."
              },
              {
                "id": "select-inc-204",
                "label": "Open INC-204",
                "type": "activate-control",
                "status": "passed",
                "detail": "Activated the matching INC-204CriticalCheckout timeoutNorthstar Bikes · Web · 6 min ago—UnassignedOpen control."
              },
              {
                "id": "assign-mira",
                "label": "Assign Mira Chen",
                "type": "set-control",
                "status": "passed",
                "detail": "Set #owner-select to the frozen recipe value."
              },
              {
                "id": "write-note",
                "label": "Record the internal resolution note",
                "type": "type",
                "status": "passed",
                "detail": "Entered the frozen recipe text into #internal-note."
              },
              {
                "id": "resolve-incident",
                "label": "Resolve INC-204",
                "type": "activate-control",
                "status": "passed",
                "detail": "Activated the matching Resolve incident control."
              }
            ],
            "stabilization": {
              "rule": "settled-window",
              "waitMs": 500,
              "animationFrames": 0,
              "status": "passed",
              "detail": "Waited 500ms and 0 animation frames before assertion and capture."
            },
            "assertions": [
              {
                "id": "resolved-present",
                "label": "Selected incident shows Resolved",
                "required": true,
                "status": "passed",
                "detail": "The frozen element contains the expected text."
              },
              {
                "id": "owner-selected",
                "label": "Mira Chen remains selected",
                "required": true,
                "status": "passed",
                "detail": "The frozen control has the expected value."
              },
              {
                "id": "note-present",
                "label": "Resolution note is visible",
                "required": true,
                "status": "passed",
                "detail": "The frozen text pattern is present in the rendered artifact."
              },
              {
                "id": "state-changed",
                "label": "Resolved workflow differs from the initial queue",
                "required": true,
                "status": "passed",
                "detail": "The post-recipe visual fingerprint differs from the reset state."
              },
              {
                "id": "runtime-clean",
                "label": "No new runtime errors during the recipe",
                "required": true,
                "status": "passed",
                "detail": "No new runtime errors were captured during the recipe."
              }
            ],
            "capture": {
              "state": "demonstration",
              "label": "Resolved critical incident",
              "caption": "The frozen recipe searched for INC-204, opened it, assigned Mira Chen, recorded the standardized internal note, resolved the incident, and captured the synchronized queue and detail state.",
              "publicationPrimaryEligible": true,
              "method": "browser-paint-evidence-recipe-v1",
              "capturedAt": "2026-08-28T07:07:05.902Z"
            },
            "failure": null
          }
        },
        {
          "id": "EV-da942f4f-73db-4706-a084-38da99314880",
          "path": "screenshots/contestant-a-c-users-burnt-downloads-pr-mtbs8sw7-bf-c-99314880.png",
          "label": "C Users burnt Downloads PR MTBS8SW7 BF contestant A.html",
          "caption": "Reviewer-supplied image; initial full queue, full page.",
          "provenance": "reviewer-upload",
          "state": "uploaded",
          "primary": true,
          "capturedAt": "2026-08-28T07:35:11.123Z",
          "width": 1574,
          "height": 1404,
          "byteLength": 371644,
          "method": "reviewer-upload-v1",
          "rendererVersion": null,
          "cameraProvenance": null,
          "camera": null,
          "recipeReceipt": null
        },
        {
          "id": "EV-e564368b-1c43-4151-b2fb-856c251e14f6",
          "path": "screenshots/contestant-a-c-users-burnt-downloads-pr-mtbs8sw7-bf-c-251e14f6.png",
          "label": "C Users burnt Downloads PR MTBS8SW7 BF contestant A.html (1)",
          "caption": "Reviewer-supplied image; mobile view, full page.",
          "provenance": "reviewer-upload",
          "state": "uploaded",
          "primary": false,
          "capturedAt": "2026-08-28T07:35:37.987Z",
          "width": 800,
          "height": 2556,
          "byteLength": 475429,
          "method": "reviewer-upload-v1",
          "rendererVersion": null,
          "cameraProvenance": null,
          "camera": null,
          "recipeReceipt": null
        }
      ],
      "generation": {
        "startedAt": "2026-08-28T06:52:22.740Z",
        "finishedAt": "2026-08-28T06:55:40.008Z",
        "wallTimeMs": 197268,
        "firstActivityMs": 5071,
        "htmlStartMs": 78247,
        "lastActivityMs": 197215,
        "activityChunks": 6035,
        "finishReason": "stop",
        "reasoningEffort": "high"
      },
      "usage": {
        "inputTokens": 612,
        "visibleTokens": 15307,
        "reasoningTokens": 6555,
        "cachedTokens": 0,
        "totalTokens": 22474,
        "cost": 0.01036694,
        "costSource": "openrouter-reported"
      },
      "evidence": {
        "artifactIssues": [],
        "runtimeErrors": [],
        "previewAudit": {
          "textCharacters": 887,
          "visibleElements": 141,
          "bodyChildren": 5,
          "documentCharacters": 276187
        },
        "probe": {
          "status": "passed",
          "durationMs": 606,
          "viewport": {
            "width": 583,
            "height": 520
          },
          "actions": [
            {
              "id": "resize",
              "label": "Narrow viewport",
              "status": "passed",
              "detail": "583 x 520 viewport exercised"
            },
            {
              "id": "click",
              "label": "Single click",
              "status": "passed",
              "detail": "Dispatched one complete pointer and mouse click sequence"
            },
            {
              "id": "drag",
              "label": "Short drag",
              "status": "passed",
              "detail": "Dispatched a bounded six-step diagonal drag"
            },
            {
              "id": "burst",
              "label": "Five-point burst",
              "status": "passed",
              "detail": "Dispatched five bounded clicks across the interaction surface"
            },
            {
              "id": "reset",
              "label": "Reset control",
              "status": "passed",
              "detail": "Activated clear filters"
            }
          ],
          "issues": [],
          "errors": []
        }
      }
    }
  ],
  "attempts": [
    {
      "side": "A",
      "ordinal": 1,
      "retryOfOrdinal": null,
      "lifecycle": "terminal",
      "artifactOutcome": "no_submission",
      "terminalCause": "protocol_exception",
      "requestedModel": "qwen/qwen3.8-flash",
      "routingMode": "pinned",
      "requestedEndpoint": "alibaba",
      "resolvedProvider": null,
      "resolvedEndpoint": null,
      "resolvedModel": null,
      "failure": {
        "source": "proofrun-server",
        "errorType": "harness_persistence_defect",
        "httpStatus": null,
        "providerCode": null,
        "message": "Proofrun could not persist the provider's immediate error because revision-3 routing normalization omitted the frozen reasoning-budget receipt during the terminal update. No model artifact, provider generation identifier, usage, or resolved-provider metadata was preserved; this attempt is excluded from model-outcome analysis.",
        "retriable": null,
        "classifiedBy": "proofrun",
        "observedAt": "2026-08-28T05:08:49.321Z"
      },
      "startedAt": "2026-08-27T17:14:26.464Z",
      "finishedAt": "2026-08-28T05:08:49.321Z"
    },
    {
      "side": "A",
      "ordinal": 2,
      "retryOfOrdinal": 1,
      "lifecycle": "terminal",
      "artifactOutcome": "no_submission",
      "terminalCause": "provider_incident_or_capacity",
      "requestedModel": "qwen/qwen3.8-flash",
      "routingMode": "pinned",
      "requestedEndpoint": "alibaba",
      "resolvedProvider": null,
      "resolvedEndpoint": null,
      "resolvedModel": null,
      "failure": {
        "source": "openrouter",
        "errorType": null,
        "httpStatus": 429,
        "providerCode": null,
        "message": "Provider returned error",
        "retriable": null,
        "classifiedBy": "proofrun",
        "observedAt": "2026-08-28T06:46:44.242Z"
      },
      "startedAt": "2026-08-28T06:46:39.841Z",
      "finishedAt": "2026-08-28T06:46:44.242Z"
    },
    {
      "side": "A",
      "ordinal": 3,
      "retryOfOrdinal": 2,
      "lifecycle": "terminal",
      "artifactOutcome": "completed_submission",
      "terminalCause": "normal",
      "requestedModel": "qwen/qwen3.8-flash",
      "routingMode": "pinned",
      "requestedEndpoint": "alibaba",
      "resolvedProvider": "Alibaba",
      "resolvedEndpoint": "alibaba",
      "resolvedModel": "qwen/qwen3.8-flash",
      "failure": null,
      "startedAt": "2026-08-28T06:52:22.740Z",
      "finishedAt": "2026-08-28T06:55:40.008Z"
    }
  ],
  "protocolExceptions": [
    {
      "code": "harness-attempt-persistence-defect",
      "detail": "Proofrun could not persist the provider's immediate error because revision-3 routing normalization omitted the frozen reasoning-budget receipt during the terminal update. No model artifact, provider generation identifier, usage, or resolved-provider metadata was preserved; this attempt is excluded from model-outcome analysis. The corrected attempt, if separately approved, must remain linked to this attempt under PREREG-001 and use the unchanged frozen condition.",
      "occurredAt": "2026-08-28T05:08:49.321Z",
      "attempt": {
        "side": "A",
        "ordinal": 1
      }
    },
    {
      "code": "corrected-attempt-launch-authorized",
      "detail": "LA-2B1EE5B6DFFE5505 records explicit approval of sha256:edcddcc147155cb6c16fab92b418df1eea79e1700c7bafca8dbc0a9564aed17c for one linked corrected attempt (PA-PR-MTBS8SW7-BF-A-02), capped at 0.030000 USD. The credential is not preserved.",
      "occurredAt": "2026-08-28T06:46:39.623Z",
      "attempt": null
    },
    {
      "code": "provider-incident-retry-authorized",
      "detail": "LA-CD069F628121E70B records explicit approval of sha256:edcddcc147155cb6c16fab92b418df1eea79e1700c7bafca8dbc0a9564aed17c for one linked provider-incident retry (PA-PR-MTBS8SW7-BF-A-03), capped at 0.030000 USD. The credential is not preserved.",
      "occurredAt": "2026-08-28T06:52:22.674Z",
      "attempt": null
    }
  ],
  "reviewAmendments": [],
  "disclosure": {
    "rawModelResponsesIncluded": false,
    "contestantSourceIncluded": true,
    "privateReviewerNotesIncluded": false,
    "credentialsIncluded": false,
    "generatedHtmlExecutedByPublicRecord": false,
    "generatedHtmlExecutedByJournal": false,
    "generatedHtmlAvailableInIsolatedSandbox": true,
    "rawSceneScriptIncluded": false,
    "compiledVoxelSceneIncluded": false,
    "editorialAssistance": {
      "evidenceAnalysis": {
        "model": "openai/gpt-5.6-sol",
        "promptVersion": "evidence-analyst-v3",
        "blind": true,
        "cost": 0.241287,
        "acceptedFindings": 10,
        "rejectedFindings": 0,
        "calibration": {
          "promptVersion": "evaluator-calibration-v2",
          "repeatConsistency": 0,
          "swapConsistency": 0,
          "swapDivergence": 1,
          "rating": "insufficient",
          "createdAt": "2026-08-28T07:40:48.741Z"
        }
      },
      "articleDrafting": {
        "model": "openai/gpt-5.6-sol",
        "promptVersion": "human-revision-v1",
        "cost": 0.109992,
        "humanRevised": true,
        "humanApproved": true
      }
    }
  }
}