{
  "schema_version": "one-dollar-showcase-v1",
  "version": 1,
  "checked_at": "2026-10-07T11:50:11.815Z",
  "status": "reviewed-results-with-access-gaps",
  "generation_requests": 3,
  "full_car": "not-run",
  "freeze_time": "2026-10-07T10:34:49Z",
  "frozen_inputs": [
    {
      "url": "/showcase/one-dollar-ai/v1/prompts/cat-runner.md",
      "sha256": "95df056b88bc57af45a9a986576c4af97afdb79ffc88f75d87a1e70477fac7e0"
    },
    {
      "url": "/showcase/one-dollar-ai/v1/prompts/subscription-calculator.md",
      "sha256": "37150048f526fef75c50de5822c7a97661dde5cedde87c08e26f192296eacb50"
    },
    {
      "url": "/showcase/one-dollar-ai/v1/prompts/packing-list.md",
      "sha256": "bac51102cbccba8e7ef9c4ba68a8f967ee7ae46d75f038f4b840d0503e1ca7ae"
    },
    {
      "url": "/showcase/one-dollar-ai/v1/prompts/packing-list-contract.json",
      "sha256": "a453b234ac5e79b5a613da8b0de64903ee202757f3aff9a1ebe8253bebcf2ee5"
    },
    {
      "url": "/showcase/one-dollar-ai/v1/prompts/freeze-manifest.json",
      "sha256": "aa3c3058dd6caf85f8bd96f24de92b701d1c855decceba91faf7c6e52644b83e"
    },
    {
      "url": "/showcase/one-dollar-ai/v1/prompts/generation-profile.md",
      "sha256": "cc33299c8828299e9191ebbe96bbb7057a1e91ebad49005ccd66792091f91864"
    }
  ],
  "hidden_answer_key_sha256": "ab1ed3c2d23aa6d37ba2bbf2b030648a9e80c2e9b15ed475420f0e6c7ede27da",
  "original_candidates": [
    {
      "lane_id": "deepseek-flash",
      "candidate": "DeepSeek V4.1 Flash",
      "model_id": "deepseek-flash",
      "provider": "DeepSeek direct API",
      "proposed_route": "deepseek/deepseek-flash",
      "access": "verified-existing-allocation",
      "allocation_evidence": "read-only official balance endpoint confirmed sufficient existing USD allocation; values private",
      "publication_permission": "reviewed API terms; AI-labelled outputs; no endorsement",
      "source": "https://api-docs.deepseek.com/quick_start/pricing/"
    },
    {
      "lane_id": "mimo-flash",
      "candidate": "MiMo V2.6 Flash",
      "model_id": "mimo-v2.6-flash",
      "provider": "Xiaomi direct API",
      "proposed_route": null,
      "access": "no-configured-direct-credential",
      "allocation_evidence": "no direct allocation verified",
      "publication_permission": "not-reviewed for direct route",
      "source": "https://mimo.mi.com/docs/en-US/price/pay-as-you-go"
    },
    {
      "lane_id": "kimi-k3",
      "candidate": "Kimi K3",
      "model_id": "kimi-k3",
      "provider": "Kimi direct API",
      "proposed_route": null,
      "access": "no-configured-direct-credential",
      "allocation_evidence": "NVIDIA endpoint mentioned by owner; no public-output-eligible allocation verified",
      "publication_permission": "NVIDIA trial output excluded from production; direct Kimi route not-reviewed",
      "source": "https://platform.kimi.ai/docs/guide/kimi-k3-quickstart"
    }
  ],
  "route_freeze_state": "deepseek-direct-frozen-before-generation; other lanes unavailable",
  "proposed_substitutes": [
    {
      "model": "opencode/mimo-v2.6-flash-free",
      "name": "MiMo V2.6 Flash Free",
      "status": "workflow-permission-unresolved",
      "attempts": 0,
      "free_access": "temporary",
      "source": "https://opencode.ai/docs/zen/"
    },
    {
      "model": "opencode/longcat-2.5-preview-free",
      "name": "LongCat 2.5 Preview Free",
      "status": "workflow-permission-unresolved",
      "attempts": 0,
      "free_access": "temporary",
      "source": "https://opencode.ai/docs/zen/"
    },
    {
      "model": "opencode/big-pickle",
      "name": "Big Pickle",
      "status": "workflow-permission-unresolved",
      "attempts": 0,
      "free_access": "temporary",
      "source": "https://opencode.ai/docs/zen/"
    }
  ],
  "generation_settings": {
    "provider_default_sampling_and_effort": "overrides omitted by verified local adapter; documented provider defaults are inferred",
    "documented_thinking": "enabled",
    "documented_effort": "high",
    "output_cap": 8192,
    "tools": "disabled",
    "automatic_retries": "second dispatch blocked before provider request",
    "cli_version": "1.18.33"
  },
  "usage": {
    "state": "reported-per-attempt",
    "records": [
      "/showcase/one-dollar-ai/v1/records/cat-runner.json",
      "/showcase/one-dollar-ai/v1/records/calculator.json",
      "/showcase/one-dollar-ai/v1/records/packing-list.json"
    ]
  },
  "budget": {
    "currency": "USD",
    "per_original_lane": 1,
    "overall": 3,
    "reported_generation_spend": 0.0145671,
    "reconciled_generation_spend": null,
    "pending_generation_reservations": 0.973212,
    "unused_experiment_ceiling_lower_bound": 2.026788,
    "provider_inference_allowance": null,
    "charge_evidence": "reported telemetry; unreconciled charges retain full reservations",
    "administrative_calls": {
      "successful_read_only_balance_requests": 1,
      "billed_auxiliary_generation_calls": 0
    },
    "deposits": 0,
    "automatic_topups_enabled_by_this_experiment": false,
    "per_lane_headroom_lower_bound": {
      "deepseek-flash": 0.026788,
      "mimo-flash": 1,
      "kimi-k3": 1
    },
    "allowance_note": "Experiment headroom only; unavailable lanes cannot subsidize DeepSeek. Remaining provider inference allowance unknown."
  },
  "spend_label": "$0.014567 reported*",
  "review_label": "1.81 agent minutes",
  "accepted_label": "1/3 task results",
  "output_review_minutes": 1.81,
  "contribution_record": [
    "Briefs, acceptance checks, fictional inputs and hidden key frozen before generation.",
    "Shared generation profile frozen before generation and published separately.",
    "Bun safety tooling, validators and editorial website authored by Codex; these are not tested model artifacts.",
    "One independent initial attempt for each task through the recorded DeepSeek route.",
    "Reviewed nonempty source, if any, published without editing; failures and repair limits visible.",
    "Missing direct lanes and unresolved free-route permission retained separately; no model ranking inferred."
  ],
  "browser_artifact_checks": "Chromium 148 desktop and emulated mobile; exact checks and limits in each task record",
  "sources": [
    {
      "url": "https://opencode.ai/docs/zen/",
      "class": "vendor",
      "checked_at": "2026-10-07T10:37:39.799Z"
    },
    {
      "url": "https://opencode.ai/legal/terms-of-service",
      "class": "vendor",
      "effective_date": "2026-08-15",
      "permission_assessment": "unresolved"
    },
    {
      "url": "https://cdn.deepseek.com/policies/en-US/deepseek-terms-of-use.html",
      "class": "vendor",
      "last_update": "2026-03-27"
    },
    {
      "url": "https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf",
      "class": "vendor",
      "section": "1.2"
    },
    {
      "url": "https://cdn.deepseek.com/policies/en-US/deepseek-open-platform-terms-of-service.html",
      "class": "vendor",
      "effective_date": "2026-04-29",
      "checked_at": "2026-10-07T11:50:11.815Z"
    }
  ],
  "tasks": [
    {
      "id": "cat-runner",
      "kind": "game",
      "title": "Cat runner",
      "description": "One HTML file: a fenced path, flowerpots, a fixed obstacle sequence and a cat that jumps, collides and restarts.",
      "prompt": "/showcase/one-dollar-ai/v1/prompts/cat-runner.md",
      "results": [
        {
          "label": "DeepSeek V4.1 Flash · direct API",
          "route": "deepseek/deepseek-flash",
          "outcome": "failed-no-artifact",
          "summary": "The response reached the 8,192-token ceiling with all tokens spent on reasoning and returned no HTML. No game or clip can be offered.",
          "accepted": false,
          "reviewed_source": false,
          "attempts": 1,
          "attempt_label": "1 initial attempt · 0 repairs",
          "spend_label": "$0.005035 reported · unreconciled",
          "review_label": "0.30 agent minutes",
          "review_minutes": 0.3,
          "usage": {
            "state": "reported",
            "accounting": {
              "generatedTokens": 8192,
              "billedOutputTokens": 8192,
              "separatelyBilledReasoningTokens": 0
            },
            "tokens": {
              "total": 8990,
              "input": 798,
              "output": 0,
              "reasoning": 8192,
              "cacheRead": 0,
              "cacheWrite": 0
            }
          },
          "source": null,
          "source_sha256": null,
          "artifact": null,
          "record": "/showcase/one-dollar-ai/v1/records/cat-runner.json",
          "binary_checks": [
            {
              "name": "Complete offline HTML and visible game",
              "status": "fail",
              "detail": "Zero source bytes; no HTML was returned."
            },
            {
              "name": "Keyboard, click and touch jumping",
              "status": "not-run",
              "detail": "No artifact."
            },
            {
              "name": "Deterministic, clearable obstacles",
              "status": "not-run",
              "detail": "No artifact."
            },
            {
              "name": "Distance scoring and collision",
              "status": "not-run",
              "detail": "No artifact."
            },
            {
              "name": "Jump and landing squash/stretch",
              "status": "not-run",
              "detail": "No artifact."
            },
            {
              "name": "Complete instant reset",
              "status": "not-run",
              "detail": "No artifact."
            },
            {
              "name": "Narrow layout and keyboard restart",
              "status": "not-run",
              "detail": "No artifact."
            }
          ],
          "checks": [
            "fail: Complete offline HTML and visible game — Zero source bytes; no HTML was returned.",
            "not-run: Keyboard, click and touch jumping — No artifact.",
            "not-run: Deterministic, clearable obstacles — No artifact.",
            "not-run: Distance scoring and collision — No artifact.",
            "not-run: Jump and landing squash/stretch — No artifact.",
            "not-run: Complete instant reset — No artifact.",
            "not-run: Narrow layout and keyboard restart — No artifact.",
            "One initial generation request; no repair reached the model.",
            "A repair request was refused before generation: retained reservations would exceed the $1 original-candidate lane.",
            "Gameplay, mobile controls and ten-second clip: not-run because no accepted game exists."
          ],
          "subjective_observations": [],
          "reserved_max_usd": 0.324404,
          "repair_state": "requested-repair-refused-before-generation"
        }
      ]
    },
    {
      "id": "calculator",
      "kind": "calculator",
      "title": "Subscription calculator",
      "description": "Monthly and annual entries, cancellation choices and local saving, with arithmetic in USD cents.",
      "prompt": "/showcase/one-dollar-ai/v1/prompts/subscription-calculator.md",
      "results": [
        {
          "label": "DeepSeek V4.1 Flash · direct API",
          "route": "deepseek/deepseek-flash",
          "outcome": "failed-incomplete-html",
          "summary": "The response ended mid-expression inside an unclosed script block. Chromium displayed the static form but Add created no entry and the total stayed $0.00. The source is available for inspection; it is not offered as a working calculator.",
          "accepted": false,
          "reviewed_source": true,
          "attempts": 1,
          "attempt_label": "1 initial attempt · 0 repairs",
          "spend_label": "$0.005044 reported · unreconciled",
          "review_label": "0.31 agent minutes",
          "review_minutes": 0.3069660992666667,
          "usage": {
            "state": "reported",
            "accounting": {
              "generatedTokens": 8192,
              "billedOutputTokens": 8192,
              "separatelyBilledReasoningTokens": 0
            },
            "tokens": {
              "total": 9051,
              "input": 859,
              "output": 3485,
              "reasoning": 4707,
              "cacheRead": 0,
              "cacheWrite": 0
            }
          },
          "source": "/showcase/one-dollar-ai/v1/artifacts/deepseek-calculator.txt",
          "source_sha256": "376ae6eedbbf2032e4af67e138e2ee0467fc0476d04006ae72e50790e58b023a",
          "artifact": null,
          "record": "/showcase/one-dollar-ai/v1/records/calculator.json",
          "binary_checks": [
            {
              "name": "Usable add/remove, periods and cancellation controls",
              "status": "fail",
              "detail": "Unfinished script is not executed; Add creates no entries."
            },
            {
              "name": "$15/month plus $120/year equals $300/year",
              "status": "fail",
              "detail": "The calculator remains at $0.00 because entries cannot be added."
            },
            {
              "name": "Cancellation and removal update totals",
              "status": "not-run",
              "detail": "No working entry handlers."
            },
            {
              "name": "Reload preserves entries and cancellation",
              "status": "not-run",
              "detail": "No working save handlers."
            },
            {
              "name": "Invalid inputs preserve valid totals",
              "status": "not-run",
              "detail": "Unfinished artifact cannot validate inputs."
            },
            {
              "name": "Unavailable storage remains usable",
              "status": "not-run",
              "detail": "Unfinished artifact is already unusable."
            },
            {
              "name": "Integer-cent arithmetic, offline operation and narrow keyboard use",
              "status": "fail",
              "detail": "Source is incomplete. Static inspection also found no safe aggregate bound for many large entries."
            }
          ],
          "checks": [
            "fail: Usable add/remove, periods and cancellation controls — Unfinished script is not executed; Add creates no entries.",
            "fail: $15/month plus $120/year equals $300/year — The calculator remains at $0.00 because entries cannot be added.",
            "not-run: Cancellation and removal update totals — No working entry handlers.",
            "not-run: Reload preserves entries and cancellation — No working save handlers.",
            "not-run: Invalid inputs preserve valid totals — Unfinished artifact cannot validate inputs.",
            "not-run: Unavailable storage remains usable — Unfinished artifact is already unusable.",
            "fail: Integer-cent arithmetic, offline operation and narrow keyboard use — Source is incomplete. Static inspection also found no safe aggregate bound for many large entries.",
            "Exact generated source retained without repairs or editing; downloaded as plain text.",
            "Chromium 148 offline file: no external requests; no console/page errors because the unclosed script is not executed.",
            "Static layout fits 1440, 390 and 320 pixels; this does not establish calculator usability.",
            "A repair request was refused before generation: retained reservations would exceed the $1 original-candidate lane.",
            "Persistence, cancellation, invalid-input and unavailable-storage behavior remain not-run."
          ],
          "subjective_observations": [],
          "reserved_max_usd": 0.324404,
          "repair_state": "requested-repair-refused-before-generation"
        }
      ]
    },
    {
      "id": "packing-list",
      "kind": "transformation",
      "title": "Twenty messages → packing list",
      "description": "One fictional batch, three products, amendments, aliases, cancellations and missing history. Orders, CSV and quantities must agree.",
      "prompt": "/showcase/one-dollar-ai/v1/prompts/packing-list.md",
      "results": [
        {
          "label": "DeepSeek V4.1 Flash · direct API",
          "route": "deepseek/deepseek-flash",
          "outcome": "accepted",
          "summary": "One fictional batch passed the frozen contract: five orders, twelve CSV lines, all three product totals and coverage of all twenty messages. Missing history stays a clarification instead of an invented order.",
          "accepted": true,
          "reviewed_source": true,
          "attempts": 1,
          "attempt_label": "1 initial attempt · 0 repairs",
          "spend_label": "$0.004488 reported · unreconciled",
          "review_label": "1.20 agent minutes",
          "review_minutes": 1.2,
          "usage": {
            "state": "reported",
            "accounting": {
              "generatedTokens": 6940,
              "billedOutputTokens": 6940,
              "separatelyBilledReasoningTokens": 0
            },
            "tokens": {
              "total": 9101,
              "input": 2161,
              "output": 2429,
              "reasoning": 4511,
              "cacheRead": 0,
              "cacheWrite": 0
            }
          },
          "source": "/showcase/one-dollar-ai/v1/artifacts/deepseek-packing-list.json",
          "source_sha256": "ce8af4171a917a2808dd5d7bac769463e4787b32c2775a0487d82c2c661fb763",
          "artifact": "/showcase/one-dollar-ai/v1/artifacts/deepseek-packing-list.json",
          "record": "/showcase/one-dollar-ai/v1/records/packing-list.json",
          "binary_checks": [
            {
              "name": "Structured schema and catalog identities",
              "status": "pass",
              "detail": "Exactly seven required top-level fields, valid customer/product IDs and five sorted orders."
            },
            {
              "name": "Amendments, aliases and cancellations",
              "status": "pass",
              "detail": "Final line quantities, statuses and audit source IDs match the frozen key."
            },
            {
              "name": "Canonical CSV",
              "status": "pass",
              "detail": "Twelve rows; required order and quoting; no final newline."
            },
            {
              "name": "Product quantities",
              "status": "pass",
              "detail": "Open/cancelled totals: notebooks 5/1, cables 5/2, mugs 4/7."
            },
            {
              "name": "Missing-history clarification",
              "status": "pass",
              "detail": "M20 asks for order ID and current lines; no unsupported order is created."
            },
            {
              "name": "Every message covered once",
              "status": "pass",
              "detail": "M01 through M20 in order, with exact dispositions, order IDs and product references."
            },
            {
              "name": "Independent exact reconciliation and source identity",
              "status": "pass",
              "detail": "Offline validator passes; independent review matches raw text events to source bytes and reconciles the key."
            }
          ],
          "checks": [
            "pass: Structured schema and catalog identities — Exactly seven required top-level fields, valid customer/product IDs and five sorted orders.",
            "pass: Amendments, aliases and cancellations — Final line quantities, statuses and audit source IDs match the frozen key.",
            "pass: Canonical CSV — Twelve rows; required order and quoting; no final newline.",
            "pass: Product quantities — Open/cancelled totals: notebooks 5/1, cables 5/2, mugs 4/7.",
            "pass: Missing-history clarification — M20 asks for order ID and current lines; no unsupported order is created.",
            "pass: Every message covered once — M01 through M20 in order, with exact dispositions, order IDs and product references.",
            "pass: Independent exact reconciliation and source identity — Offline validator passes; independent review matches raw text events to source bytes and reconciles the key.",
            "One accepted batch counts as one task; this is one candidate and one fictional dataset.",
            "JSON has no executable code; CSV is extracted verbatim from its returned csv field.",
            "No comparative model ranking or general order-processing accuracy is established.",
            "Human review minutes and full CAR remain unmeasured/not-run."
          ],
          "subjective_observations": [],
          "reserved_max_usd": 0.324404,
          "repair_state": "no-repair-needed",
          "csv": "/showcase/one-dollar-ai/v1/artifacts/deepseek-packing-list.csv",
          "csv_sha256": "ebcce47c8670986fc109c2daa7185a238d60c163aa1656f970db433ea6aabcb8",
          "csv_derivation": "Verbatim UTF-8 extraction of the returned JSON csv string; no added newline."
        }
      ]
    }
  ],
  "generation_profile": {
    "frozenAt": "2026-10-07T11:25:13.266Z",
    "sha256": "cc33299c8828299e9191ebbe96bbb7057a1e91ebad49005ccd66792091f91864",
    "publicUrl": "/showcase/one-dollar-ai/v1/prompts/generation-profile.md"
  },
  "accepted_tasks": 1,
  "review_evidence_state": "recorded-subtotal",
  "untimed_review": "Additional main-agent inspection and editorial/QA setup; human review time unmeasured."
}